JIA is a chronic inflammatory disease that can involve all synovial joints in the body. The TMJ is one of the important sites of involvement. In all subtypes of JIA, TMJs may be involved unilaterally or bilaterally at any stage of the disease, and may appear as the first involved joint or the only involved joint [4].

The frequency of TMJ joint involvement in JIA varies between 17 and 87% [4, 19]. The wide range of frequency of TMJ involvement is due to the fact that the criteria suggesting TMJ involvement are not fully established and the use of screening methods varies between centers. In our study, the prevalence of TMJ involvement in JIA patients was 52.8%. Since our study was single-centered and the number of patients was small, the frequency of TMJ involvement in our study does not reflect the frequency of involvement in our country.

TMJ arthritis in JIA is asymptomatic in 65–85% of patients [7, 11, 12]. The severity of clinical signs and symptoms is directly related to the severity of inflammation, and the sensitivity of clinical signs and symptoms is low but the specificity is high [7]. In our study, 71.5% of the patients with TMJ involvement had no clinical signs and symptoms suggestive of TMJ involvement. All of the symptomatic patients had TMJ MRI involvement. These findings obtained in our study showed that clinical signs and symptoms have low sensitivity but high specificity in the demonstration of TMJ involvement, in accordance with other studies.

Since the majority of patients with TMJ involvement in JIA are asymptomatic, arthritis shows insidious progression in the joint and in most patients, the involvement is detected after the development of a chronic and degenerative process. In a study conducted by Arvidsson et al. İn 2010 in adult patients with JIA, it was found that 80% of patients had chronic TMJ involvement [20]. In our study, 28.5% of patients with MRI findings of TMJ involvement had acute arthritis and 71.5% had chronic arthritis. Because of the anatomic feautures TMJ, degenerative processes in the joint affect the growth plate in the early period [9, 10]. Degenerative changes in the growth plate during the active growth period disrupt the growth and development of the mandible and cause irreversible, chronic, difficult to treat findings such as micrognathia, retrognathia, malocclusions, jaw asymmetry, chronic jaw pain, which significantly reduce the quality of life. Among the patients with TMJ involvement included in our study, one had micrognathia and one had retrognathia, that is, 9.52% of the patients with involvement had chronic irreversible sequelae affecting the external appearance and causing loss of function, but these findings were obtained from patient records, detailed orthodontic examination information of all patients is needed to give exact percentages.

Studies have shown that polyarticular subtype, early age of onset, long disease duration, elevated ESR and ANA positivity increase the risk of TMJ involvement, while HLA B27 positivity decreases the risk of TMJ involvement [11, 21]. In 2009, Arygropoluo et al. showed that systemic subtype, early age of onset, long disease duration, presence of multiple joint involvement and elevated ESR increased the risk of TMJ involvement [22]. In the studies conducted by Weiss et al. in 2008 and Billiau et al. in 2007, no correlation was shown between JIA subtypes, laboratory parameters, number of joints involved, age at onset and duration of disease and TMJ involvement [23, 24]. IPolyarticular JIA, high number of involved joints and high ESR are conditions associated with the presence of severe inflammation in the body in JIA patients, and it is thought that the higher frequency of TMJ involvement in these patients is related to this high inflammatory status [11, 21, 22]. HLA B27 positivity is usually seen in patients with JIA associated with enthesitis with a late age of onset and TMJ involvement is rare, and the lower incidence of TMJ involvement in these patients is explained by these conditions [21]. Inflammation in the body is more severe in JIA patients with RF positivity and joint involvement in these patients is followed by destructive chronic changes, but studies have not found a relationship between RF positivity and TMJ involvement [11, 21, 22, 24]. In our study, no association was found between RF positivity and TMJ involvement. In 2014 studies by Gorska et al. and in 2017 studies by Kalaykova et al. TMJ involvement is more common in patients with poor treatment response and in JIA patients who need multiple drug therapy [25, 26]. In our study, it was determined that the treatment compliance and responses of patients with TMJ involvement were worse than those of patients without involvement and that these patients were in need of multidrug therapy. Poor treatment response and the need for more aggressive treatment is associated with high disease activity and the higher frequency of TMJ involvement in these patients may be explained by high disease activity.

In our study, in order to determine the risk factors for TMJ involvement, logistic regression analysis was performed with the variables of age at diagnosis, mean disease duration, total number of joints involved, presence of clinical symptoms and signs suggestive of TMJ involvement, HLA B27 positivity, elevated ESR and mean ESR values, CRP positivity, use of NSAIDs, steroids, biological agents, and disease subtypes, and high number of joints involved and high sedimentation were found to be the risk factors for TMJ involvement. The risk factors for TMJ involvement differ among the studies and no definite risk factors have been identified. For these reasons, it is thought that TMJ involvement may occur in all JIA patients at any stage of the disease and screening for TMJ involvement in all JIA patients is recommended [22, 24].

There is a potential for selection bias due to the retrospective design of our study and the clinical rationale behind selecting patients for MRI. The patients who underwent MRI were those identified as having risk factors for TMJ involvement or presenting with symptoms, based on the clinical practices and literature available at the time. This selection process might have introduced bias into our study, potentially leading to a focus on more severe cases of TMJ involvement. Furthermore, our study is limited by being single-center and having a small sample size, which may affect the generalizability of our findings. Additionally, as a retrospective study, we were unable to control certain aspects of the data collection, such as the specific MRI reporting methods and sequences used, or whether two radiologists read the scans independently to assess inter-reader reliability. This limitation restricts our ability to evaluate the consistency of the MRI findings. Moreover, due to the retrospective nature of the study, it remains unclear which patients will progress and which will respond to therapies. The lack of detailed treatment protocols and follow-up in terms of treatment effects further limits the conclusions that can be drawn about treatment efficacy. These are important considerations for future research to address.