Evaluation of the dimensional changes in the mandible, condyles, and the temporomandibular joint following skeletal class III treatment with chin cup and bonded maxillary bite block using low-dose computed tomography: A single-center, randomized controlled trial

Study design and registration

The research was designed as a single-center, 2-arm, parallel-group randomized controlled trial with a 1:1 allocation ratio and guided by the CONSORT statement.¹⁹ The study was conducted in the Department of Orthodontics at the University of Damascus Dental School between January 2019 to January 2022. This study was approved by the Regional Ethical Committee on Research of the Damascus University, Faculty of Dentistry, Syria (UDDS-2788-2016PG) on January 11, 2019. The study was registered at ClinicalTrails.gov (Identifier: NCT05350306) on 28 April, 2022 after the onset of the study. Retrospective registration was done as the regulations at the Univeristy of Damascus do not oblige researchers to register their trial protocols before the commencement of their research work. However, this protocol was registered retrospectively to publish the trial's results. This article is reported in line with the CONSORT guidelines.⁴⁴

Sample size estimation

Sample estimation was calculated using Minitab^® Version 18 (Minitab Inc., State College, Pennsylvania, USA). An independent t-test, statistical power of 85%, and a significance level of 0.05 were among the assumptions for this calculation. In total, 17 children were required in each group to detect a 1 mm significant difference in total mandibular length (Co-Gn) based on the standard deviation of this measurement for a previous study (0.92).²⁰ Four patients were added to the experimental and control groups (2 patients in each group) to compensate for potential dropouts.

Patient recruitment

A preliminary screening was performed of schoolchildren during the period from April 2019 to May 2019. Ten public and private schools were randomly selected from the database of all primary schools in Damascus city, Syria. Intra- and extra-oral clinical examination was performed by the principal investigator (A.H.H.) for 1700 children with an average age of 7 years (range: 6-8 years), including 767 female and 933 male children at the medical clinics of the selected schools. Initially, 90 children with class III malocclusion (52 boys, 43 girls) were expected to be suitable for inclusion in the trial. A formal invitation letter was sent to the parents of candidate children asking them to visit the Department of Orthodontics at the University of Damascus for further analysis and to assess the need for early correction of the deformity. All invited parents came for this in-depth assessment. The following inclusion criteria were used: age between 6 and 8, edge-to-edge incisor relationship or anterior crossbite, class III relationships at the permanent first molars or mesial-step relationships at primary second molars, absence of discrepancy between the centric relation and the maximum intercuspation position, short or normal-face pattern, normal or deep overbite, and no temporomandibular joint disorders or craniofacial anomalies. The radiological inclusion criteria were (1) mild to moderate mandibular prognathism (SNB > 80°, and 0 ≤ANB≥4) and (2) normal or horizontal growth pattern (Bjork’s sum ≤396° ±5°). Fifty seven patients met the inclusion criteria. Information sheets were given to parents of eligible patients. The need for two low-dose CT images, and the possibility of postponing treatment if the child fell into the control group were elaborately explained before taking their written informed consents. Six parents refused to participate in this trial. Therefore the resultant sampling frame included 51 patients. Simple random sampling was used to select 38 patients for this trial who were then assigned randomly to the chin cup and the control groups. The CONSORT flow diagram of patients’ recruitment, follow-up, and entry into data analysis is given in Figure 1.

Figure 1. CONSORT flow diagram of patients’ recruitment, follow-up and entry into data analysis.

Randomization and allocation concealment

An online randomization service (www.randomizer.org) was used to perform simple randomization with an allocation ratio of 1:1. The allocation sequence was concealed from the researcher (AH) using sequentially numbered, opaque, and sealed envelopes, which were opened after patients completed all baseline assessments. Blinding of either patient or operator was not possible. Blinding was applied for the outcome assessment only. After the mandibular mask was constructed by one of the co-authors (MYH), the row image was resliced by cutting the region of the overjet to blind the other investigator (AH), who completed the extraction of results.

Interventions

Experimental group: Chin-cup group

Each patient in the treatment group received a bonded maxillary bite block and an occipital chin cup (Figure 2 shows the bite block). The splint consisted of a 2 mm posterior acrylic cap reinforced by a metal framework. The traction force on the mandible by the chin cup was 400–500 g at each side, in the direction of the condyles.⁸ Patients were instructed to wear the appliance for 14 hours per day. The true overjet and overbite were measured monthly, subtraction the effects of the bite block on these variables based on the assumption that every 1 mm of the thickness of the posterior bite block increases the overjet by an average of 1.3 mm and decreases the overbite an average of 2 mm.²¹ After achieving 1 mm overjet and 2 mm overbite, the acrylic splint was removed. Then, the chin cup was used with a light force for approximately 8 hours per day as a retainer. The active treatment was considered complete when the following two conditions were met: a positive overjet of 2 to 4 mm, and an active treatment time of at least 16 months.

Figure 2. The bonded maxillary bite block used intra-orally in conjunction with the chincup appliance.

Control group

After acquiring low-dose CT images, the patients allocated to the control group received no clinical intervention. They were recalled 16 months after registration for collection of the second low-dose CT image records.

Outcome measures

Computed tomography acquisition

A multiplanar spiral CT machine (Philips brilliance 64, Philips Medical Systems, Best, the Netherlands) was used to obtain pre- (T1) and post-treatment (T2) CT images. The CT scans were performed at 80 kV and 100 mAs, one pitch, 2.5 mGy (CTDIvol), and a 1.25 mm slice thickness.¹⁷ The patients were seated in a supine position with their Frankfort horizontal (FH) plane perpendicular to the floor and their teeth in maximum intercuspation. Once the low-dose CT images were stored in the Digital Imaging and Communications in Medicine (DICOM) format, they were transferred into MIMICS 21.0 software (Materialise, Leuven, Belgium) to create 3D volumetric mandibular models.

Mandible and condyles reconstruction

Each image was resliced to make the head orientation uniform along the midsagittal and Frankfort horizontal planes. The basion, crista galli, and glabella landmarks defined the midsagittal plane. The Frankfort horizontal plane was defined by the right and left orbitale landmarks and the right porion landmark.²²

The image processing procedures are shown in Figure 3. Briefly, a bony mask was created using a global threshold value of 226 to 2976 Hounsfield units [HU] for the software algorithm. Semiautomatic segmentation using a split mask tool was utilized to isolate the mandibular bone. The mandible mask was rendered in a high-quality 3D model. Due to the differences in the lower permanent molar eruption, the crowns of the teeth were removed by a plane passing through 1 mm inferior to the alveolar bone and 10 mm distal to the second primary molars. Finally, the condyles were cut at neck level using a plane parallel to the Frankfort horizontal plane at the most inferior part of the sigmoid notch, and the post-treatment cut was parallel to the first cut.²³ The condyle and mandible volumes and surface sizes were automatically calculated in mm³ and mm², respectively.

Figure 3. Workflow of the image processing procedures.

Linear and angular measurements

A new template was created from the Measure and Analyze tool of Mimics™, which was used to define the anatomical landmarks and planes and produce the desired measurements (Tables 1 and 2).^22,24,25 Ikeda and Kawamura suggested that the condylar and glenoid fossa landmarks were made on the corrected sagittal view.²⁶ After identifying the landmarks, the software automatically calculated the distances and the angles. Then, the data was exported in.cvs format.

3D mandibular regional superimposition and comparison analysis

The paired resulting reconstructed mandibular models were exported to 3-Matic software (3-matic13.0, Materialise NV, Leuven, Belgium), and their surfaces were warped using the Warp tool of Mimics™. Initially, The T1 and T2 3D volumes were superimposed manually by approximating similar anatomical regions of the mandible using Interactive Translate and Rotate tools of the software, followed by automatic global registration in 3-Matic software (Figure 4). The superimposition was repeated three times to enhance accuracy and reproducibility. The point-based analysis was performed to assess the changes in 3D mandibular models between T1 and T2, and a color map was produced to assess the mandibular shape changes. The threshold was set at 2 mm: green areas indicated differences within 2 mm (between −2 and 2 mm), red surfaces indicated positive values displacement more than 2 mm, and blue surfaces indicated negative values displacement greater than −2 mm between two 3D models (Figure 5). Quantitative changes were done by reporting the mean, minimum, and maximum values of part analyses on a spreadsheet and used for comparative analyses

Figure 4. Isometric view of registered 3D mandibular models T1 (green mandible) and T2 (red mandible).

Figure 5. Isometric view of the registered 3D mandibular color-coded map after part analysis between T2-T1.

Interim analyses and stopping guidelines

No interim analyses were performed during this trial.

Statistical analysis

Statistical analyses were performed using IBM SPSS Statistics for Windows, version 26.0 (IBM Corp., NY, USA). The average value of each bilateral measurement was calculated to achieve the statistics of this study. Shapiro–Wilk test was used to test the normality of data. Accordingly, non-normally distributed data were analyzed by the Wilcoxon signed-rank test and the Mann–Whitney U test for in- and intergroup comparison, respectively. Additionally, a paired-samples t-test was used to evaluate the changes in the treatment group, and an independent t-test was used for the intergroup comparisons of parameters with a normal distribution. Bonferroni's correction was used to adjust the alpha level due to multiplicity. The adjusted alpha level was 0.0125 for the volumetric and surface area measurements and 0.005 for the cephalometric analysis. To assess the reliability of the measurements, 17 (25%) images were randomly selected and re-measured after a 2-week interval by the same examiner (AH). Intraclass correlation coefficients (ICCs) with the absolute agreement were used to assess intra-rater reliability. Error of the method was analyzed with Dahlberg's formula.²⁷

Reliability of the measurements

Intra-examiner reliability was high, and ICCs ranged from 0.83 to 0.999. The error of the method was acceptable for both linear and angular measurements, which were smaller than 0.5 mm and 0.6 degrees, respectively. The measurement error ranged from 18.84 to 54.3 mm³ for the volumetric measurements and 18.81 to 44.58 mm² for the surface area measurements (Table 3).

Basic sample characteristics

The CONSORT flow diagram of patients' recruitment and follow-up is given in Figure 1. In total, 38 patients (21 boys and 17 girls) were enrolled.⁴⁴ However, three patients were not included in the statistical analysis: one in each group was lost to follow-up, and the third one was extracted from the control group due to the low-quality low-dose CT image. Table 4 shows the basic sample characteristics. The mean treatment and observation periods were 16.67±0.52 months and 16.67±0.52 months, respectively. There were no significant differences in the age distribution and treatment/observation periods between the experimental and the control groups (p=0.675 and p=0.179, respectively).

Outcome measures

The mean mandibular volume increased significantly by 777.24 mm³ and 1,221.62 mm³ in the chin cup and control groups, respectively. Also, the mean condylar volume increased significantly by 94.57 mm³ and 132.54 mm³ in the chin cup and control groups, respectively (Table 5). The linear measurements of the mandible significantly increased in both groups (p<0.05). No statistically significant differences between the chin cup and control groups were observed regarding the volumes, superficial areas, and linear changes of the mandible and condyles (Tables 5 and 6). Although the mean mandibular angle decreased by 1.63° in the treatment group, there was no statistically significant difference between the treatment and the control groups regarding this variable (Table 6).

There was a significant posterior and superior displacement of the condyles following treatment by a mean of 1.29 mm (p=0.001) and 1.68 mm (p<0.001), respectively. Similar changes were observed in the glenoid fossa, but to a lesser extent by a mean of 0.82 mm posteriorly (p<0.001) and 0.81mm superiorly (p<0.001). The superior joint space (SJS) and posterior joint space (PJS) showed a significant mean decrease of 0.71 mm and 0.93 mm, respectively. However, no statistically significant differences were observed in the relative positions of the condyles, glenoid fossa, and joint spaces in the control group (p>0.05), except for a significantly superior displacement of the condyles (0.84 mm; p=0.001). The changes in the relative sagittal and vertical positions of condyles and glenoid fossa were significantly smaller in the treatment group than in the control group. No statistically significant differences were observed between the treatment and control groups regarding the average mean, maximum, and minimum part analysis of the point-based analysis (p>0.05; Table 7).

Harms

No serious harm was observed.

Limitations

One of the limitations of the current trial is its short-term evaluation. Secondly, the temporomandibular disk was not studied in this trial. However, this aim needs another imaging method.

Generalizability

Randomized controlled trials (RCTs) provide the highest level of evidence.¹⁰ Additionally, three-dimensional quantification gives a true mandibular, condyle, and glenoid fossa changes after orthodontic treatment by the chin cup.⁴³ Moreover, significant changes occur in the craniofacial region during the mixed dentition that can be utilized for orthodontic therapy.³² Our study designed as RCT with the help of the three-dimensional quantification and patients were in the early mixed dentation, as a result, the current research findings revealing the effects of the chincup are generalizable. In this randomized controlled trial, the chin cup treatment failed to control mandibular growth, resulting only in orientation to condylar positions and compression of joint spaces.