Goldenhar syndrome is a rare disease with hemifacial microsomia and craniofacial disorders originating from the first and second branchial arches, such as ocular, auricular, and vertebral anomalies. The complexity and variety of the ways in which the disease presents itself usually need several examinations. In this study, we aimed to evaluate both craniofacial and vertebral skeletal anomalies and asymmetries between the nonaffected and affected sides in patients with Goldenhar syndrome by using cone-beam computed tomography.
Ten patients (7-14 years old; 6 boys, 4 girls) were evaluated via NewTom 5G cone-beam computed tomography (QR srl, Verona, Italy) with a large field of view (18 × 16 cm). Ten anatomic facial landmarks were identified to measure the following distances bilaterally: sella turcica (ST)-mandibular angle, ST-condyle, ST-mastoid, ST-mental foramen, ST-fronto zygomatic suture, ST-zygomatic temporal suture, ST-zygomatic facial foramen, ST-sphenopalatine fossa, mandibular angle-mandibular symphysis, and mandibular angle-condyle. The following 6 volumes were calculated bilaterally: orbit, maxillary sinus, condyle, external ear canal, middle ear, and internal auditory canal. These measurements were performed to assess skeletal asymmetries to compare the nonaffected side with the affected side by the Wilcoxon test. Cervical spine anomalies were classified into fusion anomalies and posterior arch deficiencies.
All patients showed a deficit of skeletal development on the affected side. Statistically significant differences (0.001 ≤ P value ≤ 0.043) between the nonaffected and affected sides were recorded for all measurements, except for ST-frontozygomatic suture, mandibular angle-mandibular symphysis, and maxillary sinus volume. Vertebral fusion anomalies and posterior arch deficiencies were found in 7 and 4 patients, respectively.
Cone-beam computed tomography with a large field of view was able to accurately identify craniofacial and vertebral skeletal anomalies, and to quantify asymmetries between the nonaffected and affected sides for an efficient maxillofacial treatment planning.
Goldenhar syndrome is a rare disease characterized by craniofacial disorders.
Patients are very young when they need several examinations and treatment planning.
Head and neck multislice spiral computed tomography is usually avoided for radio-protection reasons.
CBCT can exactly identify craniofacial and vertebral skeletal anomalies at 1 time.
CBCT could become the elective imaging technique for patients with Goldenhar syndrome.
Oculo-auriculo-vertebral dysplasia is a phenotypically variable developmental anomaly of facial structures originating from the first and second branchial arches. In addition to the mandibular and auricular abnormalities, the most severe cases show ocular or vertebral involvements. This is known as the Goldenhar syndrome. The typical presentation of Goldenhar syndrome includes epibulbar dermoids, microtia, vertebral anomalies, and hemifacial microsomia (85% monolateral).
Vertebral malformations are mainly represented by vertebral fusions and scoliosis; in the cervical area, they are detected in about half of the patients. Anomalies of the maxillary sinus and inner or medium ear are also frequent, whereas systemic abnormalities and psycho-physical motor underdevelopment are less common. The birth prevalence rate varies greatly among authors, ranging from 1 in 5,000 to 1 in 25,000. Etiology is multifactorial. It is partly linked to environmental factors during the fetal period, such as smoking, diabetes, and vasoactive drugs, and also partly due to genetic causes. Most chromosomal aberrations are sporadic, whereas only 2% to 10% of them are autosomal dominant or recessive.
The complexity and variety of the alterations pose the need for several examinations, such as panoramic radiography, cephalometric and cervical spine x-rays, and sometimes a volumetric imaging technique that allows a detailed analysis of bone structures, such as multislice spiral computed tomography (MSCT).
Although magnetic resonance imaging (MRI) does not depict cortical bone well, it has been recently used as an alternative tool because the MSCT of the cervical spine and head delivers a high x-ray radiation dose. Nothing has been published about a possible diagnostic role of cone-beam computed tomography (CBCT) in Goldenhar syndrome. This technique provides an adequate volumetric study of the bone structures of the head and neck areas with a high spatial resolution (0.075-0.4 mm isotropic voxel), allowing easy and accurate 2-dimensional and 3-dimensional (3D) measurements. Moreover, it involves relatively low radiation doses compared with MSCT and is only slightly affected by metal artifacts, which often occur in young people with ortho-gnatho-dontic oral metallic treatment. The aims of this retrospective study were to identify craniofacial and vertebral skeletal anomalies and to quantify the asymmetries between the nonaffected and affected sides in patients with Goldenhar syndrome by using CBCT with a large field of view (FOV) as a diagnostic tool.
Material and methods
From April 2012 to October 2016, 10 patients (6 boys, 4 girls; mean age, 9.9 years) ( Table I ) affected by Goldenhar syndrome with unilateral craniofacial involvement were examined via CBCT volumetric imaging. This study was approved by the research ethics committee of the University of Florence in Italy, and informed written consent was obtained from each patient. Scans were performed with NewTom 5G (QR srl, Verona, Italy), a horizontal CBCT unit equipped with an amorphous silicon flat-panel detector (20 × 25 cm). The patients lay in a supine position with competent lips and the mouth in habitual occlusion. The same protocol—called “standard regular” by the producer—was used for all examinations. It provided 110 kV, 3.39-5.38 mA, 18 × 16 cm FOV, 18-second scan time, and 3.6-second exposure time with the acquisition of 360 images (1 image for each rotation degree) at the lowest doable automatically set radiation dose. The largest FOV available was chosen to include both the facial skeleton and the vertebral spine up to the sixth cervical vertebra. All CBCT volumes were reconstructed with 0.3-mm isometric voxel size. The 0.4-mm thick axial sections were exported in DICOM format and analyzed by using OsiriX software (version 7.0; OsiriX, Geneva, Switzerland), implemented in a Macintosh operating system with a 21.5-in monitor (Power Macintosh G3; Apple, Cupertino, Calif).
|Patient||Age||Sex||Hemifacial microsomia||Aural abnormalities||Eye abnormalities||Vertebral abnormalities|
|1||7 y 1 mo||M||+||+||+||+|
|2||8 y 4 mo||M||+||+||+||−|
|3||9 y 2 mo||M||+||+||−||+|
|4||14 y 0 mo||F||+||+||+||−|
|5||13 y 6 mo||F||+||+||−||+|
|6||10 y 8 mo||F||+||+||−||+|
|7||11 y 6 mo||M||+||+||+||+|
|8||8 y 2 mo||M||+||+||+||+|
|9||7 y 11 mo||F||+||+||+||−|
|10||7 y 10 mo||M||+||+||−||+|
The anatomic points chosen for the 10 linear measurements and the 6 calculated volumes are shown in Figures 1 and 2 , respectively. Gonion, pogonion, and zygomatic-facial foramen were identified on the 3D surface-rendering reconstructions. The identification of the remaining 7 points took place on the 2-dimensional reconstruction planes (coronal, sagittal, and axial) by using the 3D multiplanar reconstruction function that allowed the simultaneous display of any structure on those 3 planes. The distances between sella turcica and the other points, except for pogonion, were calculated by the 3D curved multiplanar reconstruction function. The distances between gonion and pogonion and between gonion and the condyle were calculated also. The last 2 corresponded with the length of the mandibular body and the mandibular ramus, respectively.
Orbit, maxillary sinus, and condyle volumes were measured on the axial images, whereas the external ear canal, middle ear, and internal auditory canal volumes were measured on the coronal images. All volumes were calculated using the volume region of interest function after a region of interest surrounding the anatomic structures was depicted in each slice in which the same structures could be seen.
The cervical vertebrae anomalies were assessed by adaptation of the conventional x-ray classification of Sandham to CBCT. Sandham divided the vertebral anomalies into 2 groups: fusion anomalies and posterior arch deficiencies. Fusion anomalies were defined as incomplete separation with osseous continuities between 2 vertebral bodies or arches. Posterior arch deficiencies were represented by a discontinuity or interruption of the posterior arch of the vertebra due to incomplete ossification or dehiscence.
All examinations were evaluated by 3 skilled maxillofacial imaging observers (S.C., C.N., V.S.) (with 21, 10, and 4 years of experience, respectively). The assessment was carried out twice by each observer, with an interval of 3 months. Collected data were analyzed using SPSS statistical analysis software (version 23.0; IBM, Armonk, NY). Intraobserver and interobserver reliability values were determined for each designated parameter (affected side, nonaffected side, reduction factor, and differences between the nonaffected and the affected sides). For both linear and volumetric measurements, the intraclass correlation coefficient was used. According to Zidan et al, intraclass correlation coefficient values of 0.00 to 0.10, 0.11 to 0.40, 0.41 to 0.60, 0.61 to 0.80, and 0.81 to 1.0 represent no, slight, fair, good, and very good agreement, respectively.
The mean, median, standard deviation, and maximum and minimum values of each parameter were calculated for both linear and volumetric measurements among the 60 observations (2 evaluations by 3 observers for 10 patients). The reduction factor was the ratio between the nonaffected and affected sides as stated by Hirschfelder et al. Then, a descriptive analysis was performed. The differences between the nonaffected and affected sides for all linear and volumetric measurements were assessed by using the Wilcoxon signed rank test for paired data. For each analysis, a P value ≤0.05 was considered to be statistically significant.
Alterations of bone development were detected on the affected side in all 10 patients ( Tables II and III ). Statistically significant differences (0.001 ≤ P value ≤0.043) between the nonaffected and affected sides were observed for all linear and volumetric measurements, except for maxillary sinus volume and the distances between gonion and pogonion, and between sella turcica and the fronto-zygomatic suture ( P = 0.5). Intraobserver and interobserver reliabilities for both linear and volumetric measurements were very good. The intraclass correlation coefficient values were 0.84 ± 0.26 and 0.83 ± 0.28 for the linear measurements, respectively, and 0.96 ± 0.08 and 0.91 ± 0.13 for the volumetric measurements, respectively.
|Distance||Parameter||Mean (cm)||Median (cm)||SD (cm)||Minimum (cm)||Maximum (cm)|