This study aimed to evaluate the morphologic and positional features of the mandible in children, adolescents, and adults with skeletal Class I and unilateral posterior crossbite.
The sample included cone-beam computerized tomography images of 76 subjects, divided in 3 groups: (1) children (aged 6.77 ± 1.5 years; n = 25), (2) adolescents (aged 14.3 ± 1.7 years; n = 26), and (3) adults (aged 32.66 ± 13.4 years; n = 25) with unilateral posterior crossbite. Condylar and mandibular linear distances and angles were performed using a mirrored 3-dimensional overlapped model. Intragroup asymmetries were determined by a comparison between crossbite and no crossbite sides. The differences between both sides of all measurements were compared among groups and correlated to mandibular horizontal rotation (yaw) and age.
The crossbite side showed shorter distances in the condyle and mandibular regions. Asymmetries were slightly but significantly greater in adults, as expressed by the lateromedial condylar distance, total ramus height, and mandibular length with an average 0.7 mm, 2.0 mm, and 1.5 mm, respectively. The mandibular yaw rotation was not correlated to age but moderately associated (r = 0.467) to asymmetry in mandibular length and total ramus height.
Patients with skeletal Class I and unilateral crossbite showed small mandibular asymmetries and these conditions were slightly greater in adults, specifically in lateromedial condylar distances and mandibular body and length.
Unilateral posterior crossbite in skeletal Class I patients presents small asymmetries in all ages.
The crossbite side presented slightly shortened distances.
Adults showed more asymmetric condyles and hemimandibular length.
Mandibular yaw rotation is moderately correlated to asymmetries.
Mandibular yaw rotation is not correlated to age.
Unilateral posterior crossbite is reported to occur in up to 22% of the population. The etiology of unilateral crossbite is related to a combination of dental, skeletal, and neuromuscular disturbances in association to narrower maxilla and/or wider mandible. The presence of oral habits such as thumb sucking, abnormal swallowing, and respiratory disorders during childhood are considered the main etiological factors. In terms of occlusal dynamic, the mandibular shift is 1 of the key mechanisms that lead to a unilateral posterior crossbite in the early stages of life. It is considered a response to uncomfortable occlusal contacts, resulting in a buccal position of the lower posterior teeth. This event is confirmed by lower midline deviation to the crossbite side and no association between centric relation and the maximum intercuspal position of the mandible.
There is a consensus that unilateral crossbite does not self-correct during growth and aging, and early treatment in childhood is recommended in orthodontic practice. This functional shift of the mandible, commonly present in unilateral posterior crossbite, is considered an etiological factor for irregular growth and the development of temporomandibular joint asymmetry. In addition, some studies have suggested that the natural history of a functional unilateral posterior crossbite in childhood would lead to the development of permanent skeletal asymmetry in adults. , , This statement justifies the earlier intervention in patients with unilateral posterior crossbite to correct the direction of the craniofacial growth. However, some studies have reported that unilateral crossbite does not show significant differences in condylar position within the glenoid fossa between the crossbite side and control groups. , In addition, no lateral functional mandibular shift in adults was reported, suggesting that condyles and glenoid fossa in older patients may have an adaptive repositioning. ,
The anteroposterior relationship of the maxilla and the mandible can influence their transverse relationship and asymmetry. There is evidence of a higher occurrence of mandibular asymmetry in skeletal Class III ; for this reason, it is important to assess mandibular asymmetry in a sample with the homogenous maxillomandibular anteroposterior relationship. The controversial relation between posterior crossbite with skeletal asymmetry during different stages of life can be explained by the absence of longitudinal studies design, samples with heterogenous malocclusions, small sample size, and image analysis methods. , , With the advent of computer technology to aid in imaging diagnosis, anatomical superimposition overlay using software has brought about new ways to compare condylar morphology and positional changes of the mandible. , Such superimpositions can provide researchers and clinicians with new information. In addition, mirroring 3-dimensional (3D) models registered at mandible and/or the skull base allow the overlay of both sides and direct visualization, respectively, of the morphologic and positional differences in the symmetry of both sides.
Systematic reviews of studies on posterior crossbite and skeletal asymmetries have reported the need for further evidence-level studies. , Comparison groups with 3 different stages of life (children, adolescents, and adults) in patients with a balanced anteroposterior relationship, skeletal Class I pattern, without the influence of abnormal mandibular growth will elucidate whether posterior crossbite is a condition that leads to progressive asymmetry or to skeletal compensations that mask asymmetry of the craniofacial complex throughout life. This study aims to evaluate the asymmetry of the mandible in children, adolescents, and adults with unilateral posterior crossbite and skeletal Class I, using mirrored 3D overlaid models.
Material and methods
The study protocol was approved by the institutional review board of the Federal University of Goiás (approval no. 60702316.3.0000.5083).
This study sample consisted of secondary data analysis of 76 cone-beam computed tomography (CBCT) scans of patients (28 male and 48 female) selected from deidentified scans taken for other clinical purposes. The sample size calculation (23 patients in each group) was based on measurements of a previous study. It was used the statistical power of 80% with an alpha of 0.05, large effect size (0.80), and a difference of 10% between the age groups for the following measurements: condylar lateromedial, mandibular body length and rami height asymmetries.
The following inclusion criteria had to be fulfilled: (1) skeletal Class I according to ANB angle (from 0° to 4.5°) , ; (2) unilateral posterior crossbite involving at least 2 posterior teeth for primary dentition, and more than 2 posterior teeth for mixed and permanent dentition, and (3) CBCT scans presenting no distortion or movement artifacts, an appropriate field of view and with teeth in centric occlusion. The exclusion criteria were: (1) condylar imaging features of degenerative disorders, erosion, subchondral cyst, generalized sclerosis or osteophytes, as defined by Ahmad et al and Schiffman et al, as well condylar abnormal size suggestive of condylar hyperplasia; (2) signs of facial or dental trauma; (3) syndromes or congenital craniofacial anomalies, such as cleft lip palate; (4) previous orthodontic or facial surgical procedures, (5) early loss of primary teeth or absence of permanent teeth, except the third molar, and (6) anterior crossbite.
The sample was divided according to the predictor variable, age, and was distributed as follow: children (aged ≥4 and <11 years; n = 25); adolescents (aged ≥12 and <18 years; n = 26), and adults (aged ≥18 years; n = 25).
Cone-beam computed tomographic scans had been taken for all subjects, using an iCat unit (Imaging Sciences International, Hatfield, PA) at these settings: 3.8 mA, 120 kV, exposure time of 8.9 seconds for children and 40 seconds for adolescent and adult patients, voxel size of 0.4 mm 3 , and field view of 17 cm in height × 23 cm in depth. All the images were exported as DICOM files.
An orthodontist examiner (K.E), previously trained in this method, performed all 3D analysis. Three-dimensional surface models were created using the following steps:
Conversion of DICOM files into GIPL files to decrease the file size using ITK-SNAP, an open-source software (version 2.4.0; www.itksnap.org ).
Downsize to a 0.5-mm 3 voxel the original scan 0.4 mm 3 voxel size using 3D Slicer (version 4.0; www.slicer.org ) to standardize the scan resolution and decrease the computational power and time for cranial base registration.
Create a volumetric label map of all cranium complex using ITK-SNAP.
Create a virtual 3D surface model using 3D Slicer.
Head positioning of all samples using the stable 3D coordinate system of the 3D Slicer. The midsagittal plane (MSP) defined by glabella, crista galli, and basion must be matched and perpendicular with the horizontal reference plane, defined by bilateral orbitale (most inferior point of the left and right orbitals) and bilateral porion (most superior point of the left and right external acoustic meatus). ,
Prelabeling: create a mandible volumetric segmented label map with landmarks using ITK-SNAP. All the landmarks are shown and described in Figure 1 .
Mirroring the prelabeled mandible volumetric label and corresponding scans using 3D Slicer. ,
Three-dimensional cranial base superimposition was performed approximating the mirrored scan manually to the original scan previously oriented, using the center of anterior cranial fossa as a best-fit reference. This step was followed by the registration of the mirrored segmented and scans files and construction of the mirrored and original models with pre-labeled landmarks using 3D Slicer.
Landmark identification: the prelabeled landmarks ( Fig 1 ) were detected at the original oriented and mirrored surface models using the Q3DC tool in the 3D Slicer.
Assessment of quantitative linear distances and angles ( Table I ), and the number of directional changes in mediolateral, anteroposterior, and superoinferior axes in Q3DC tool. The variables were measured in both sides. The yaw mandibular rotation was assessed in original and mirrored models ( Fig 2 ).Table I
Variable Identification Unit Definition Condylar linear measurements Condylar process height Co-Sig` 2D linear (mm) Superoinferior distance between condilium and the correspondent point of sigmoid notch in the posterior region of condylar neck Condylar height (Co-CtCo) Co-CtCo 2D linear (mm) Superior-inferior distance between condilium and the center of the condyle Lat-med width (lat-med poles) CoL-CoM 3D linear (mm) Lateromedial distance between the lateral and medial poles of condyle Ant-post width (ant-post poles) CoA-CoP 3D linear (mm) Anteroposterior distance between the anterior e posterior poles of condyle Mandibular linear measurements Chin deviation Me-MSP 2D linear (mm) Lateromedial distance between the Me point to MSP line Ramus height Sig`-Go 2D linear (mm) Superoinferior distance between the correspondent point of sigmoid notch in the condylar neck and gonion Ramus anteroposterior width (R ant-R post) 3D linear (mm) Anteroposterior distance between the Ramus anterior and Ramus posterior points Total ramus height Co-Go 2D linear (mm) Superoinferior distance between the condilium and gonion Mandibular length Co-Gn 3D linear (mm) Distance between condilium and gnathion Mandibular body Go-Gn 3D linear (mm) Distance between gonion and gnathion Ramus inclination Anteroposterior CtCoGo.GoGn Degree ( o ) Anteroposterior inclination of mandibular ramus Lateral CtCoGo.MSP Degree ( o ) Lateromedial inclination of the mandibular ramus to MSP line Lateral hemimandibular angle Lateral CtCoMe.MSP Degree ( o ) Lateromedial inclination of the mandibular body to MSP line Mandibular position Yaw (mandibular horizontal rotation) CoMe(or).CoMe(mir) Degree ( o ) Angle between the CoMe line at the right original side and CoMe at the “right” mirrored side
The reproducibility of this method was verified after repeating all pre-labeled landmarks and measurements of 20 patients with 15-day intervals, randomly selected from all groups, and using intraclass correlation coefficients test with a confidence level of 95%. The statistical analysis was performed with the SPSS (version 23.0; IBM, Armonk, NY). All variables distributions were tested using Kolmogorov-Smirnov and showed normal distribution for each side measurement, ANB angle and chin deviation, and non-normal data for the difference between both sides. The intergroup comparisons of sagittal relationship (ANB) and chin deviation were performed using analysis of variance test followed by Bonferroni post hoc test. Intergroup comparisons (using the differences between crossed and non-crossed sides) were performed with the Kruskal-Wallis test. For each statistically significant result of comparison tests, the effect size was calculated to determine if the difference has a strong clinical effect. The effect size was considered weak when values were <0.20, moderate when between 0.50 and 0.80, and strong when >0.80. Spearman correlation was used to test the following possible associations: (1) differences between both sides to a mandibular horizontal rotation (yaw), and (2) differences between both sides to age. The level of significance was set at 0.05.
All 3D measurements showed a high intraexaminer agreement. The lowest intraclass correlation coefficients value was for lateral mandibular body angle (CtCoMe.MSP; 0.920), and the highest was for ramus height (Sig`-Go; 0.999).
Sample distribution (age, ANB, chin deviation, sex, and crossbite side) is described in Table II . Children showed greater ANB values than adults and adolescents. Chin deviation characterized a sample with a relative and moderate mandibular asymmetry.
|Characteristics||Unilateral crossbite groups|
|4-10 y||12-17.9 y||≥ 18 y|
|Mean (SD)||6.77 (1.5)||14.13 (1.7)||32.66 (13.4)|
|Mean (95% CI)||3.7 (2.9-4.0) a||2.5 (1.9-3.1) a||2.5 (1.8-3.3) a|
|Median (25%, 75%)||1.65 (0.8, 2.6) a||1.0 (0.4, 2.1) a||1.96 (0.5, 3.3) a|