Researchers have shown that the reliability and validity of MRIs are acceptable relative to the anatomy of the TMJ when they are interpreted by experienced radiologists.^2–4 However, one of the primary indications for treatment of patients with TMDs is pain, including that of the TMJ. Therefore, it is necessary to determine whether MRI-depicted TMJ anatomical findings are related to TMJ pain.

Researchers have shown that disk displacements, or DDs, of the TMJ are present in pain-free people as well as on the pain-free side of people with unilateral TMJ pain.^5–8 However, in studies using T2-weighted MRIs, researchers have demonstrated an association between the presence of fluid density in the joint space and the presence of TMJ DDs.^9–17 In addition, Schellhas and colleagues¹³ speculated that this fluid density is an effusion and, as such, represents a pathological inflammatory process within the TMJs that causes TMJ pain and indirectly results in jaw and facial pain including headaches. The question arises whether MRI-depicted effusions are a radiographic marker for TMJ pain. Finally, although it is clear that TMJ DDs are present in pain-free people, it would be of interest to know if effusions are related to the presence or absence of a TMJ DD or to different stages of TMJ DD.

All subjects reported that digital palpation of the temporomandibular joint duplicated the jaw pain on the side of reported pain.

The purpose of this study was to evaluate the relationship between TMJ pain and clinical vs. MRI findings. These questions helped guide the investigation:

– When a patient complains of pain over the TMJ, what are the odds that palpating the TMJ will identify it as the source of the pain? What are the odds that the TMJ will have an MRI-detected effusion? What are the odds that an MRI-detected DD will be present?

– When a patient complains of TMJ pain when the TMJ is palpated, what are the odds that the TMJ will have an effusion?

– When an effusion of the TMJ is present, what are the odds that a DD also will be present?

	PATIENTS AND METHODS

TOP ABSTRACT PATIENTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
The subjects were recruited from the TMJ and Orofacial Pain Clinic, University of Minnesota, Minneapolis, and from the Department of Dentistry, Hennepin County Medical Center, Minneapolis. Eighty-five consecutively seen, consenting female patients, aged 16 to 49 years, who had unilateral jaw pain and who agreed to undergo MRI of the TMJ were studied. All patients complained of unilateral jaw pain localized to the TMJ/preauricular area and all had concurrent masticatory muscle pain. Two of us (D.H., E.S.) performed complete stomatognathic examinations in all patients; the interrater reliability for joint palpation was in the acceptable range and was consistent with earlier reports.¹⁸ The examinations included palpation of the TMJ and masseter muscle, according to the operational definitions for each examination item specified by the craniomandibular index.¹⁸

Palpation and MRI. The examination also included palpation of both TMJs when between two to three pounds of digital pressure was applied to the lateral and then posterior aspects of the TMJ. All subjects reported that digital palpation of the TMJ duplicated the jaw pain on the side of reported pain. Within one week of the clinical examination, all patients underwent bilateral MRI of the TMJs. No anti-inflammatory medications were prescribed and no self-help advice was given during the time between the clinical examination and the bilateral MRI.

T1- and T2-weighted MRIs were interpreted by the same board-certified radiologist (Q.A.), who was blinded to the patients’ complaints. Images were made in an open- and closed-mouth position in the sagittal plane as well as in a closed-mouth position in a coronal plane. The imaging was performed with a 1.5 T signal MR scanner (General Electric).

If any fluid density was identified within either joint space in the open- or closed-mouth T2-weighted images, we interpreted it to represent an effusion. A TMJ DD was considered to be present if previously reported criteria were fulfilled.³ The reliability of the radiologist to detect TMJ effusions and disk diagnoses yielded coefficients of 0.85 and 0.74, respectively.

Statistical analysis. The experimental group consisted of the 85 patients with unilateral TMJ pain. The contralateral pain-free TMJ, in these same patients, served as the matched controls. Complete data were missing for two patients. ² analysis and McNemar’s matched-pairs tests were used to evaluate whether any significant relationships existed between a patient’s report that palpation of the TMJ was painful and the presence of effusions, a patient’s report of a painful TMJ on palpation and the presence of a DD, and a patient’s report of a painful TMJ on palpation and subjective complaints of TMJ pain. P values were calculated using the matched analysis method. The odds ratio, or OR, is presented as a measure of the association between two variables. To control for multiple comparisons, we set the level of statistical significance at .01.

	RESULTS

TOP ABSTRACT PATIENTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
Seventy (84 percent) of 83 subjects reported experiencing pain on palpation on the side of the face reported to have pain at the onset of the study, which is a statistically significant relationship (P = .001). Only 22 (27 percent) of 83 patients complained of pain on palpation on the side of the face not reported to be painful at the start of the study. The OR for this relationship is 49. This means that if a patient complains of unilateral pain in the area of the TMJ, it is 49 times more likely that he or she will complain of pain on palpation on the side with pain than on the pain-free side.

The authors found that effusions and disk displacement often are present even when the temporomandibular joint is not painful.

Fifty-nine (69 percent) of 85 patients had an MRI-detected effusion on the side of the face reported to be painful. This is a statistically significant relationship (P = .001). By contrast, only 37 patients (44 percent) had an MRI-detected effusion on the side not reported to be painful. The OR for this relationship is 3.8, meaning that if the patient complains of unilateral TMJ pain, it is 3.8 times more likely that an effusion will be present on the painful side than on the pain-free side.

Finally, 68 (80 percent) of 85 patients had a DD on the side reported to be painful and 63 patients (74 percent) had a DD on the pain-free side. Thus, we found no statistically significant relationship between the patient’s side of reported pain and the presence of an MRI-detected DD (P = .332). The OR of 1.8 indicates that it is 1.8 times more likely to have a DD on the side with pain than on the side without pain.

Tables 1 through 4 demonstrate other relationships between variables, none of which are statistically significant.

	DISCUSSION

TOP ABSTRACT PATIENTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

We did not find a statistically significant relationship between a patient’s report of pain on palpation and the presence of effusions. One could speculate that effusions and pain on palpation are indirectly related via the inflammatory process that is variably present with painful joints. Joint pain may arise from development of abnormal mechanical stresses within the joint tissues or from accumulation of inflammatory mediators within the joint capsules.¹⁹ If either mechanical or inflammatory factors are present, then putting more tension on the joint capsule via joint palpation should be a more accurate method of detecting pain than using an indirect method such as MRI. This may be why joint palpation better identifies the painful joint than does MRI-detected effusion or DD. Future research may further evaluate whether any relationship between these variables actually exists.

Determining whether MRI-depicted fluid in TMJs represents an inflammatory process requires further investigation through synovial fluid analysis of effusions. It may turn out that effusions sometimes contain inflammatory mediators or that variable amounts of fluid normally are present in all TMJs; if so, detecting effusions would be without clinical consequence. A further confounding factor is that the size of the effusion may determine, in part, whether this fluid is inflammatory. Therefore, although previous investigations have suggested that TMJ effusions are a prominent source of TMJ pain, orofacial pain and even headaches, MRI-detected effusions should be considered of unknown diagnostic value until they are more fully investigated.^11–13

Other authors have proposed a strong relationship between DD and TMJ effusions and between DD and jaw pain.^10–17 We found no significant relationship between DD and the patient’s reported TMJ pain on one side and only a trend between the presence of a DD and TMJ effusions on the side of reported TMJ pain. Therefore, DD clearly is unrelated to the patient’s complaint of TMJ pain, since a DD can be present on either side (68 [80 percent] of 85 patients had a DD on the side with pain, while 63 patients [74 percent] had a DD on the pain-free side).

When reviewing the TMD literature, we found studies that showed that DD with reduction in the TMJ is common in the general population, and that the vast majority of these people are pain-free.^5–8 Recent MRI studies have shown that among asymptomatic subjects aged 9 through 25 years (that is, people without a history of TMJ noise, pain or limited opening), approximately 27 percent had a DD in the TMJ, and that among asymptomatic subjects aged 23 through 43 years, 32 percent had a DD in the TMJ.^7,8 Furthermore, TMJ noises (including clicking and popping), which typically indicate the presence of a DD with reduction, can be detected clinically in up to 30 percent of the general population.^20–24 It becomes clear, then, that the total number of adults with DD with reduction, with or without audible clicking and popping noises, probably is more than 50 percent of the population. Therefore, it is no surprise that finding a DD in a patient with TMJ pain can be an incidental finding.

DD with and without reduction. It is also clear that a small minority of patients with DD with reduction have a progressive DD that can advance to DD without reduction.^25,26 In this study, we divided the number of cases of DD into those with and without reduction and found a trend toward DD without reduction being associated with the patient’s side of reported TMJ pain (Table 1). When we divided the number of cases of DD into those with and without reduction and compared them with instances of effusions, a trend also was found between DD without reduction and effusions (Table 4). It is unclear whether these associations have any clinical significance.

It may be that DD without reduction produces more abnormal mechanical stresses or inflammatory chemical mediators than does DD with reduction and, therefore, results in pain and effusions in some TMJs. Conversely, the displaced disk in patients with DD without reduction is less mobile and more deformed than it is in patients who have DD with reduction. These more advanced anatomical changes simply may interfere with the normal synovial fluid circulation, causing localized collection of fluid. According to this reasoning, then, an effusion is simply a pooling of synovial fluid with no direct relationship to inflammation or pain. All of the above explanations may be true for different patients. To try to answer the above questions, the clinician needs to conduct TMJ fluid analysis.

Contradictory findings. The results of this study contradict those of previous studies. Other studies have used few or no asymptomatic controls, used the less painful side as controls for patients with bilateral jaw pain or ignored the less painful joint entirely.^10–14 In our study, the patients themselves served as their own controls. Prior investigations also did not state clearly how they determined the presence of joint pain, leaving in doubt whether the pain might be due to another source, such as muscular pain. Our patients clearly identified the TMJ as the primary area of their jaw pain (that is, when asked to localize their jaw pain, they pointed to the TMJ on the painful side and reported that when it was palpated, the pain duplicated their initial pain complaint). Therefore, we believe our patients had unilateral TMJ pain.

Another possible reason our results differ from those of other studies is that previous studies analyzed data as if the TMJs were independent. Such an analysis assumes that one TMJ has no influence over the contralateral TMJ. This perspective ignores the fact that the two TMJs are physically connected, and it has been shown that one TMJ can influence the other.⁵ If one analyzes the total number of joints rather than the number of subjects, the significance of any finding will increase. This is because the P value is a function of the sample size, and when the sample size increases, the P value decreases.

This fact is illustrated by the data in Table 3, in which the relationship between MRI-depicted effusions and DD on either the painful or nonpainful side is compared. In this instance, if the sample size is 170 joints rather than 85 people, the statistical significance of this relationship increases (P = .006), yet the OR remains approximately the same. Thus, if there is a dependency between joints, then this increase in significance is artificial. This may have been a problem in previous studies that ignored this potential dependency issue, which was controlled for in our study by having a matched case-control design. This design takes into consideration that the TMJs are physically connected and, therefore, may affect each other. This is an ideal situation in which to use the matched case-control design because each subject is his or her own control.

We have not proven that the two TMJs are dependent; it is difficult to do so given that the two sides of each subject differ in regard to pain. A better approach for evaluating this dependent-vs.-independent issue would be to identify subjects with two TMJs that are as similar as possible. For example, researchers could compare subjects with bilateral pain and pathology with subjects with no pain and pathology and then compare the frequency of clinical or radiographic findings between the two joints.

	CONCLUSION

TOP ABSTRACT PATIENTS AND METHODS RESULTS DISCUSSION CONCLUSION REFERENCES

 
The results of this study suggest that TMJ DD was not significantly related to a patient’s complaint of TMJ pain. MRI-depicted effusions were significantly related to the presence of TMJ pain. However, we do not recommend that clinicians use MRI-depicted effusions to assess for TMJ pain because of the large number of false-positive and false-negative findings, as well as the significant costs associated with conducting an MRI examination. This study indicates that TMJ palpation that elicits a complaint of pain that duplicates the patient’s initial pain complaint is superior to MRI as an indicator that the TMJ is the source of the pain.

Although we question the clinical utility of using MRI findings to predict the source of pain for patients with TMJ pain, MRI should be considered for patients if the initial history, examination findings or plain film radiographs suggest clinically significant pathology, such as a neoplasm. Finally, MRI of the TMJ can be considered when patients with pain are refractory to conservative therapy and if assessment of the hard and soft tissues via MRI will affect the diagnosis, prognosis or treatment plan.²⁵

View larger version (96K): [in this window] [in a new window]   Dr. Haley is an associate clinical professor, Division of TMJ and Orofacial Pain, Department of Diagnostic and Surgical Sciences, School of Dentistry, University of Minnesota, Minneapolis, and a staff dentist, Department of Dentistry, Hennepin County Medical Center, 701 Park Ave. S., Minneapolis, Minn. 55415. Address reprint requests to Dr. Haley.

View larger version (97K): [in this window] [in a new window]   Dr. Schiffman is an associate professor, Division of TMJ and Orofacial Pain, Department of Diagnostic and Surgical Sciences, School of Dentistry, University of Minnesota, Minneapolis.