Association between the clinical features of and types of temporomandibular joint ankylosis based on a modified classification system

This study aimed to describe the clinical features of different types of traumatic temporomandibular joint (TMJ) ankylosis. Seventy-one patients with 102 ankylosed joints were retrospectively reviewed and categorized into four groups according to the grades of severity: type I, non-bony ankylosis of the joint with almost-normal joint space; type II, lateral bony ankylosis marked by a normal joint space that coexists with a radiolucent line; type III, complete bony ankylosis of the joint characterized by only a radiolucent line; and type IV, extensive bony ankylosis without any radiolucent line. The period of ankylosis, maximal mouth opening (MMO), rate of complications, and histopathological changes were compared among groups. Intergroup comparison showed significant differences in the clinical features of MMO and the incidence of complications (p < 0.05). Younger trauma patients tended to develop more severe types of ankylosis than older patients. Additionally, long post-trauma periods were related to the development of severe ankylosis. MMO was highly negatively correlated with the severity of ankylosis. Significant differences were noted among the four types of ankylosis. Younger trauma patients with long post-trauma periods tended to develop more severe TMJ ankylosis, experience more complications, and face more challenges in treatment than older patients.

www.nature.com/scientificreports www.nature.com/scientificreports/ Therefore, this study aimed to describe the clinical features of different types of TMJ ankylosis as per our proposed classification and validate the correlations between the severity of clinical features and the type of disease.
Results types of tMJ ankylosis. As per our method of classification of the 71 cases ( Fig. 1), 11 patients had type I, 21 patients had type II, 25 patients had type III, and 14 patients had type IV ankylosis. The coverage rate was 100%, and the Kappa value was 0.922 (p < 0.05), indicating an extremely high interobserver agreement. The mean patient age was 24 (range 5-63) years, and the mean age at the occurrence of maxillofacial trauma was 16 (range 1-60) years (Fig. 2). Type IV ankylosis was observed in patients with facial trauma aged <20 years alone; similarly, type III ankylosis was observed in patients injured age <30 years alone. On the other hand, type I ankylosis was found in injured patients aged >20 years alone, whereas type II ankylosis was found in injured patients of all ages.
Clinical characteristics. Of the 71 patients, 21 presented with mandibular asymmetry, 22 presented with mandibular retrusion, and 5 experienced OSAS. All patients presented with limited mouth opening; the maximum interincisal distance at presentation ranged 0-30 mm. Table 1 summarizes results of clinical examinations as well as rates of complications in each group. χ 2 and Fisher's exact tests showed no statistical significance in the mandibular asymmetry among types II, III, and IV (p > 0.05). In contrast, significant differences were observed in mandibular retrusion and combined OSAS between type III and type IV patients (p < 0.05).

Figure 1.
Computed tomography images and representations of the four types of temporomandibular joint ankylosis. Type I is nonbony ankylosis of the joint, with the fossa and condyle clearly seen but with scattered callus. Type II is lateral bony ankylosis of the joint, with bony fusion on the lateral side of the joint. The medially displaced condyle, residual disc, and fossa form a pseudarthrosis. Type III is complete bony ankylosis of the entire joint with a radiolucent line inside the fusion area but no recognizable condyle and fossa. Type IV is extensive bony ankylosis, with complete disappearance of the joint and with no radiolucent line. Histopathological features. In type I ankylosis, the joint space was composed of irregular fibrous tissue.
The cartilage cell layer on both sides was thickened, similar to the fibrocartilage tissue, and the middle of the joint space was filled with irregular fibrous connective tissue. Cracks and fissures were observed between fibrous tissues (Fig. 5). In type II ankylosis, the joint space was composed of layers of mostly fibrous and some cartilaginous tissue. The cartilage cell layer on both sides of the bone was composed of calcified cartilage close to the bone and uncalcified cartilage toward the joint space. One or more undulating hematoxyphil lines were present between the two layers of cartilage tissue. Fibrous tissue and capillary vessels were present in the residual joint space. The joint space in type III ankylosis was predominantly filled with cartilage, which was similar to the hyaline cartilage tissue. The chondrocytes were markedly hypertrophic and calcified, and extended toward the bone marrow cavities and joint space, suggesting that the osseous was active. In addition, fibrous connective tissue was observed and seemed to be involved vascular invasion. In type IV ankylosis, the joint space disappeared and was therefore not considered in the histological examination.

Discussion
It is important to correctly estimate the severity of TMJ ankylosis in order to accurately evaluate the difficulty of surgery, precisely plan the resection, and assess the amounts of autogenous or alloplastic material required to fill the gap created in gap arthroplasty 11 . Extensive research has focused on determining the ideal algorithm for classification of the severity of TMJ ankylosis 12 . Our study proposed a modified classification system for TMJ ankylosis to allow provision of classification-based treatment suggestions for the condition. Using our classification system, all the patients in this study were grouped according to differences in the change in joint space and the radiolucent zone, and we found significant correlations and remarkable differences in the clinical features among the four groups.
In general, TMJ ankylosis can be classified by location (internal or external), the type of tissue involved (fibrotic or bony), and extent of ankylosis (partial or total) 5 . In 1986, Sawhney 13 categorized TMJ ankylosis into four types according to radiography results. This method based on plain radiography has several advantages such as objective evaluation of scope of bony mass. However, based on our clinical experience, plain radiography has certain limitations in evaluating the condition of the residual condylar head and the degree of ossification, which thereby affects the prediction of probable surgical difficulties. The 2002 classification proposed by El-Hakim et al. 14 was based on CT scans and focused on adjacent anatomic structures. In 2011, He et al. 15 proposed a classification for evaluating the medially displaced condyle. Although previous systems have been useful for classification of TMJ ankylosis, these do not analyze sufficient data on manifestations of clinical features and disease development. www.nature.com/scientificreports www.nature.com/scientificreports/ One of the classification standards-the joint space-coexists with a medially displaced condyle and constitutes the pseudarthrosis of the condyle-fossa 16 . Zheng et al. 17 investigated the disc using magnetic resonance imaging and showed that the disc was discernible in all joints in lateral bony ankylosis, but with a degree of deformity and an intermediate position. Operative observation showed that the disc was generally intact and reduction was possible by traction. The optimal inter-positional material may have served as a barrier to prevent fusion of the condyle with the glenoid fossa [18][19][20] . Another classification standard-the radiolucent zone-is   www.nature.com/scientificreports www.nature.com/scientificreports/ similar to the joint space, but differs due to its demonstration of insufficient calcification of the bony fusion area. Radiography showed that the radiolucent zone presented as a low-density line inside the fusion area. Its histopathological feature was the residual joint space manifested as a compound tissue structure of fibrous, cartilaginous, and osseous tissue. In our previous research 21 , CT findings were consistent with histologic results. Type II and III ankyloses were cartilaginous bony ankylosis, with similar components but with different degrees of severity. This could explain why some patients with bony ankylosis of TMJ can open their mouth to some degree but others cannot.
Ferretti et al. 22 evaluated the joint morphology on coronal CT images and concluded that all the ankylosed joints exhibited a persistent rudimentary joint space. Ferretti et al. 22 and Casanova et al. 23 inferred that the radiolucent area inside the lesion represented a remnant of the inter-articular disc. We believe that the radiolucent zone may represent the course of bone healing of two traumatic articular surfaces with interference from the mouth-opening movements. This zone can be regarded as a mark of bony fusion development to a certain stage. Further research using magnetic resonance imaging is needed to verify this hypothesis.
In our study, when the radiolucent zone disappeared, the ankylosis progressed to the end stage (type IV in our classification). Both young age at occurrence of facial trauma (<10 years) and time after trauma occurrence (>10 years) were risk factors for the severe phase of ankylosis. Therefore, the growth and development of the maxillofacial region were inevitably affected, resulting in a high percentage of facial deformity. The treatment of type IV ankylosis should be directed toward the release of ankylosis and correction of the facial deformity. Furthermore, simultaneous or secondary mandibular advancement should be considered to enlarge the collapsed posterior airway space and relieve OSAS.
This study found a correlation of age at injury with prevalence and severity of TMJ ankylosis and concomitant complications, which is consistent with the finding of Kaur 24 . In addition, Anyanechi 25 showed that patient age was correlated with TMJ ankylosis. Compared to mature adults, children had greater growth and reparative potential in the lamellar bony structures because they have rich vascularization. Allori et al. 4 reported that the pediatric mandible is characterized by a broad condyle and thin cortical bone, which predispose children to intra-capsular comminuted fractures. In contrast, the elongated condylar neck in adults limits fractures to the extracapsular space. Bello et al. 1 found that 17 of 23 patients with TMJ ankylosis had combined facial deformity, all of which had suffered maxillofacial trauma at <5 years of age.
In conclusion, this study highlights significant differences among the four types of traumatic TMJ ankylosis in terms of clinical features. Younger trauma patients with long post-trauma periods tend to develop more severe TMJ ankylosis, manifest lower MMO, experience more complications, and face more challenges in treatment than older patients. Our proposed modified classification system is efficient in evaluating the disease severity and developmental stage of traumatic TMJ ankylosis on the basis of CT findings. Moreover, all four types of ankylosis have precise definitions and distinctive criteria, which make it possible to achieve highly reproducible results and a high Kappa value.
Despite our important findings, this study has a few shortcomings. First, although the collection of imaging data on facial trauma would provide a deeper understanding and establish the relationship between the type of traumatic TMJ ankylosis and features of facial fracture, it was difficult to achieve. Second, MRI analysis was the best way to describe the joint disc of different type of ankylosis, but unfortunately, most of patients lacked of MRI examinations. In our further investigation, we will attempt to evaluate the articular disc based on MRI examinations. Thirdly, we were unable to conduct a follow-up for our patients, even though a follow-up is essential to further optimize our classification-based therapeutic strategy. Restricted by being a retrospective study, the conclusions of this study are less convincing than those of prospective studies. In further work, animal experiments could be used to test the conclusion whether younger trauma patients with long post-trauma periods tended to develop more severe TMJ ankylosis. Fourthly, TMJ ankylosis is an uncommon disease entity so this study is constrained by small numbers. A multicenter international work is needed to reconfirm the conclusions of this study. The study sample consisted of 42 men with 66 ankylosed joints and 29 women with 36 ankylosed joints. Of the 71 patients, 40 had unilateral ankylosis and 31 had bilateral ankylosis.
Basic information on age, gender, period after facial injury, MMO, concomitant facial deformity, and OSAS was obtained for all patients. CT examinations (helix with 1.25-mm slice thickness; Bright Speed 16, GE Healthcare, Buckinghamshire, UK) were performed, and multiplanar reformation was used to generate coronal CT images of the TMJ.
Classification. Considering the change in joint space and the radiolucent zone on coronal CT, all ankylosed joints were categorized into one of four types (Fig. 1): type I, non-bony ankylosis of the joint with an almost-normal joint space and without bony fusion or a radiolucent line; type II, lateral bony ankylosis of the joint with lateral bony fusion and a radiolucent line inside the fusion area (medially discernible joint space coexisting with a displaced condyle); type III, complete bony ankylosis of the whole joint with a radiolucent zone inside the fusion mass but no clearly defined joint space; and type IV, extensive bony ankylosis with no clear www.nature.com/scientificreports www.nature.com/scientificreports/ definition of the joint or no radiolucent zone. When a patient with bilateral TMJ ankylosis had a different type of ankylosis on each side, the classification was based on the more critical side. All ankylosed joints were categorized by two surgeons who were members of the Department of Oral and Maxillofacial Surgery and were engaged in TMJ ankylosis research for >5 years. Any difference in opinion between the two surgeons was resolved by a third surgeon. The CT images were evaluated by three surgeons in a single-blind test.
surgery. During the surgery, the ankylosed mass was carefully exposed and harvested using osteotomy planes placed 3 mm above and 10 mm below the fusion area. The whole bony fusion mass was cut from the condyle to the temporal bone, with the residual joint space inside. Intraoperative findings such as movement of the remnant condyle, the characteristics of the residual joint space, and the condition of the articular disc were recorded.
Histopathological examination. All specimens were fixed in 10% buffered formalin for 48 h and decalcified with 20% buffered ethylenediaminetetraacetic acid on a shaking table until elastic deformation occurred. After dehydration in a series of graded alcohols, the specimens were paraffin-embedded, cut into 4-μm serial sections, and stained with hematoxylin and eosin. statistical analysis. The coverage rate was calculated, and the reliability of the test was determined by the coefficient of internal consistency using the Kappa value. The Spearman correlation coefficient was used to evaluate the relationship between the severity of ankylosis and clinical features, including age at occurrence of maxillofacial trauma, duration after facial trauma, and MMO. The χ 2 and Fisher's exact tests were used for intergroup comparisons of the proportion of facial deformity and OSAS among the different types of ankylosis. Statistical analysis was conducted using SPSS 19.0 software (SPSS Inc., Chicago, IL, USA). The level of statistical significance was set at p < 0.05. ethical approval. This study was approved by the ethics committee of Peking University School and Hospital of Stomatology [No. PKUSSIRB-201416095]. All study procedures were conducted in accordance with relevant guidelines and regulations. All participants provided written informed consent to publish identifying information/images and clinical records.