Correlation between clinical and magnetic resonance imaging (MRI) findings in temporomandibular disorders

This study was carried out to determine the value of Magnetic resonance imaging (MRI) as a diagnostic tool in patients with temporomandibular disorders. The clinical presentation and MRI findings on 88 temporomandibular joints belonging to 44 symptomatic patients were retrospectively studied. The disk position, configuration and signal intensity; mandibular condyle morphology and signal intensity; temporomandibular joint space and surrounding soft tissue abnormality were assessed. The correlation between the clinical and MRI findings was statistically analyzed using Fisher’s exact (1-sided) test. Pain in the temporomandibular region was the most common clinical presentation, it accounts for 64% of cases. There was significant correlation between pain, and disc displacement with no reduction (DDWNR) and condylar hyperlaxity (p = 0.04, 0.03, respectively), as well as between clicking and each type of DD (p = 0.00). Statistically significant relationship was also found between tenderness, and DDWNR and presence of joint effusion (p = 0.02, 0.03, respectively) as well as between limitation of mouth opening and condylar marrow edema (p = 0.02). Causes of temporomandibular disorders can be well defined by clinical examination. However, MRI can be preserved for patients with pain in whom an initial medical conservative oral treatment failed in order to exclude other pathological process.


INTRODUCTION
Tempomandibular joint (TMJ) is a synovial joint and the diseases that affect other joints such as disk displacement (DD), degenerative joint disease, inflammatory arthritis, infection and synovitis can affect TMJ.Temporomandibular disorders are the most common causes of facial pain after toothache (Parnes et al., 2006).It had been reported that its etiology is multi-factorial and still widely disputed in literature (Emshoff et al., 2003).However, several studies demonstrated that DD (Tallents et al., 2002;Katzberg et al., 1980) and muscular disorders affecting the masticatory system are the most common *Corresponding author.E-mail: salryalat@yahoo.com.
the most common causes of these disorders (Emshoff et al., 2003;Carlsson, 1999).The initial examination used to image TMJ is usually plain radiograph and conventional tomography, since arthritic changes and congenital bone abnormalities are visualized well on these imaging modalities.Computerized tomography (CT) scan has the advantage in allowing a perfect visualization of the osseous components of the TMJ (Baily et al., 1990).
Several authors considered that MRI is the imaging modality of choice in temporomandibular disorders as it provides detailed information regarding the disc, joint space, and adjacent soft tissue structures (Emshoff et al., 2003;Rao, 1995).Therefore, the aims and reasons of this retrospective study determined the correlation between clinical presentation and MRI findings, to identify the

Selection criteria
The criteria for including a patient in the study were the presence of pain in the temporomandibular region and presence of TMJ pain during palpation as well as with jaw function.Patients with ear problems and typical or atypical neuralgic facial pain were excluded.The patients were referred to our MRI unit for the evaluation of presence of DD or adjacent soft tissue anomalies that could be the source of patients' symptoms.

Imaging technique and interpretation
All MR imaging were obtained with a 1.5 T Magnetom vision plus machine (Model of machine: Siemens, Germany) using bilateral TMJ surface coil.Our protocol consisted of oblique sagittal plane proton density and T2 weighted images at closed and then at open mouth.The images were taken for each side in each mouth position (closed and open) at angles perpendicular to the long axis of the mandibular condyle as determined by axial scout view image.A total of nine slices for each side in open and close position were obtained.The parameters used for proton density images were, slice thickness of 3 mm; repetition time, 2500 ms; echo time, 20 ms; field of view, 160 mm; and acquisition matrix size, 202 × 256.For T2-weighted images, the repetition time was 2900 ms, and the echo time was 80 ms.Both TMJs were examined for disk position, disk configuration, signal intensity; morphology and signal intensity of mandibular condyle, presence or absence of joint effusion in the temporomandibular joint space, and signal intensity of surrounding soft tissues.Disk mobility was not assessed as CINE MRI is not available in our machine.The disk was considered normal if its posterior band was at 12 o'clock position relative to the mandibular condyle on close mouth position according to the criteria proposed by Katzberg and Westesson (1993); dumbbell-like configuration and hypointense homogenous signal.It was considered an abnormal position if the posterior band of the disk was in an anterior position relative to the superior part of the condyle.It was considered displaced anteriorly with reduction (DDWR) when the disk returns back to normal position on opened mouth.However, disk displacement without reduction (DDWNR) was considered when the displaced disk had the same position in close or open position.
Disc configuration was considered abnormal if it was of uniform thickness (biplanar), having a thicker central part (biconvex), or showing an enlargement of its posterior band.Mandibular condyle was considered normal if it was rounded shape; it was considered edematous if its signal was bright on T2 weighted sequence.All MRI examinations were reported by two general radiologists who were unaware of clinical information and working together in consensus with MRI experience of 15 to 18 years.

Statistical analysis
Fisher's exact (1-sided) test was used to define the relationship between each clinical presentation and MRI findings.It was also used to define the presence of an association among patients' symptoms as well as among MRI findings.P value < 0.05 was considered statistically significant using SPSS 16 software package for statistical analysis.

Statistical results
On testing the relationship between the clinical presentation and MRI findings, a statistically significant relationship was found between pain and DDWNR and condylar hyperlaxity (p = 0.04, 0.03, respectively), as well as between clicking and each type of DD (p = 0.00).Statistically significant relationship was also found between tenderness and DDWNR and presence of joint effusion (p = 0.02, 0.03, respectively) as well as between mouth opening limitation and condylar marrow edema (p = 0.02).Detailed statistical relationship and percentage rates of association of each sign and symptom, and MRI findings are shown in Table 3.There was no statistically significant association neither among patients' symptoms (p = 0.3 to 0.6), nor among MRI findings (p = 0.09 to 1).A significant relationship between tenderness and disk morphology was found (p = 0.02).

DISCUSSION
Dysfunction of the TMJ is a common clinical problem, and imaging of the temporomandibular region has become essential in identifying the origin of patients' symptoms.Seventy percent of our symptomatic patients demonstrated abnormalities in the temporomandibular region on MRI examinations.It had been reported that temporomandibular disorders are more common in female patients; the results of these studies were based on history and clinical examination (Gesch et al., 2004;Nassif and Hilsen, 1992).Although 70% of symptomatic patients in this study were females, abnormal MRI findings were seen in 45% females, and only in 25% male patients, respectively.Several authors described a relationship between psychological status of the patient such as depression and stress and temporomandibular disorders that may explain the difference in the frequency of symptoms and MRI abnormalities (Selaimen et al., 2007;Korszun et al., 1998).
It has been reported that DD can be seen in up to onethird of asymptomatic individuals (Kircos et al., 1987).Haley et al. (2001) demonstrated that 26% of DD were at the side without pain while this rate in our study was 43%.The results of the present study demonstrated that DD was the most common finding in symptomatic patient and that it compares favourably with the results of other studies (Emshoff et al., 2003;Tasaki et al., 1996).Farina et al. (2008) found a significant correlation between TMJ pain and MRI findings of DD, and that was only observed in our patients with DDWNR (0.04).The incidence of DD in painful subjects in their study was 82%, and in ours was 54%.Whyte et al. (2006) reported that DD is usually unilateral and reducible in asymptomatic patients while in symptomatic patients, it is bilateral and reducible in 76% of cases.Our results demonstrated that 83% of bilateral DD were reducible.In general, the reducible displaced disks were more common than the non-reducible disks