Spectroscopic differences in posterior insula in patients with chronic temporomandibular pain

2.1 Participants

Thirty-six female patients with chronic regional TMD pain [rTMD; 17 patients; median (25/75 percentile) age: 40 (30/44) years] or generalized pain including TMD pain [gTMD; 19 patients; median (25/75 percentile) age: 43 (40/56) years] were included in this study. All patients were referred to the Department of Orofacial Pain and Jaw Function, Malmö University due to orofacial pain. The inclusion criteria for the chronic rTMD pain were TMD pain duration for more than 3 months; TMD pain intensity of 4 or higher on a 0–10 Numerical Rating Scale (NRS) where 0=no pain and 10=most possible pain; a diagnose of myalgia or myofascial pain with referral according to Diagnostic Criteria for Temporomandibular Disorders (DC/TMD) [25]. In the rTMD group, no chronic pain in other locations than in the orofacial and neck area was allowed. The inclusion criteria for the gTMD group were to fulfill the same criteria as for the group with chronic rTMD and in addition to have a generalized pain condition diagnosed by a medical doctor with general pain intensity of NRS 4 or higher and duration of generalized pain for more than 3 months (Table 1).

Ten healthy women [HI; median (25/75 percentile) age: 36 (26/51) years] were included in the study as a control group and recruited at the Faculty of Odontology, Malmö University, Sweden. Inclusion criteria for HI were female gender; absence of pain condition and absence of chronic disease (Table 1).

Exclusion criteria for patients and HI were metal splinters in the body, pacemaker, other implants or claustrophobia contradicting the MR examination, age younger than 18 years, pregnancy, malignancy, pulmonary disease, psychiatric disorder (except depression and anxiety), opioid medication and obesity (body mass index >30).

2.2 Questionnaires

All participants completed a DC/TMD questionnaire prior to the clinical examination. This questionnaire comprised assessment of degrees of depression (Patient Health Questionnaire-9), anxiety (Generalized Anxiety Disorder-7), number of physical symptoms (Patient Health Questionnaire-15), stress (Perceived Stress Scale-10), catastrophizing (Pain Catastrophizing Scale), characteristic pain intensity (mean of pain intensity for reported worst, current and average pain; Graded Chronic Pain Scale), pain-related disability (mean of how much facial pain changed the patient’s ability to participate in daily activities, social activities and work; Graded Chronic Pain Scale), and grade (Graded Chronic Pain Scale) as well as pain locations (Pain Drawing) [25].

2.3 Clinical examination

All participants underwent a clinical DC/TMD examination [25]. This examination comprised standardized assessment of pain locations, headache, mouth opening capacity, pain on mandibular movement, pain on palpation and temporomandibular joint noises. The examiners in the present study had all went through the 2-day course and were calibrated in the use of DC/TMD on a specialist level. The department of Orofacial Pain and Jaw Function at Malmö University, Sweden is one of three official DC/TMD Training and Reliability Centers in the world, accredited by the International Network for Orofacial Pain and Related Disorders Methodology (INfORM, part of IADR: http://www.iadr.org/INfORM/DC-TMD). One of the five dentists performed the examinations on each individual. In addition to the DC/TMD examination, the following factors were assessed:

Pressure-pain threshold (PPT) over the masseter muscle was assessed using a handheld electronic algometer (Somedic Sales AB, Sollentuna, Sweden) with a stimulation area of 1 cm². The PPT was defined as the amount of pressure needed to produce the slightest sensation of pain. The pressure was applied with linearly increased pressure with a rate of 30 kPa/s perpendicularly to the skin surface over the body of the right and left masseter muscles [26]. The participant was asked to press a button when the sensation changed from pressure to pain and the pressure value was recorded. PPT was assessed three times for each site and the mean of the two last measurements was used. The first dorsal interosseous muscle of the dominant hand was used as an extra-segmental control site.

Temporal summation to pressure stimuli was assessed by applying a pressure corresponding to PPT 10 times with a pressure rate of 30 kPa/s and an interval between each pressure application of 1 s over the masseter muscle with the lowest PPT. The participants were asked to grade the pain intensity on a 0–10 NRS for the first and last application of pressure.

Neck range of movement capacity (hypo-, normal or hyper-mobility), pain on neck movement (yes/no), pain on neck palpation (yes/no) and pain radiating from the neck to the face (yes/no) were recorded by examining the levator scapulae, sternocleidomastoid, trapezius and the occipitalis muscles.

2.4 Magnetic resonance imaging

Within a week from the clinical examination all participants underwent non-contrast enhanced MRI on a 3T MR scanner (Siemens MAGNETOM Trio, Erlangen, Germany) using a 12-channel head coil. The following sequences were acquired: T1-w sagittal and transverse images (magnetization prepared rapid gradient echo) with TI=900 ms, TR/TE=1,900/3.5 ms on 176 slices with isotropic resolution 1.0×1.0×1.0 mm and T2-w coronal images fluid attenuation inversion recovery with TI=2,500 ms, TR/TE=9,000/89 ms, 30 slices with 5 mm thickness, in-plane resolution 1.2×0.9 mm².

2.5 Magnetic resonance spectroscopy

All participants went through a single-voxel ¹H-MRS examination using the following parameters: point-resolved spectroscopy repetition time 2,000 ms, echo time (TE) 30 ms, voxel size 2.0×1.0×1.0 cm. To enable calculation of absolute metabolite concentrations, water-suppressed and non-water-suppressed spectra were acquired. In this study, we chose the single voxel spectroscopy technique with a short TE of 30 ms, as this technique result in a spectrum with more metabolite peaks for MI, Glu and Gln and also allow a more accurate quantification of the metabolites. The individual volumes were placed in the right and left posterior insula cortices. The metabolite concentrations of NAA, tCr, Cho, MI, Glu and Glx were analyzed. The data were analyzed on a computerized analysis system, i.e. the LC model (LCModel; Oakville, ON, Canada) [27]. Metabolite levels were calculated as absolute concentrations, normalized to the water signal and reported in arbitrary institutional units (AU). Metabolite concentrations were used for statistical analysis from the LC model if the Cramér–Rao bounds were less than 20%. In addition, the ratios NAA/tCr, Cho/tCr, Glu/tCr and MI/tCr were calculated.

2.6 Statistics

Non-parametric statistics were used due to the characteristics of the pain-related variables. For descriptive statistics, median and 25th and 75th percentiles as well as number of observations are reported.

The significance of differences between two or three groups was calculated using Mann-Whitney U-test and Kruskall-Wallis test, respectively. The correlation coefficients between two variables and their significance were calculated using Spearman’s ranked correlation test (r_s). A probability level of p<0.05 was considered as significant. There was no compensation for multiple testing performed in order to reduce the risk of missing relevant results.

3.1 Clinical characteristics of TMD pain groups

Demographic data and distribution of diagnoses for chronic TMD pain are presented in Table 1. Table 2 shows the clinical findings in the three groups.

Patients with gTMD pain had significantly higher pain-related disability (p=0.008), degree of depression (p=0.001), anxiety (p=0.017), physical symptoms (p<0.001), stress (p=0.005) and number of sites with referred pain on palpation (p=0.026) than rTMD pain (Table 2). In addition, patients with gTMD pain had significantly lower masseter muscle PPT (p=0.038) and maximum voluntary mouth opening capacity with (p=0.047) and without pain (p=0.025) than patients with rTMD pain (Table 2).

Patients with rTMD or gTMD pain had significantly higher characteristic pain intensity (p<0.001), pain-related disability (p<0.001), chronic pain grade (p<0.001), degree of depression (p<0.001), anxiety (p=0.001), number of physical symptoms (p<0.001), stress (p=0.002) and number of sites with referred pain on palpation (p=0.002) than HI. In addition, patients with rTMD and gTMD pain had significantly lower maximum voluntary mouth opening without pain (p=0.002) than HI (Table 2).

3.2 Magnetic resonance imaging findings

When evaluating the conventional MR images for pathology, mild white matter lesion (1–5 white matter lesions) burden were present in three of 17 patients with chronic rTMD pain and in three of 11 HI. Moderate lesion burden (5–10 white matter lesions) was present in one of 17 patients with chronic rTMD pain, in two of 19 patients with gTMD and in one of 11 HI. One of the HI had a 1 cm² large area of unspecific gliosis in the parietal lobe and a corpus pineal cyst was seen in one of the patients. No brain atrophy, infarct or other pathological brain lesions where present in the remaining study participants examined.

3.3 Relations between spectroscopy and clinical findings

Table 3 shows spectroscopy data for absolute concentrations of metabolites and ratios in the right and left posterior insula. The only significant difference was found between patients with rTMD or gTMD, who had significantly higher levels of tCr in the right posterior insula than HI (p=0.029; Table 3). All other differences between the groups did not reach the level of significance.

Because there were no other differences between the TMD pain groups, the rTMD and gTMD groups were pooled to test for relationships between clinical characteristics and spectroscopy. There were significant positive relations between metabolite concentrations and metabolic ratios in the posterior insula and clinical findings (Table 4).

Cho was negatively correlated to maximum mouth opening capacity with or without pain [r_s=−0.42, n=28, p=0.031; (Fig. 1) and r_s=−0.48, n=28, p=0.034, respectively] as well as PPT at the hand [r_s=−0.41, n=28, p=0.031; (Fig. 2)]. Glu was positively correlated to temporal summation (r_s=0.42, n=26, p=0.034).

Figure 1:

Scatter plot showing the relation between metabolic activity of Cho in the right posterior insula and the pain intensity on maximum mouth opening with pain (r_s=−0.42, n=28, p=0.031).

Figure 2:

Scatter plot showing the relation between pressure-pain threshold of over dominant hand and metabolic activity of Choline in the right posterior insula (r_s=−0.41, n=28, p=0.031).

4.1 Posterior insular cortex metabolism

The present study found that the concentration of tCr was higher in the chronic TMD patients than in the HI. All other investigated substances could not be shown to differ. The spectroscopic assessment was performed during rest and not during any painful stimuli. This could be an explanation to the lack of differences, i.e. there are probably only relatively small differences between chronic TMD pain patients and HI in the posterior insula in a resting state.

The Glu concentration in the posterior insular cortex in patients with rTMD pain and in HI was in accordance with what previously have been demonstrated in patients with local TMD pain and healthy volunteers [22]. The present study could not find differences in metabolite concentrations or ratios in the insular cortices between patients with rTMD pain versus patients with gTMD, which is supported by previous studies [22], [28]. However, other studies have demonstrated reduced Gln concentration in patients with TMD pain and increased Glu concentration in fibromyalgia patients [17], [22]. A possible explanation for this discrepancy is that the previous reports predominately measured Glu and Gln during or after painful stimulations, whereas the present study acquired spectroscopy data without prior painful stimulation of the patients. Potential changes in Glu and Glx concentrations would then probably have been more obvious in the present study if the patients had been subjected to painful stimulation prior to or during the spectroscopy examination.

4.2 Posterior insular cortex metabolism and chronic temporomandibular disorder pain

In the present study, emphasis was given to potential relations between spectroscopy findings and clinical characteristics of TMD pain. The study tested systematically, for the first time, if metabolic changes in the posterior insular cortex were associated with pain intensity, pain-related disability, referred pain, and TMD pain provoked by masticatory system function.

Cho and Glu were both related to aspects of pain sensitivity. Cho was related to lower general PPTs and lower capacity of mouth opening, with or without pain. In turn, Glu was related to stronger temporal summation of nociceptive mechanical stimulus. This means that lower levels of Cho and higher levels of Glu in the posterior insula are associated with a more pronounced pain sensitivity. Similar findings of increased concentration of Glu in the posterior insula have been seen in patients with fibromyalgia and found to be associated with PPTs [17]. Our and as well as those findings seen in fibromyalgia patients suggest that Glu is related to pain processing that can potentially be a result of focal increased levels of Glu in this region and an elevated excitatory glutamatergic neurotransmission within the insula.

The findings of higher levels of tCr in TMD pain patients than in HI is surprising since tCr is normally considered as a stable marker of cell energy. Nociceptive signaling within the spinal cord and parts of the brain is mediated through the spinothalamic pathway, which probably is modulated by higher cortical brain regions likes the insular cortex and ACC. The alterations in tCr might thus be explained by neuroinflammation, which has been suggested to be involved in chronic pain. Microglial activation and autoantibodies have been found in chronic regional pain syndrome [29], [30]. This is also supported by earlier studies demonstrating the concentration of both Cho and tCr to be two to three fold higher in glial cells than in neurons and may be elevated in conditions of neuroinflammation [31]. The elevated tCr in our TMD pain patients could relate to the fact that these patients presented with not only higher chronic pain intensity but also had more stress, anxiety and were more depressed than HI. Previous studies demonstrating elevated glial metabolites like MI, tCr and Cho in chronic pain patients with spinal injury, especially in those with high psychosocial impact [32], further supports our findings. Others have demonstrated relations between higher levels of glial markers such as MI with pain intensity in patients with spinal cord injury [33]. Also depression seems to influence tCr as both animal and human studies have demonstrated a change in tCr in hippocampus before and after electroconvulsive therapy [34], [35].

4.3 Measures of pain sensitivity in TMD pain groups

In the present study the rTMD and gTMD pain groups showed a more pronounced temporal summation than the HI. In addition, the gTMD pain group had lower PPT over the masseter and on the hand than the rTMD pain group. These findings correspond to central sensitization, where changes in functional properties of neurons in the central nervous system causes reduced pain thresholds and increased responses to temporal wind-up [36]. Also pain referral to areas not affected by stimuli is a sign of central sensitization [37], which in the present study is demonstrated by sites with referred pain on palpation. The gTMD pain group had a median of four sites with referred pain on palpation, which was more than what was found in the rTMD pain group and HI. The present study thus supports that central sensitization is clinically significant and manifest in patients with regional pain conditions and perhaps even more so in generalized pain conditions, as could be expected [38]. Temporal summation of mechanically-induced pain is generally elevated in fibromyalgia compared to HI [39]. This is supported by the present findings where the gTMD pain group had a higher increase in pain intensity during the wind-up test compared the HI. The technique, in this study, to assess temporal summation using a handheld algometer has previously been shown to successfully induce both superficial and deep temporal summation on the hands [40].

In the gTMD pain group, 89% were diagnosed with myofascial pain with referral and 10.5% with myalgia, compared to the rTMD pain group where 41% were diagnosed with myofascial pain with referral and 59% myalgia. This was an expected distribution of diagnoses since the patients in the gTMD pain group have widespread pain and most likely an associated more pronounced central sensitization than the regional TMD pain group.

The gTMD pain group had the highest pain-related disability, depression and stress scores. It was evident in the present study that psychological symptoms as well as augmented pain (wind-up and referral) were more common in the gTMD pain group compared to both the rTMD pain group and HI. This also coincides with the fact that cognitive-affective factors; reappraisal, anxiety and depression are critical for the experience of chronic pain [41]. This means that the symptoms of the gTMD pain group have a more severe impact of their social activities and be a handicap in every-day-life [42], [43]. Despite these differences in clinical phenotypes between rTMD and gTMD pain patients, then there were no significant differences in spectroscopic metabolic activity within the posterior insula suggesting that other brain areas or biomarkers are involved in the pathophysiology of chronic TMD pain. Further studies will be needed to elucidate the underlying neurobiology of rTMD and gTMD pain.

4.4 Methodological considerations

All but one patient in the gTMD pain group had fibromyalgia. The remaining one had a diagnosis of generalized pain syndrome. We included that patient in the gTMD pain group since fibromyalgia is a generalized pain disorder by definition [44].

In this study, we focused on the posterior insular cortices as these areas have previously demonstrated metabolic changes in TMD, fibromyalgia and systemic lupus erythematosus (SLE) patients. These are all diseases with a variable degree of chronic pain. In future studies a larger voxel volume is recommended as previous studies in patients with fibromyalgia [17] or TMD have demonstrated that this works well in the posterior insula region [22].

We chose not to compensate for multiple testing in order to reduce the risk of type II errors, i.e. the risk of excluding interesting and possibly relevant results. On the other hand, the risk of type I-errors, including false positive findings, increased accordingly. We therefore discussed each finding regarding its biological plausibility and relevance and we made only cautious conclusions from the findings that were considered to possess biological plausibility and relevance.

A technical limitation of the present study was that spectroscopic data were only obtained in the resting state and without painful stimulation. In addition, the small voxel volumes used in order to avoid influence of cerebral spinal fluid and mixing of signals from gray and white matter in the adjacent sulci, may have made the signal-to-noise ratio of the spectra low.