Power analysis concluded at least 11 patients (n = 11) were required.¹⁰ The inclusion criteria were: mandibular prognathism and maxillary retrognathia due to skeletal Class III malocclusion; bimaxillary operation with manual TMJ positioning; no asymmetry (menton [Me] deviation < 4 mm); no TMJ disorder; no systemic disease, trauma, ankylosis, or syndromes; and no history of cyst, tumor, or prior surgical operation in jaws except BSSRO. The exclusion criteria were low image quality, images not demonstrating the area of interest, and absence of preoperative (maximum two weeks prior) or postoperative (maximum 7 months after) images. All 44 patients who met the inclusion criteria were included to reduce the margin of statistical error. Initial evaluation of the subjects’ skeletal characteristics revealed an average anterior-posterior and vertical discrepancy of 8.91 ± 2.48 mm (negative overjet) and 1.87 ± 0.71 mm (overbite), respectively. All patients underwent preoperative and postoperative orthodontic treatments.

During bimaxillary orthognathic surgery, the ramus was manually adjusted so that the condylar position provided an ideal centric relationship. All surgical procedures were performed between September 2016 and September 2020 by the same right-handed oral and maxillofacial surgeon (EC) with 14 years of experience.

All images including TMJ regions were gathered with a CBCT device (SCANORA^® 3Dx Dental CT, SOREDEX Nahkelantie, Tuusula, Finland; field of view: 140×165 mm, tube voltage: 90 kVp, tube current: 13 mA, and scan time: 26 seconds) between September 2016 and February 2021. Before the scans, the natural head position was determined, and the teeth were positioned at maximum intercuspation. The thicknesses of the cross-sectional images were 0.3 mm.

Cone-beam computed tomography images obtained 1–2 weeks (mean: 1.6 ± 0.5 weeks) before (T0) and approximately 6 months (ranging between 5.4–6.9 months, mean: 6.1 ± 0.2 month) after (T1) the surgery, were converted in digital imaging and communications in medicine (DICOM) format. Both condylar positions were analyzed using open-source and non-commercial 3D Slicer software (3D Slicer, Brigham and Women’s Hospital, Boston, MA, USA).

Because the natural head position might vary due to changes in the stabilization of the craniocervical complex before measurements, the head orientation was arranged in an anatomic frame of reference. In the axial plane, the Frankfort horizontal plane (FHP; passing through the right portion [Po], left Po, and midpoint between the right and left orbita [Or]) was parallel to the floor. The basion-nasion (Ba-Na) plane was perpendicular to the FHP in the sagittal plane. The coronal plane passing through sella (S) was set perpendicular to the other planes. Me deviation, the distance from Me to the Ba-Na plane, was measured for asymmetry. The maxillary and mandibular positional change was evaluated as the vertical distance from points A and B to the line passing perpendicular to the FHP through the Ba.

The reference landmarks as points, lines, and angles used to determine the condylar position in three planes are shown in Figures 1–3. The distance, ratio, and angle measurements were performed on the CBCT images to objectively compare the pre- and postoperative condylar positions.

The FHP was used as a reference plane in the axial plane, and the slice with the most significant mediolateral dimension of the condyle was identified. In terms of α-angle, also known as a horizontal condylar angle, increased angulation indicated inward rotation, while decreased α-angle indicated outward rotation of the condyle postoperatively (Figure 1).

In the sagittal plane, slices parallel to the Ba-Na plane and passing through the mediolateral midpoint of the condyle established on the axial plane were identified on both sides. Vertical and horizontal positional change of the condyle in the glenoid fossa were examined using the distances between the lowest points of the articular eminence (D1) and the temporal squamotympanic fissure (D2, M); D1 and the point of the line drawn perpendicular to the line passing through D1 and D2 (Line3) from the highest point of the condylar head (C2, m); the highest point of the glenoid fossa (C1) and the point of the line drawn perpendicular to Line3 from point C1 (D3; N); and C2 and the point of the line drawn perpendicular to Line3 from C2 (n). An increased n/N ratio indicated vertical upward movement, whereas a decreased n/N ratio indicated vertical downward movement of the condyle. An increased m/M ratio indicated a horizontal backward movement, whereas a reduced m/M ratio indicated a horizontal forward movement of the condyle (Figure 2).

In the coronal plane, the plane passing perpendicular to the FHP was used as the reference plane, and the slice parallel to the reference plane with the most significant mediolateral dimension of the condyle was identified. The β-angle formed between Line4 (the line passing between the highest point of the right [R] and left [L] glenoid fossae) and Line5R and 5L (the lines passing between medial and lateral poles of right and left condyles, respectively) and the C distance between the medial poles of both condyles were examined. An increased β-angle in the postoperative period indicated inward rotation, while a decreased β-angle indicated outward rotation of the condyle. An increased or decreased C-distance indicated an increased or decreased postoperative distance between the condyles (Figure 3).

The images were evaluated between December 2021 and June 2023 by two observers: an oral surgery resident (OK) and a dentomaxillofacial radiologist (NE) with 12 years of experience. After at least 1 month, all images were re-evaluated to determine intra- and inter-observer agreements.

Statistical Package for the Social Sciences (SPSS version 25; IBM Corp., Armonk, NY, USA) was used for statistical analysis. The mean and standard deviation values were calculated, and preoperative and postoperative values were compared for the related pairs. After testing the homogeneity of variance and normality using the Shapiro–Wilk test, the dependent sample t test was applied. Intra- and interobserver agreements on all distance, ratio, and angle measurements were analyzed for the whole sample using the Intraclass Correlation Coefficient (ICC) test that values < 0.5, 0.5–0.75, 0.75–0.9, and > 0.9 were indicative of poor, moderate, good, and excellent reliability, respectively.¹¹ A significance level of 95% was accepted.

Ueki et al.²³ showed a mean α-angle (horizontal condylar angle) of 12.0° ± 7.9° and 11.8° ± 5.1° on the right and left, respectively, in Class III patients with facial symmetry. These findings were much lower than those reported in other studies, including our study. This might be attributed to the different populations and differences in the condylar shapes in the study populations.

Lee and Park,²⁰ who also placed the posterior segment empirically during BSSRO and unlike our study, evaluated the TMJ region using CT instead of CBCT, 1–4 months postoperatively instead of 6 months, and measured condylar angulation as the acute angle of a right-angled triangle other than the α-angle reported an inward rotation of the condyle with a mean value of 4°. However, they did not report the side that demonstrated rotation. Choi et al.¹⁰ reported an increase in the α-angle postoperatively, which returned to the preoperative position after 6 months. Kim et al.²⁴ reported an increase in the mandibular condyle angle of 2.23° and 2.18° on the right and left sides, respectively, which resulted in an inward rotation. Nishimura et al.²⁵ stated that in patients with mandibular prognathism, the lateral part of the condyle rotates anteromedially due to internal fixation of the proximal segment, and there is less change with nonrigid fixation than with rigid fixation. Kim et al.²⁶ reported that the condyles inclined outward insignificantly. Podčernina et al.⁸ evaluated TMJ position, and in axial condylar angle, they found similar values (T0: 19.6 ± 7.5 and T2-after 1 year: 22.2 ± 7.6) as in our study. However, they did not indicate whether these measurements were from the right or left side or the overall mean. Jung et al.,²⁷ who used condylar positioning plate and performed evaluations on CT images reported significant differences in the left condylar axis angle (T0: 19.89° ± 5.40°, after 3-months: 21.21° ± 5.52°, after 1-year: 20.72° ± 5.44°), whereas no significance was demonstrated on the right side (T0: 20.93° ± 5.55°, after 3-months: 21.79° ± 6.04°, after 1-year: 21.24° ± 5.58°).

In our study, while α-angleL increased and demonstrated an inward rotation 6 months postoperatively, the increase in α-angleR was insignificant. This difference between the right and left condyles may be because the operator was always in the same operating position; thus, the direction and amount of force applied during the fixation phase differed. Inward rotation of the condylar head could also be due to screw insertion during rigid fixation procedures.

In addition, Omar and Bamber²⁸ reported an unplanned rotation of the TMJ toward the right side in manual model surgery compared with a passive Robot Arm. Mediolateral errors toward the right side in manual model surgery may be due to the operator’s right-handedness. In our study, all surgical procedures were performed by a right-handed surgeon. The standing position and right-handedness of the operator could be the reasons for the inward rotation.

Choi et al.,¹⁰ who included patients with and without asymmetry, reported non-significant m/M ratio changes, representing horizontal movement pre- and postoperatively. Regarding the n/N ratio, which shows vertical movement, a significant decrease was reported immediately after surgery, indicating that the mandibular condyle had a perpendicular orientation to the inferior. However, at 6 months postoperatively, the n/N value also returned to the initial value, resulting in a non-significant positional change. Regarding the ‘m’ distance calculation, Choi et al.¹⁰ defined the distance between D1 and D3 points as the ‘m’ distance. Because these points are fixed anatomical points in which positional changes were not anticipated postoperatively, this might lead to an erroneous calculation that there will be no alteration in the sagittal ‘m/M’ ratio. Consequently, the horizontal movement change cannot be determined. In our research, ‘m’ distance was defined as the distance between D1 and D4, where D4 may change position due to surgical operation.

Ueki et al.,²⁹ who used bent plates to maintain the condyle in its original position, reported no significant change in the anteroposterior direction after BSSRO on cephalograms. Lee and Park²⁰ reported that the condyle moved down and forward, albeit in small amounts, which was non-significant at 6 months postoperatively. Kim et al.²⁶ concluded that the superior movement of both mandibular condyles and anterior movement of the right condyle were substantial. Our study showed no significant change in either the anteroposterior or superoinferior direction. The manual positioning maneuver during BSSRO (directing the proximal part backward from the lower level of the vertical corpus osteotomy line and extraorally upward from the mandibular angulus) might be sufficient to provide the initial anteroposterior and superoinferior relationship of the condyle. No additional devices or processes during the operation are necessary for cost, time, or operation success.

Choi et al.¹⁰ reported that in patients with facial symmetry, the β-angle significantly decreased 1 month after BSSRO, which returned to the initial mean value 6 months postoperatively. Choi and Lee³⁰ reported a decrease of 0.65° on the right and 0.5° on the left sides. Kim et al.²⁶ reported that the condylar angle decreased by a mean of 0.92° on both sides. Compatible with the literature, our findings showed that the right β-angle decreased significantly. Besides, no significant change was observed on the left side at T1 time, which means that the right condyle rotated outward. Both the effects of the fixed standing position of the operator, which might also affect the axial plane, and the direction and amount of force during the fixation process could have contributed to this outward rotation.

Regarding intercondylar distance, Kawamata et al.¹⁷ reported a 2 mm increase following SSRO on CT images. Choi et al.¹⁰ reported that the distance between the condyles in the facial symmetry group (Me deviation < 4 mm) showed a significant increase at 1 month postoperatively and tended to return to the initial distance at 6 months postoperatively. Choi and Lee³⁰ reported no significant differences, although the distance between the condyles tended to increase. In our study, no significant changes were observed, which is consistent with the results of previous studies.

One limitation was that there were two-time points; therefore, information regarding the later stages must be provided, considering radiation protection principles. Since TMJ remodeling is expected in these patients, changes in the condylar position could be better evaluated using landmarks designated by the condylar shape. Parameters such as condylar remodeling, asymmetry, TMJ disorders, and maxillary or mandibular movement, which might affect the condylar position, were not calculated. Although the data of the patients operated on by a single surgeon eliminates variables dependent on the operator, involving data obtained from patients operated on by other surgeons and verifying reproducibility between these surgeons is necessary.