Subjects

This prospective study included 55 female patients whose jaw deformities had been corrected surgically at the university hospital in Niigata, Japan, between July 2012 and September 2014. Agreement to participate in this clinical study was obtained. Only female patients were included because the normal levels of bone metabolic markers and cephalometric items differ between males and females. No cases of cleft palate or craniofacial syndrome were included. The mean age of the subjects at surgery was 23 years (range 15–39 years). The study protocol was approved by the ethics committee of the study institution in Niigata, Japan, and informed consent was obtained from the subjects.

Cephalometric analyses

Skeletal morphology was examined using a lateral cephalogram (CX-150W; Asahi Roentgen Industrial Co., Ltd, Kyoto, Japan) that was taken with the teeth in centric occlusion immediately before surgery. The lateral cephalogram was traced and digitized by translating the landmark points. Nine measurements were made using cephalometric analysis software (CephaloMetrics AtoZ; Yasunaga Computer System Co., Ltd, Fukui, Japan): facial angle, A–B plane angle, mandibular plane angle, y -axis, gonial angle, ramus inclination, sella–nasion–A-point (SNA), sella–nasion–B-point (SNB), and A-point–nasion–B-point (ANB) ( Fig. 1 ).

Measurements made on a lateral cephalogram. (1) Facial angle: angle between the FH plane and facial plane. (2) A–B plane angle: angle between line A–B and line N–Pog. (3) Mandibular plane angle: angle between the FH plane and mandibular plane. (4) Y -axis: angle between the FH plane and line S–Gn. (5) Gonial angle: angle between the ramus plane and mandibular plane. (6) Ramus inclination: angle between the ramus plane and FH plane. (7) SNA: angle between line S–N and line N–A. (8) SNB: angle between line S–N and line N–B. (9) ANB: difference between SNA and SNB. (FH, Frankfort horizontal; A, A-point; B, B-point; N, nasion; Pog, pogonion; S, sella; Gn, gnathion.).

The patients were divided into three groups according to the type of anteroposterior skeletal pattern: a group of 39 patients with skeletal class III malocclusions with or without open bite (class III group), a group of nine patients with skeletal class II malocclusions with or without open bite (class II group), and a group of seven patients with skeletal class I malocclusions who had facial asymmetry and/or open bite (class I group). The mean ± standard deviation (SD) age at surgery was 22 ± 6 years (range 15–39 years) in the class III group, 27 ± 9 years (range 17–38 years) in the class II group, and 22 ± 5 years (range 16–28 years) in the class I group.

Measurements of bone metabolic markers

As bone formation markers, the serum levels of osteocalcin (OC) and bone alkaline phosphatase (BAP) were measured by immunoradiometric assay (IRMA) and chemiluminescent enzyme immunoassay (CLEIA), respectively. As bone resorption markers, tartrate-resistant acid phosphatase isoform 5b (TRACP-5b) in serum and deoxypyridinoline (DPD) in urine were measured by enzyme immunoassay (EIA). Blood and urine samples were collected between 9:00 a.m. and 11:00 a.m. because bone metabolic markers are subject to circadian influence. Reference values used were the normal ranges established in healthy Japanese subjects.

Measurement of bone properties at the calcaneal bone

Quantitative ultrasound (QUS) is a portable and accurate technology for measuring bone properties without radiation exposure. QUS measurements at the calcaneal bone were made using an ultrasonometer device (Achilles A-1000 EXP II; GE Healthcare Japan Corporation, Tokyo, Japan). The patient’s right foot was placed on the foot positioner of the device, because it is considered that there is no significant difference in healthy human calcaneal bone density between the left and right foot. A transducer on one side of the heel converted an electrical signal into a sound wave, which passed through the patient’s heel. A transducer at a fixed distance on the opposite side of the heel received the sound wave and converted it into an electrical signal. This device measured the speed of sound (SOS; metres per second (m/s)) passing through the heel and the broadband ultrasound attenuation (BUA; decibels per megahertz (dB/MHz)) as a measure of frequency-dependent attenuation of the ultrasound wave passing through the heel. The stiffness index, which is a variable derived from the combination of SOS and BUA and accurately reflects BMD, was calculated using the following equation: stiffness index = [(0.67 × BUA) + (0.28 × SOS)] − 420.

Measurements of bone volume and bone properties at the mandibular condylar process

Bone volume and bone density of the mandibular condylar process were measured in 43 patients for whom helical computed tomography (CT) of the maxillofacial region was performed around the same time as the other examinations. Helical CT data were obtained in DICOM (digital imaging and communication in medicine) format. A multi-detector CT scanner was used (Aquilion ONE Global Edition; Toshiba Medical Systems Co. Ltd, Ohtawara city, Japan), with a tube voltage of 120 kVp, average tube current of 125 mA, 500 ms scan time, and thickness of 1 mm.

Three-dimensional images of the maxillofacial region were reconstructed and re-aligned to extract images of the mandibular condylar process using image processing software (ZioTerm 2009; Ziosoft, Inc., Tokyo, Japan). The Frankfort horizontal (FH) plane was used as a reference plane; this was formed by the bilateral uppermost points on the bony external auditory meatus (porion) and lowest point on the right inferior borders of the bony orbit (orbitale). The three points were determined manually using the sagittal plane, coronal plane, and axial plane of a multiplanar reconstruction (MPR) image ( Fig. 2 ). As the target region, the mandibular condylar process was defined as the area above the base plane that was parallel to the FH plane and passing through the most inferior point of the mandibular notch ( Fig. 3 ).

Setting of the Frankfort horizontal plane on the reconstructed three-dimensional CT images. (A) Axial view. (B) Coronal view. (C) Midsagittal view. The bilateral uppermost points on the bony external auditory meatus (porion) and lowest point on the right inferior borders of the bony orbit (orbitale) were identified three-dimensionally and the Frankfort horizontal plane was defined as the plane passing through the bilateral porions and right orbitale.

Definition of the mandibular condylar process. The mandibular condylar process was defined as the area between the Frankfort horizontal plane (A) and the plane parallel to the Frankfort horizontal plane and passing through the mandibular notch (B).

Measurements of the mandibular condylar process were performed using image analysis software (ImageJ; National Institutes of Health, Bethesda, MD, USA). Voxel data from above the base plane were used ( Fig. 4 ). The outer shape of the condylar process was clearly extracted with a threshold of over 350 Hounsfield units (HU), classified as D3 according to the Misch classification. The bone region in the extracted outer shape was then identified if voxel data were over 150 HU, classified as D4 according to the Misch classification. The size of voxel data was 0.43 × 0.43 × 0.50 mm, and the total volume (TV), bone volume (BV), and bone density (BV/TV) of the condylar process were calculated from the voxel data. The average value of measurements obtained on both sides was taken as the representative value for the subject.

Measurements of bone volume and bone properties at the mandibular condylar process. The A plane is the base plane and the dotted line B is one of the cross lines between CT slices and the base plane. Voxel data from above the base plane were used. The outer shape of the condylar process was clearly extracted with a threshold of over 350 Hounsfield units (HU). The bone region in the extracted outer shape was identified if voxel data were over 150 HU. The size of voxel data was 0.43 × 0.43 × 0.50 mm, and the total volume (TV), bone volume (BV), and bone density (BV/TV) of the condylar process were calculated from the voxel data.

Statistical analysis

Mean and SD values of all measured parameters were calculated. Since it could not be assumed that some parameters had a normal distribution, Kruskal–Wallis one-way analysis of variance and pairwise comparisons were used to assess the significance of differences among the groups; Spearman’s correlation analysis was performed to assess the correlations between parameters. Probabilities of less than 0.05 were accepted as significant. Data were analyzed using IBM SPSS Statistics 21 for Windows (IBM Japan Ltd, Tokyo, Japan).