Patients

This study included 64 Japanese adults (21 men and 43 women) who presented with jaw deformities diagnosed as mandibular prognathism with and without a maxillary deformity. At the time of orthognathic surgery, the patients ranged in age from 16 to 51 years, with a mean age of 29.2 years (standard deviation 10.9 years). For this retrospective study, informed consent was obtained from all patients in accordance with the Declaration of Helsinki; the study was approved by the ethics committee of Yamanashi University Hospital.

Surgery

Lateral, frontal, and submentovertex (SV) cephalograms were obtained before surgery, as described previously. All 64 patients underwent bilateral SSRO setback with the modified fixation. Of the 64 patients, 26 underwent SSRO with Le Fort I osteotomy to advance or impact the maxilla. At the time of fixation, the dental arch of the distal segment was secured to the maxillary arch with an interpositional splint and 0.4-mm wire.

For seven patients (14 sides), a titanium miniplate was used for internal fixation of the mandible (long miniplate (four holes/burr 8 mm, thickness 1.0 mm) and four screws (2 × 14 mm and 2 × 5 mm); Universal Mandible fixation module, Stryker Leibinger Co., Freiburg, Germany). These patients comprised group 1 (titanium group).

For three patients (six sides), a titanium miniplate (long miniplate (four holes/burr 8 mm, thickness 1.0 mm) and four screws (2 × 14 mm and 2 × 5 mm); Universal Mandible fixation module) was used for internal fixation of the mandible, and self-setting α-TCP (Biopex) was inserted in the anterior part of the gap between the segments after plate fixation. These patients comprised group 2 (titanium with α-TCP group).

For seven patients (14 sides), a PLLA miniplate (28 × 4.5 × 1.5 mm with four screws (2 × 8 mm); FIXSORB-MX, Takiron Co., Osaka, Japan) was used for internal fixation of the mandible. These patients comprised group 3 (PLLA group).

For 11 patients (22 sides), a PLLA miniplate (28 × 4.5 × 1.5 mm with four screws (2 × 8 mm); FIXSORB-MX) was used for internal fixation of the mandible, and self-setting α-TCP (Biopex) was inserted in the anterior part of the gap between the segments after plate fixation. These patients comprised group 4 (PLLA with α-TCP group).

For 19 patients (38 sides), a uHA/PLLA miniplate (28 × 4.5 × 1.5 mm with four screws (2 × 8 mm); SuperFIXSORB-MX; Takiron Co.) was used for internal fixation of the mandible. These patients comprised group 5 (uHA/PLLA group).

For 17 patients (34 sides), a uHA/PLLA miniplate (28 × 4.5 × 1.5 mm with four screws (2 × 8 mm); SuperFIXSORB-MX) was used for internal fixation of the mandible, and self-setting α-TCP (Biopex) was inserted in the anterior part of the gap between the segments after plate fixation. These patients comprised group 6 (uHA/PLLA with α-TCP group).

The groups were divided by time-frame, so the sample number was statistically variable.

To prevent intraoperative inward rotation of the condylar long axis, model surgery was performed preoperatively with reference to the SV projection. Before surgery, an SV cephalogram was obtained for all patients, followed by simulation. A distal segment including the lower dental arch was first set back according to the position of the upper dental arch on the SV cephalometric trace.

When the proximal and distal segments are fixed with straight plates after bilateral sagittal split osteotomy (BSSO), proximal segments containing the condylar head cause internal rotation ( Fig. 1 ). To prevent internal rotation of the proximal segments, the overlapping cortical bone at the anterior site of the proximal segment was not removed to keep the contact area between the proximal and distal segments, and was fixed with a bent plate and four screws on each side of the mandible. At the posterior part, a 3–7-mm gap was maintained between the proximal and distal segments.

Simulation of plate bending. (A) Straight plate: when the proximal and distal segments are fixed with straight plates after bilateral sagittal split osteotomy, the proximal segments containing the condylar head cause internal rotation, and the posterior aspect of the distal segment on the deviated side sometimes interferes with the proximal segment. (B) Bent plate: the plates were bent to prevent the proximal segments from rotating internally; note the gap between the osteotomy surfaces.

After surgery, elastics were placed to maintain an ideal occlusion. All patients received orthodontic treatment before and after surgery. CT images were obtained for all patients preoperatively, immediately after surgery, and 1 year after surgery.

CT data acquisition

For CT scanning, the patients were placed in the gantry with the tragal–canthal line perpendicular to the floor. They were instructed to breathe normally and to avoid swallowing during the scanning process. CT scans were obtained in the radiology department by skilled radiology technicians using a high-speed, advantage-type CT generator (Light Speed Plus; GE Healthcare, Milwaukee, WI, USA) with each sequence taken 1.25 mm apart for three-dimensional (3D) reconstruction (120 kV, average 150 mA, 0.7 s/rotation, helical pitch 0.75). The resulting images were stored in the attached workstation computer (Advantage Workstation version 4.2; GE Healthcare, Milwaukee, WI, USA), and 3D reconstruction was performed using the volume-rendering method. ZedView version 7.0 (LEXI Co., Tokyo, Japan) medical imaging software was used for 3D morphological measurements.

CT image analysis software (DIANA DICOM free version 1.8.1.6; Luke-System, Chiba, Japan) was used to measure the CT value (pixel value). CT values of the lateral cortex, medial cortex, and osteotomy site were measured preoperatively, immediately after surgery, and at 1 year postoperative using horizontal CT images with the mandibular foramen parallel to the Frankfort horizontal (FH) plane. For each CT image, three randomly selected points at the anterolateral site of the lateral cortex, three at the anteromedial site of the medial cortex, and three at the anterolateral site of the mandibular canal were assessed. The mean pixel value in the region of interest (ROI; approximately 1.0 mm ² circle) was measured automatically. The mean value for the three points was then calculated and assigned as the pixel value for each part. Measurement of the screws was performed for just one screw on each side for all patients. The measurement area for the screw was very small for determining the ROI. Therefore, the distance between the long axis line of the screw and the outline of the lateral cortical surface was measured three times using another horizontal CT image without artefacts. The mean value was calculated and determined as the pixel value for the screw ( Figs. 2–4 ).

Preoperative horizontal CT image at the mandibular foramen parallel to the FH plane. The red points indicate where the pixel values were measured (three points in the lateral cortex, three at the osteotomy site, and three in the medial cortex).

Horizontal CT image at the mandibular foramen parallel to the FH plane obtained immediately after surgery. The red points indicate where the pixel values were measured (three points in the lateral cortex, three at the osteotomy site, and three in the medial cortex). The red arrows indicate where pixel values were measured at the centre of the screw head. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Horizontal CT image of the mandibular foramen parallel to the FH plane obtained at 1 year postoperative. The red points indicate where the pixel values were measured (three points in the lateral cortex, three at the osteotomy site, and three in the medial cortex). The red arrow indicates where the pixel value was measured at the centre of the screw head. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Measurement of pixel values of unused absorbable screws (in vitro)

The pixel value of the unused absorbable screws was unknown. Therefore, six unused PLLA screws and five uHA/PLLA screws were implanted in impression paste and their CT image taken under the same conditions. The pixel value was then measured three times at the centre of the screw head. The mean value was calculated and determined as the pixel value of the screw.

Statistical analysis

Data were analyzed statistically with Dr. SPSSII for Windows (SPSS Japan Inc., Tokyo, Japan). The time-course changes were compared using repeated measures analysis of variance (ANOVA). The data from each period were compared by a paired comparison method using the Student’s t -test. Differences were considered significant at P < 0.05.