Use of Cone Beam Computed Tomography in the Differential Diagnosis of Jaw Bone Lesions

Mailon Cury Carneiro (  mailoncury@gmail.com ) USP Campus de Bauru: Universidade de Sao Paulo Campus de Bauru https://orcid.org/0000-0003-39526002 Elen de Souza Tolentino Universidade Estadual de Maringa Lorena Borgononi Aquaroni Universidade Estadual de Maringa Milenka Gabriela Quenta Huayhua USP Campus de Bauru: Universidade de Sao Paulo Campus de Bauru Bernardo da Fonseca Orcina USP Campus de Bauru: Universidade de Sao Paulo Campus de Bauru Verônica Caroline Brito Reia USP Campus de Bauru: Universidade de Sao Paulo Campus de Bauru Izabel Regina Fischer Rubira-Bullen USP Campus de Bauru: Universidade de Sao Paulo Campus de Bauru


Introduction
The introduction of cone beam computed tomography (CBCT) has greatly changed the dentists' practice in diagnosing maxillofacial alterations, since access to sectional images has become faster and easier.
Previously, multiplanar images were obtained mainly by multislice helical computed tomography (MHCT) and magnetic resonance imaging (MRI), which are costly equipment usually restricted to hospitals. However, CBCT units have the smallest physical dimension, lowest cost, lowest radiation dose, and easier operation, which led to their rapid acceptance [1].
Before the advent of CBCT, panoramic radiography (PR) was the most common imaging tool in oral and maxillofacial surgery [2]. Although surgical practice is successful in the vast majority of cases, the limitations of this technique include variable magni cation, distortion, overlapping structures, air shadows, and inaccurate images of structures that are not within the focal plane [2][3][4]. CBCT has overcome these limitations due to its wide range of applications in dentistry, from diagnosis to treatment planning [1,2]. Depending on the eld of view (FOV), the CBCT exam covers a large focal area of the facial skeleton, as well as a small focal area of clinical interest [1].
For the evaluation of cysts or tumours, the CBCT originates three dimensions that are registered through multiplanar reformatting (axial, coronal, and sagittal) [5], which provides important information about the presence and buccal-lingual extension of bone pathologies, bone resorption, sclerosis of the adjacent bone, cortical expansion, presence of internal or external calci cations, and proximity to anatomical structures [6].
In addition, small cut thicknesses, such as 0.1 mm, allow a better visualization of the bone margins of a lesion [5,7].
Although CBCT has been widely studied in recent years, studies on its applicability and diagnostic value in intraosseous lesions of the jaws are scarce. Some studies have reported isolated clinical cases [8,9], speci c pathologies [10][11][12], or are based on MHCT images [13][14][15]. Thus, the present study aims to describe in detail the imaging ndings of CBCT examinations of patients with intraosseous lesions in the maxillofacial region.

Methods
The research was approved by the Human Research Ethics Committee (number 1,683,193/2016).

Sample
A retrospective study was carried out with a sample consisting of CBCT exams belonging to the Department of Stomatology of a Brazilian University. All preoperative images of intraosseous pathologies, odontogenic or not, con rmed via microscopic examination, of patients presented between 2008 and 2017 were reviewed.
There were no age restrictions. Endodontic apical lesions and bro-osseous lesions were excluded from the study.

CBCT
The patients were examined using two CBCT systems (3D Accuitomo® 170 -J. Morita Corp., Kyoto, Japan, and New Generation i-Cat® -Imaging Sciences International, Hat eld, PA, USA), according to the protocols recommended by the manufacturer and the speci cations of each system (120 kVp, 3-8 mA, 0.3 mm voxel size and 80 kVp, 6 mA, 0.08 mm voxel size, respectively). The FOV varied according to the location and extension of the lesion. Images were obtained in different planes (sagittal, coronal, and axial) at 1 mm slice thickness and 0.5 mm slice interval in both devices.

Data analysis
For all cases, coronal, sagittal, axial, and panoramic reconstructions were evaluated, as well as parasagittal cuts and 3D reconstruction. The following data were collected: sex, age, location of the lesion, density and size of the lesion, relation to an unerupted tooth, margins, cortical expansion, cortical perforation, tooth displacement, root resorption, presence of calci cations, and proximity to anatomical structures. The assessed variables are shown in Table 1. In addition, Fig. 1 was designed to illustrate the main characteristics evaluated.

Subdivision of the lesions
The lesions found were categorized as aggressive and non-aggressive according to their behaviour. The Odontoma (17.86%) was also more prevalent in females (60%) and in the mandibular region (80%), with a mean age of 20.4 years. The average lesion size was 1.3 cm. In 60% of the cases, the lesion was related to an unerupted tooth, and the lesion was hyperdense in all the cases, with well-de ned hypodense halos. In addition, tooth displacement caused by the lesion was observed in 60% of cases. Cortical perforation, cortical expansion, and root resorption were not observed in any tomographic examinations. All lesions were diagnosed as complex odontomas.
The dentigerous cyst (10.71%) had an average size of 2.3 cm and occurred at an average age of 20.3 years in affected individuals. Females were more affected (66.7%) than males and the mandible was the only affected region (100%). All lesions were related to unerupted teeth and had well-de ned limits. Cortical expansion was absent in most lesions (66.7%) and there was no cortical perforation or tooth displacement. Root resorption was present in two cases (66.7%).
Tomography revealed ameloblastomas (7.15%) with an average size of 4.14 cm, being more prevalent in females (57.1%), with an average age of 43.5 years, presenting in 100% of hypodense and multilocular cases, affecting exclusively the mandible. Additionally, in all the cases, there was a relationship between the lesion and an unerupted tooth, and an expansion of the bone cortices. The borders were well de ned and half of the cases had a hyperdense halo. Perforation of the cortical bone, as well as root resorption, was present in 50% of the cases. It was not possible to observe calci cations or dental displacements in the analysed lesions. No cases of unicystic ameloblastoma were observed.
The lesions were subdivided into aggressive (odontogenic keratocyst and ameloblastoma) and nonaggressive (odontoma, ameloblastic bro-odontoma, dentigerous cyst, cementoblastoma, adenomatoid odontogenic tumour, and calcifying epithelial odontogenic cyst), based mainly on ndings related to clinical behaviour and recurrence of lesions. When compared, there was a statistically signi cant difference only for the bone cortex perforation variable, which was more prevalent in aggressive lesions. For the other variables, no statistically signi cant differences were observed (Table 2).

Discussion
CBCT images provide important information about the presence and extent of bone and/or tooth resorption, cortical expansion, presence of calci cations, tooth displacements, and involvement of anatomical structures [4,6]. With the different reformats, it is possible to obtain a better view of the bone margins of the lesion in three dimensions [2,5].
Measurements on CBCT images are acceptably accurate, in addition to revealing the direction in which the expansion is taking place, contributing to the planning of surgical treatment, especially in the early stages of expansion, when it may be di cult to observe the direction of growth by only clinical examination [4], in contrast to the PR where the image is enlarged [5].
In the present study, odontogenic keratocyst was more prevalent in the mandible than in the maxilla, in line with previous studies [18 -23]. Likewise, ameloblastomas were also located in the mandibular region [18,20,[22][23][24]. All dentigerous cysts occurred in the mandible; however, they may be common in the maxillary region [25]. For odontomas, 80% of the cases were observed in the mandible, contrary to previous studies that demonstrated a higher prevalence in the maxilla [18-20, 22, 23].
Among the lesions considered aggressive, odontogenic keratocyst and ameloblastoma caused cortical bone perforation in 71.4% and 50% of cases, respectively. Root resorption and displacement were observed in 22% and 29% of odontogenic keratocyst cases, respectively, and in half of ameloblastoma cases, root resorption of the teeth involved was observed. The main characteristic that differentiated both lesions was the association with unerupted teeth and the expansion of bone cortices, present in all cases of ameloblastoma.
As for odontogenic keratocysts, in most cases (64.3%), there was no expansion of the cortices or association with an unerupted tooth (57.1%).
Among the non-aggressive lesions, the dentigerous cyst was not associated with cortical perforation or tooth displacement. However, most cases showed root resorption (66.7%). The margins have always been well de ned in non-aggressive lesions, as well as in odontomas, which in 60% of cases, were associated with unerupted teeth or caused tooth displacement without root resorption. Calci cations were observed in ameloblastic bro-odontoma, adenomatoid odontogenic tumour, and calcifying epithelial odontogenic cyst.
CBCT allows the detection of subtle hyperdensities, favouring the orientation of the hypotheses for the diagnosis of calci ed lesions, which may also include the calcifying epithelial odontogenic tumour, cementoossifying broma, and bro-osseous lesions.
On comparing aggressive and non-aggressive lesions, there was a statistically signi cant difference only for the perforation of cortical bone, which was more prevalent in aggressive lesions, suggesting that this characteristic is an important indicator of pathological behaviour, guiding the elaboration of diagnostic hypotheses and the plan of treatment.
Oral surgeons can rely on PR if the margins of benign lesions are well de ned. However, when the margins are not well de ned, CBCT is the best tool for diagnostic assistance [5]. A 'benign' lesion appearance in a PR can reveal characteristics of malignancy in thin slices scanned on CBCT. Tomographic images can identify such irregular margins and provide accurate and reliable information in the early stages of a malignant lesion [2,7]. CBCT is as reliable as MHCT for detecting bone invasion by malignant lesions [26,27]. However, they are not applicable for the analysis of soft tissue tumours, and in this case is more appropriate acquisition of MHCT in 'soft tissue windows' or MRI [7,28].
The absence of cases of compound odontomas, simple bone cysts, or even the small number of cases of dentigerous cyst, can be justi ed by the fact that CBCT is not requested in all cases. Generally, the indication