INTRODUCTION

Advancements in imaging technology have generalized the use of cone-beam computed tomography (CBCT) in the field of dentistry. CBCT convertes two-dimensional image data to three-dimensional (3D) images, effectively overcoming the limitations of the former modality, particularly in orthodontic treatment and orthognathic surgery, greatly facilitating patient diagnosis, treatment planning, and treatment-outcome assessment.1 Kau et al.,2 in comparing the many types of 3D imaging devices, asserted that 3D imaging is most effectively used for examination and treatment relating to the oral and maxillofacial region. Bayram et al.3 reported that 3D modalities are necessary for quantification of mandibular condyle volumes. Recently, a number of studies employing pre- and postoperative 3D images have been reported.

For patients with skeletal Class III deformities, corrective orthognathic surgery improves both oral functionality and related esthetics. However, one orthognathic-surgical procedure, sagittal split ramus osteotomy (SSRO), is known to induce postoperative anteroposterior movement and condylar rotation.1 During surgery, the condylar axis rotates to maintain the intersegmental contact after the distal segment moves anteriorly or posteriorly. Kim et al.4 reported that in patients who had undergone SSRO with rigid fixation, the condyle showed an inward rotation on the axial plane and was positioned posteriorly, on the basis of short- and long-term observation of condylar location. These movements reportedly do not have a negative impact on postsurgical stability.

Orthognathic surgery–induced changes of condylar location impart physical stress on the condylar surface through condylar remodeling of the temporomandibular joint (TMJ) structure, which is an adaptation to the new functional requirements.5 This functional stress can result in TMJ changes postoperatively. In some cases, however, severe condylar change is incurred, deteriorating skeletal stability, though the cause remains mostly unknown.6,7 Many other studies have reported on postoperative condylar changes as well.8–10

For assessment of condylar remodeling, several CBCT-based analysis methods have already been introduced.11,12 Carvalho et al.13 evaluated 3D changes of location in the ramus, condylar, and mental region in patients who had received mandibular advancement. Motta et al.14 superimposed a 3D CBCT model, executed mandibular advancement, and evaluated the patients' condyles. They stressed an individual variability in condylar displacement after surgery.13,14 A study by da Motta et al.15 used a 3D CBCT overlapping technique to evaluate mandibular anatomy and condylar location, proving its effectiveness. However, there are still only a few postoperative condylar remodeling analyses of skeletal Class III patients using 3D-reconstructed images and condylar superimpositions.3,13,14,16

The purpose of this study was to evaluate the effect of postoperative condylar axis changes on mandibular condylar remodeling by comparing the condylar head in 3D surface reconstructions before and after surgery (condylar pre- and postoperative 3D surface reconstructions). In addition, the purpose was to determine the relationship between condylar inward rotation and condylar surface remodeling after orthognathic surgery.

MATERIALS AND METHODS

This was a retrospective analysis of all patients who underwent mandibular setback SSRO with or without Le Fort I osteotomy for Class III malocclusion at the Department of Orthodontics, Pusan National University Dental Hospital, within the period beginning January 2010 and ending December 2012. Rigid internal fixation was achieved in the maxilla and mandible by means of plates and screws. The sample consisted of 30 adults: group 1 underwent mandibular setback SSRO with rigid fixation (10 males, five females, age 22.4 ± 3.3 years), and group 2 underwent mandibular setback SSRO and Le Fort I osteotomy (10 males, five females, age 22.3 ± 2.2 years). Patients who had been diagnosed with any syndromes, facial trauma, or degenerative joint disease were excluded. The study was reviewed and approved by the Ethics Committee at Pusan National University Hospital (E-2012055).

Images were obtained using a CBCT scanner (DCT Pro; Vatech, Seoul, Korea) within an average of 1.0 month (range, 0.1–1.5 months) prior to surgery and 12.4 months (range, 10–19 months) after surgery. The CBCT was acquired with the subject patients in an upright position for maximum intercuspation and with the FH plane adjusted parallel to the floor. The maxillofacial regions were CBCT scanned according to the following parameters: 20 × 19-cm field of view, 90-kVp tube voltage, 4.0-mA tube current, and 24-second scan time. On the basis of the CBCT data, 60 condyles total were evaluated pre- and postoperatively.

3D images of the proximal segments of the mandible were reconstructed and reformatted into the stereolithography (STL) data format using 3D imaging software (Vworks 4.0; Cybermed Co, Seoul, South Korea). The segmented proximal segment images were then imported into 3D scan data processing software (Rapidform XOS3; Inus Technology Inc, Seoul, South Korea).

To investigate the surface remodeling of the condylar head, pre- and postoperative segmented images were superimposed over three registration areas (condylar neck, mandibular notch, and posterior border of ramus).16 This function of Rapidform XOS3, designated as 3D surface-to-surface matching (best fit method), employs a least-mean-squared algorithm.17 According to the coordination between the registration areas, the reconstructed images were automatically fitted. Since the procedure is fully automated, the influence of the observers' variability on the accuracy of measurements is completely eliminated. The superimposed condylar heads were divided by a plane connecting the median and lateral poles. The surface displacements of the heads were calculated using Rapidform XOS3 (Figure 1). In order to calculate the surface difference on the 3D-superimposed images, average absolute deviation, the averages of the absolute value of the deviation between the pre- and postoperative 3D images, was first obtained; second, average deviation representing the average values of deviation between the pre- and postoperative 3D images were taken.

View Figure

• Download as Powerpoint
• Download Figure

To evaluate the surface remodeling on the superimposed condylar heads, six areas (the anterolateral, anteromiddle, anteromedial, posterolateral, posteromiddle, and posteromedial regions) were investigated and classified into three categories (bone resorption, unchanged, bone formation; Figure 2).

View Figure

• Download as Powerpoint
• Download Figure

To evaluate the effect of postoperative condylar axis change on the condylar surface changes, the changes of the condylar axis angles (the axis passing through the medial and lateral pole of the condylar head) were measured from the midsagittal reference plane (MSR; Figure 3).

View Figure

• Download as Powerpoint
• Download Figure

RESULTS

The intraoperator reliability for condylar axis changes was high according to the ICC (average, .993). The Cohen kappa index for remodeling signs also showed substantial agreement (average, .805). The mean of the average absolute deviation of condylar surface change before and after surgery was 0.22 ± 0.03 mm, and the mean of the average deviation was 0.01 ± 0.09 mm. The pre- and postoperative images revealed a significant condylar surface change difference (P < .05), though not for the average deviation (P > .05).

In a comparison of the groups according to surgery (one-jaw vs two-jaw), there was no significance in either the average absolute deviation or the average deviation (P > .05; Table 1).

Table 1. Condylar Surface Change (Preoperative to Postoperative)^a

View Fullscreen

In order to uncover the relationship between the condylar axis changes and the condylar surface changes, a Pearson correlation analysis was performed. It was found that after the surgery, the condylar axis changes showed significant changes with inward rotation (P < .05; Table 2). There were significant coefficients for the average absolute deviation (r  =  .70; P < .05); however, there were none for the average deviation (Table 3).

Table 2. Difference of Condylar Axis Angle

View Fullscreen

Table 3. Correlation Between Extents of Condylar Axis and Condylar Head Changes

View Fullscreen

In the assessment of the distribution of the surface remodeling signs from the superimposed condylar heads, all of the areas (ie, the anterolateral, anteromiddle, anteromedial, posterolateral, posteromedial, and posteromedial regions) were changed in diverse ways. Among the condylar remodeling signs, bone resorption showed a higher frequency than did bone formation (Table 4).

Table 4. Frequency of Condylar Head Remodeling Signs

View Fullscreen

DISCUSSION

The relationship between orthodontic treatment and the TMJ has long proved elusive,18 despite the determined researchers who have set out to find it.19,20 Post–orthognathic surgery stability and relapse can be influenced by the TMJ,21–23 in the sense that surgical correction of skeletal deformities requires soft-tissue responses and condylar remodeling to adapt to the new environment.5 In some studies, the location of the condylar head during orthognathic surgery has caused condylar head remodeling, which is related to recurrence.24,25 Nonetheless, the exact cause of condylar head change remains unknown.

Some studies26,27 using lateral cephalograms have reported anterosuperior condylar resorption and remodeling after orthognathic surgery for cases of anterior openbite, and a 3D study16 employing CT data reported the effect of orthognathic surgery on a condyle. However, for comprehensive evaluation of condylar surface changes, 3D reconstructed images and a method of superimposition over the ramus region are necessary. In the present study, the 3D reconstructed images were obtained from pre- and postoperative CBCT data.3 Preparatory to assessment of condylar head surface changes resulting from orthognathic surgery, 3D images were superimposed over the base of the registration areas (condylar neck, mandibular notch, and posterior border of ramus).16

In the results, the average deviation did not manifest significant change, whereas the average absolute deviation did (P < .05; Table 1). These changes were found to be unrelated to the surgical types, either one- or two-jaw surgery (Table 1), similar to the result of Park et al.16

The condylar axis showed significant inward rotation (P < .05; Table 2). There were significant coefficients for the average absolute deviation (r  =  .70; P < .05), though none for the average deviation (Table 3). As the condylar axis change increased, the average absolute deviation showed significant increases (P < .05). The inward rotation of the condylar axis produced condylar surface remodeling. However, it also did not show which remodeling signs were related to that rotation.

To investigate which remodeling sign is prominent on the condylar surface after orthognathic surgery in skeletal Class III deformities, surface remodeling was assessed with reference to superimposed 3D images of the condylar heads. It was found that all of the areas (ie, the anterolateral, anteromiddle, anteromedial, posterolateral, posteromiddle, and posteromedial regions) were changed in diverse ways. Among the condylar remodeling signs, bone resorption showed a higher frequency than did bone formation (Table 4).

Overall, inward rotation after mandibular setback SSRO or LeFort I osteotomy with rigid fixation in skeletal Class III deformities induced condylar surface remodeling. This condylar rotation can lead to changes of the condylar surface (Table 4). Park et al.16 showed that bone resorption and bone formation occurred more frequently in specific areas after mandibular setback SSRO and two-jaw surgery. In the present study, specific areas did not show distinct signs of bone remodeling. A previous study16 used multiplanar reformation images to evaluate condylar remodeling signs, reporting specific change sites (ie, resorption areas, anterior and superior areas on sagittal plane, superior and lateral areas on coronal plane, anterolateral and posterolateral areas on axial plane, bone formation area, and anteromedial area on axial plane). Nonetheless, Hwang et al.28 suggested that the posteriorly inclined condylar neck should be considered as a relevant nonsurgical risk factor in condylar resorption following orthognathic surgery. Our study used 3D surface images, which could include all two-dimensional MPR images, because 3D surface images are more accurately representative of condylar head remodeling. From the comparison of 3D model accuracy between CBCT and MSCT, the artifacts were mostly located at the mandibular border and the posterior margin of the scan volume, and they appeared at similar positions on all CBCT systems. All comparisons were statistically significant, yet very small differences were obtained, which do not necessarily have a clinical significance.29 Therefore, further studies should be needed for the various methodologies.

The obtained data established that condylar remodeling had occurred after orthognathic surgery, thus effecting positional changes of the proximal segments. There were no specific remodeling signs in the condylar head. Even though the inward rotation of the condylar axis was related to the condylar head remodeling, it was a part of the process of adaptation to the changed TMJ environment. However, some anterior openbite cases appear to be sensitive to functional stress on the condyle, and so their adaptive capacity to the changed TMJ environment probably is smaller than that of deep-bite cases.30,31 Therefore, patients with predisposing factors of condylar adaptive capacities should be informed about the condylar remodeling, specifically condylar resorption and its adverse effects on skeletal stability and relapse.

CONCLUSIONS

• In orthognathic surgery for skeletal Class III deformity correction, 3D-superimposed images of the pre- and postoperative condyles were useful for observing and assessing condylar surface changes such as remodeling signs.

• The postoperative inward rotation of the condyles was correlated with the average absolute deviation of the condyles, regardless of surgery type (one- or two-jaw surgery; r  =  .70; P < .05).

• However, there was condylar surface remodeling over the entire condylar head surface. No specific remodeling signs appeared in the condyles.