The purpose of this study was to apply a novel method to evaluate surgical outcomes at 1 year after orthognathic surgery for Class III patients undergoing two different surgical protocols. Fifty patients divided equally into two groups (maxillary advancement only and combined with mandibular setback) had cone beam computed tomography (CBCT) scans taken pre-surgery, at splint removal, and at 1-year post-surgery. An automatic cranial base superimposition method was used to register, and shape correspondence was applied to assess, the overall changes between pre-surgery and splint removal (surgical changes) and between splint removal and 1-year post-surgery at the end of orthodontic treatment (post-surgical adaptations). Post-surgical maxillary adaptations were exactly the same for both groups, with 52% of the patients having changes >2 mm. Approximately half of the post-surgical changes in the maxilla for both groups were vertical. The two-jaw group showed significantly greater surgical and post-surgical changes in the ramus, chin, and most of the condylar surfaces ( P < 0.05). Post-surgical adaptation on the anterior part of the chin was also more significant in the two-jaw group ( P < 0.05). Regardless of the type of surgery, marked post-surgical adaptations were observed in the regions evaluated, which explain the adequate maxillary–mandibular relationship at 1-year post-surgery on average, with individual variability.
Class III skeletal malocclusion can be corrected with orthodontic–surgical approaches, which, in addition to improving aesthetics and function, are expected to be stable over time. For correction of skeletal Class III, hard tissue long-term follow-up with two-dimensional (2D) cephalometric analyses have shown that maxillary advancement only or advancement combined with mandibular setback are stable procedures. However, despite improvements in surgical techniques for mandibular setback, post-operative stability still leaves something to be desired.
Short and long-term three-dimensional (3D) superimposition studies using cone beam computed tomography (CBCT) have evaluated mandibular changes for Class II patients. For Class III patients, Cevidanes et al. evaluated changes in the condyles and rami with the closest point distances method at 1 week after surgery. Other studies on surgical outcomes of jaw surgery have evaluated short-term soft and hard tissue changes with landmark coordinates and linear or angular measurements in anatomical crossections. In 2011, Paniagua et al. introduced shape correspondence as a means of quantifying surgical displacements to the field of orthognathic surgery. Shape correspondence has previously been used to analyse brain morphology and condylar resorption. This method does not rely on specific 2D or 3D landmarks or closest point surface distances, but aims to automatically obtain correspondence points in ‘before’ and ‘after’ surgery 3D models, and is able to quantify the changes that have occurred in three axes of space ( x , y , z ). The purpose of this study was to apply the shape correspondence technique to evaluate the 1-year post-operative stability of the maxillary–mandibular complex with two different surgery protocols in Class III patients.
Materials and methods
Fifty patients (23 male, 27 female, mean age 24.7 years) with Class III skeletal malocclusions were selected for this prospective study, which was approved by the Institutional Review Board of the University of North Carolina. All patients had a skeletal Class III jaw relationship with edge-to-edge or negative overjet. For each study participant, orthognathic surgical treatment was the most recommended treatment option to fully correct the malocclusion and skeletal imbalances. The patients were divided into two groups according to the type of surgery (25 patients in each group). The first group had maxillary advancement only and the second group had maxillary advancement combined with mandibular setback (bilateral sagittal split osteotomy); both groups were stabilized with rigid internal fixation. Patients with cleft lip and palate, syndromes, and disharmonies due to trauma were excluded.
CBCT images were acquired before surgery, at splint removal from 4 to 6 weeks after surgery, and at 1-year follow-up. The scanning protocol involved a 36-s full head exposure, using the NewTom 3G scanner (Aperio Services, Sarasota, FL, USA) with a 12-in. field of view. Ten patients had at least one CBCT taken with a NewTom 9000 (Aperio Services) with a 9-in. field of view. Due to this smaller field of view, either the chin or the condyles were missing in these patients. The voxel dimension was an isotropic 0.5 mm × 0.5 mm × 0.5 mm. During the scan all the subjects were biting a thin wax record to maintain centric occlusion.
After the CBCT acquisitions, semi-automatic segmentations were completed for the volumes using open-source software ITK-Snap ( www.itksnap.org ). This software allows the construction of models based on voxel grey scale intensity. To superimpose different time-points, all the post-surgical models were registered to the cranial base of the pre-surgery volume. The registration method consists of a fully automated voxel-wise rigid registration that compares and matches the intensities of the voxel grey scales at the cranial base between different time-points and relocates the image with 6 degrees of freedom.
Point-distributed correspondent models of all the regions of interest were computed using SPHARM-PDM toolbox. This method establishes correspondences based on the inherent geometry of the population.
Shape analyses and measurements of surgical outcomes were computed by subtracting pre- and post-surgery point-based correspondent models and were displayed via colour-coded distance magnitude and vector maps ( Fig. 1 ). The distance maps show the magnitude of the position changes between two point-based correspondent models, while the vector maps provide the directionality of these positional displacements. Positive and negative numbers represent outward and inward displacement, respectively. Thirteen regions of interest were selected ( Fig. 2 ), and the largest displacement for each region was calculated for pre-surgery to splint removal (S1) and splint removal to 1-year post-surgery (S2).
Analyses were done with SPSS 17.0 (Chicago, IL, USA). The greatest displacement for each anatomic region was measured in two superimpositions. To assess observer error, 10 randomly selected patients were measured twice by a single observer in a 10-day interval using intra-class correlation (ICC).
For each anatomic region, the independent Student’s t -test was used to test whether there were statistically significant differences in the outcomes of the two types of surgery, (1) between pre-surgery and splint removal (surgical displacements, S1), and (2) between splint removal and 1 year post-surgery (post-surgical adaptations, S2). In addition, percentages of patients who experienced positive or negative displacement greater than 2 mm were calculated. Since the corresponding distances between the superimpositions have positive and negative signals that represent inward and outward movement rather than a single direction, descriptive statistics and the independent Student’s t -test were done with the absolute value of the distances. Statistical significance was set at 0.05.
The agreement between repeated measures was with all the areas having ICC coefficients greater than 0.90.
Descriptive statistics of displacement in each one of the regions of interest are shown in Tables 1 and 2 for S1 and S2, respectively.
|Anatomic surface||Type of surgery||n||Mean surface distances (mm)||SD|
|Right condyle posterior surface||Maxilla-only||24||1.56||0.84|
|Left condyle posterior surface||Maxilla-only||24||1.89||0.92|
|Right condyle medial pole||Maxilla-only||24||1.54||0.80|
|Left condyle medial pole||Maxilla-only||24||1.69||0.84|
|Right condyle lateral pole||Maxilla-only||24||1.65||0.92|
|Left condyle lateral pole||Maxilla-only||24||1.50||0.67|
|Right condyle superior surface||Maxilla-only||24||1.52||0.91|
|Left condyle superior surface||Maxilla-only||24||1.61||0.69|
|Right posterior border ramus||Maxilla-only||24||2.64||1.83|
|Left posterior border ramus||Maxilla-only||24||3.27||1.65|
|Anterior surface of the chin||Maxilla-only||22||3.43||1.92|
|Inferior border of the chin||Maxilla-only||19||4.44||2.84|
|Anatomic surface||Type of surgery||n||Mean surface distance (mm)||SD|
|Right condyle posterior surface||Maxilla-only||24||1.43||0.76|
|Left condyle posterior surface||Maxilla-only||24||1.44||0.61|
|Right condyle medial pole||Maxilla-only||24||1.37||0.76|
|Left condyle medial pole||Maxilla-only||24||1.33||0.86|
|Right condyle lateral pole||Maxilla-only||24||1.43||0.78|
|Left condyle lateral pole||Maxilla-only||24||1.57||0.89|
|Right condyle superior surface||Maxilla-only||24||1.43||0.91|
|Left condyle superior surface||Maxilla-only||24||1.49||0.53|
|Right posterior border ramus||Maxilla-only||24||2.14||1.07|
|Left posterior border ramus||Maxilla-only||24||2.11||1.39|
|Anterior surface of the chin||Maxilla-only||22||2.80||1.35|
|Inferior border of the chin||Maxilla-only||19||2.77||1.51|
Maxilla-only surgery group
In the maxilla-only surgery group ( Figs 3 and 4 ), the maxillary region was displaced forward and downward or upward more than 4 mm for 96% (mean displacement 6.78 ± 2.32 mm) of the patients in S1. Post-surgical adaptations (S2) included mostly vertical changes of 2 to 4 mm or −2 to −4 mm for 39% of the cases, and >4 mm or <−4 mm for 13% of the patients, and there were greater horizontal and vertical components for negative and positive values, respectively.
Surgical movements (S1) >2 mm and <−2 mm of the left and right condyles in the maxilla-only surgery group occurred in 33% of the cases at the posterior surface, in 27% of the cases at the medial and lateral condylar poles, and in 23% of the cases at the superior surface. Post-surgical adaptations (S2) >2 mm and <−2 mm at the posterior condylar surface occurred in 19% of the cases, in 12% and 21% of the cases at the medial and lateral poles, respectively, and in 15% of the cases at the superior surface ( Fig. 5 ). The left and right posterior ramus borders presented post-surgical adaptation changes >2 mm and <−2 mm, respectively, in 38% and 50% of the cases in S2. The chin and inferior border of the mandible were the regions with greatest post-surgical adaptation in S2, showing adaptive displacements >2 mm or <−2 mm in 64% and 68% of the cases, respectively.
Two-jaw surgery group
In the two-jaw group ( Figs 6 and 7 ), the maxillary region was advanced more than 4 mm for all patients in S1 (mean displacement 7.45 ± 2.84 mm). Post-surgical adaptations >4 mm in the maxillary position were not experienced, and changes between 2 and 4 mm occurred in 52% of the cases.
Surgical movements (S1) of the condyles in the two-jaw surgery group often led to some degree of antero-posterior and lateral condylar rotation. Surgical movements of >2 mm and <−2 mm occurred in 55% of the patients at the posterior surface of the condyles, in 62% and 49% of the cases at the medial and lateral condylar poles, respectively, and in 42% of the cases at the superior surface ( Fig. 8 ). The left and right posterior ramus borders presented >2 mm and <−2 mm movements with surgery in 96% of the cases and post-surgical adaptive changes in 77% of the cases. The chin and inferior border of the mandible were the regions with greatest post-surgical adaptation in S2, showing adaptive changes >2 mm or <−2 mm in 83% (mean 4.15 ± 2.71 mm) and 75% (mean 3.63 ± 2.44 mm) of the cases, respectively.
The Student’s t -test indicated that changes at the ramus and the anterior surface of the chin were significantly different in the two types of surgical procedure, both for the surgical movements ( Table 3 ) and the post-surgical adaptive changes ( Table 4 ). The mean condylar change differences between the two groups were smaller than 0.6 mm for all condylar anatomic regions. Maxillary changes with surgery and post-surgical adaptations were not statistically different when the two groups were compared.
|Anatomic surface||Mean difference (mm)||Standard error of the difference||95% CI of the mean||P -Value|
|Right condyle posterior surface||−0.40||0.25||−0.91||0.11||0.12|
|Left condyle posterior surface||−0.44||0.27||−0.98||0.11||0.12|
|Right condyle medial pole||−0.60||0.27||−1.14||−0.06||0.03 a|
|Left condyle medial pole||−0.61||0.29||−1.20||−0.02||0.04 a|
|Right condyle lateral pole||−0.43||0.25||−0.93||0.06||0.09|
|Left condyle lateral pole||−0.62||0.26||−1.13||−0.10||0.02 a|
|Right condyle superior surface||−0.54||0.25||−1.04||−0.05||0.03 a|
|Left condyle superior surface||−0.46||0.26||−0.98||0.06||0.08|
|Right posterior border ramus||−1.72||0.55||−2.83||−0.60||0.00 a|
|Left posterior border ramus||−2.26||0.50||−3.27||−1.25||0.00 a|
|Anterior surface of the chin||−5.29||0.77||−6.85||−3.74||0.00 a|
|Inferior border of the chin||−5.78||1.28||−8.40||−3.16||0.00 a|