Accepted: 24 January 2018

DOI: 10.1111/joor.12611

ORIGINAL ARTICLE

Clinical diagnosis of temporomandibular joint arthritis

P. Alstergren1,2,3,4 | M. Pigg1,5 | S. Kopp4

1
Scandinavian Center for Orofacial
Neurosciences (SCON), Malmö, Sweden Summary
2
Specialized Pain Rehabilitation, Skåne Evidence-­based clinical diagnostic criteria for temporomandibular joint (TMJ) arthri-
University Hospital, Lund, Sweden
tis are not available. To establish (i) criteria for clinical diagnosis of TMJ arthritis and
3
Orofacial Pain Unit, Faculty of
(ii) clinical variables useful to determine inflammatory activity in TMJ arthritis using
Odontology, Malmö University, Malmö,
Sweden synovial fluid levels of inflammatory mediators as the reference standard. A cali-
4
Section for Orofacial Pain and Jaw brated examiner assessed TMJ pain, function, noise and occlusal changes in 219
Function, Department of Dental Medicine,
Karolinska Institutet, Huddinge, Sweden TMJs (141 patients, 15 healthy individuals). TMJ synovial fluid samples were obtained
5
Department of Endodontics, Faculty of with a push–pull technique using the hydroxycobalamin method and analysed for
Odontology, Malmö University, Malmö, TNF, TNFsRII, IL-­1β, IL-­1ra, IL-­1sRII, IL-­6 and serotonin. If any inflammatory mediator
Sweden
concentration exceeded normal, the TMJ was considered as arthritic. In the patient
Correspondence group, 71% of the joints were arthritic. Of those, 93% were painful. About 66% of the
Maria Pigg, Department of Endodontics,
Faculty of Odontology, Malmö University, non-­arthritic TMJs were painful to some degree. Intensity of TMJ resting pain and
Malmö, Sweden. TMJ maximum opening pain, number of jaw movements causing TMJ pain and latero-
Email: [email protected]
trusive movement to the contralateral side significantly explained presence of arthri-
Funding information tis (AUC 0.72, P < .001). Based on these findings, criteria for possible, probable and
Swedish Council of Medical Research;
National Institute of Dental and Craniofacial definite TMJ arthritis were determined. Arthritic TMJs with high inflammatory activ-
Research, Grant/Award Number: R01- ity showed higher pain intensity on maximum mouth opening (P < .001) and higher
DE15420; Schering-Plough Corporation;
Swedish Dental Society number of painful mandibular movements (P = .004) than TMJs with low inflamma-
tory activity. The combination TMJ pain on maximum mouth opening and Contralateral
laterotrusion <8 mm appears to have diagnostic value for TMJ arthritis. Among ar-
thritic TMJs, higher TMJ pain intensity on maximum mouth opening and number of
mandibular movements causing TMJ pain indicates higher inflammatory activity.

KEYWORDS
arthritis, diagnosis, pain, synovial fluid, temporomandibular joint

1 | BAC KG RO U N D (function) and if present in children and adolescents, mandibular

growth arrest that may lead to micrognathia.3-6
Arthritis, ie articular tissue inflammation, in the temporomandibu- Inflammation is a complex, rapid, first-­line and highly unspecific
lar joint (TMJ) is a disorder due to either local or systemic factors. immune system response with the purpose to locate and eliminate
Examples of local factors are micro-­or macrotrauma secondary to pathogens and injured tissue as well as to promote tissue healing.
disc displacement or degenerative joint disease and infection.1,2 This reaction has a clear and important biologic purpose in the acute
Systemic factors may be systemic inflammatory disorders such as phase but may transfer into a chronic state with very unclear, if
rheumatoid arthritis (RA), psoriatic arthritis (PsA) or reactive arthri- any, biologic purpose. The unspecific nature of the response means
tis. Clinically, TMJ arthritis may present with articular pain and pain that regardless of cause, the reaction involves to a great extent the
in adjacent structures and reduced jaw mobility. Cartilage and bone same cells, mediators, enzymes, etc. The same inflammatory medi-
tissue destruction may result in occlusal changes (loss of anterior ators are therefore most likely involved in local and systemic TMJ
contacts between the upper and lower jaw) with impaired chewing arthritis,7 whereas the local concentrations of these mediators as

J Oral Rehabil. 2018;45:269–281. wileyonlinelibrary.com/journal/joor

© 2018 John Wiley & Sons Ltd | 269
270 | ALSTERGREN et al.

well as the specific composition may differ. Autoantibodies and au- activity, it may be possible to establish a diagnostic grading system
toinflammation may also contribute to maintenance of the chronic where the presence of TMJ arthritis can be considered. To achieve
inflammation.8 this, the objectives of the study were (i) to identify the clinical vari-
Since ancient times, inflammation has clinically been described ables with the highest sensitivity and specificity to diagnose TMJ
and diagnosed by the presence of the five cardinal signs swelling, arthritis using synovial fluid levels of inflammatory mediators as ref-
redness, warmth, pain and impaired function. This is sometimes erence standard and (ii) to establish variables that are clinically use-
adequate regarding acute inflammatory conditions such as peri- ful to determine the degree of inflammatory activity in the arthritic
coronitis and sun-­b urned skin. However, for chronic inflammation TMJ.
as well as for many other acute inflammatory states, these cardi-
nal signs are neither sufficient nor correct to describe, diagnose
2 | M ATE R I A L S A N D M E TH O DS
or monitor the inflammatory activity. These classical clinical signs
may be found in some sites with chronic inflammation at a certain
2.1 | Patients and healthy individuals
time-­p oint but in other sites they may be absent. For example,
chronic periodontitis seldom displays any of the cardinal signs A total of 219 TMJs from 141 patients and 15 healthy individuals
despite there being an ongoing inflammatory process, local im- were included in this study (Table 1). The samples were included
mune system activation and disease progression. In the case of from a large (approximately 1050 TMJ synovial fluid samples) set of
TMJ arthritis, ongoing and progressive chronic TMJ inflammation samples obtained between 1995 and 2009 at Karolinska Institutet,
causing tissue degradation and/or growth disturbance may also Department of Dental Medicine, Section for Orofacial Pain and Jaw
be pain-­free for substantial periods. At a given time-­p oint, the Function in Huddinge, Sweden.
clinical presentation may be anywhere on a continuum from no Inclusion criteria for the patients/TMJs were (i) ongoing TMJ
sign or symptom whatsoever to any combination of pain, swelling/ pain or impaired TMJ function or recent development of anterior
exudate, tissue degradation or growth disturbance. In addition, open bite (see below for definition) or event/disorder that may initi-
there is temporal variation in the inflammatory activity which ate or maintain TMJ arthritis; (ii) TMJ synovial fluid sample that ful-
also may cause a fluctuation in symptoms and signs in the chronic filled the sample quality criteria from one or both TMJs (if several
condition.9 eligible samples existed, the latest samples were included) and (iii)
For diagnosis of TMJ arthritis, clear and ideally specific clinical verbal consent. Exclusion criteria were age ≤18 years, current ma-
criteria should be defined. In rheumatology, a swollen or painful lignancies, TMJ surgery or trauma within 2 years before inclusion
joint leads to a diagnosis of definite synovitis in that particular joint. or recent intra-­articular corticosteroid injection in the TMJ (within
The TMJ differs to some extent from other synovial joints as the 3 months). Patients whose symptoms could be mainly related to dis-
TMJ is seldom swollen, seldom shows redness and the mechani- ease in other components of the trigeminal system (eg, toothache,
cal pain sensitivity over the TMJ is only weakly, if at all, related myalgia, neuralgia) were also excluded.
to an inflammatory intra-­articular milieu but rather to systemic in- For patients with rheumatic disorders, diagnoses were deter-
flammatory factors.10 The recently published Expanded taxonomy mined by a medical doctor with specialty in rheumatology at the
for diagnostic criteria for temporomandibular disorders (DC/TMD)9 Clinic of rheumatology, Karolinska University Hospital, Huddinge,
lists Arthritis under the classification Temporomandibular joint dis- Sweden. Diagnoses were determined using the American College
orders—Joint pain and states the following criteria: “TMJ pain and of Rheumatology (ACR) or European League Against Rheumatism
swelling, redness and/or increased temperature in front of the ear (EULAR) criteria.12-15
or dental occlusal changes resulting from articular inflammatory Fifteen healthy adults (>18 years of age): six females and nine
exudate (eg, posterior open bite)”. However, TMJ swelling, redness males, with a median (25th/75th percentile) age of 36 (31/44)
or increased temperature occurs very rarely,11 which severely limits years and recruited from staff and postgraduate students at the
the possibilities to identify TMJ arthritis based on cardinal signs. In Department of Dental Medicine, Karolinska Institutet voluntarily
fact, ongoing chronic TMJ inflammation, that is arthritis, may not agreed to participate. Inclusion criteria were a statement from the
show any of the cardinal signs although there is ongoing disease subjects that they did not have any chronic pain or inflammatory
progression. condition and that they considered their orofacial area to be healthy.
Earlier diagnostic studies on TMJ arthritis have been hampered Exclusion criteria were symptoms and signs of a chronic pain con-
by the lack of an established, valid and reliable reference standard. dition, TMJ dysfunction (TMJ resting pain, TMJ pain on jaw move-
Today, true synovial fluid concentrations of inflammatory mediators ment, painful TMJ clickings) and pharmacological treatments that
can be determined from TMJ synovial fluid samples obtained with a may interfere with pain or inflammation. Subjects with non-­painful
joint washing technique,7 and cut-­off values for healthy and inflamed joint clicking were allowed. Table 1 shows age and gender distribu-
7,10
joints are available. tions, diagnoses and medications.
The aim of this study was to establish clinical criteria for the diag- This project was approved by the regional ethical committee at
nosis of TMJ arthritis. By identifying valid clinical markers for arthri- Karolinska Institutet, Stockholm, Sweden (176/91; 310/97; 142/02;
tis per se together with determinants for the degree of inflammatory 03-­2004).
ALSTERGREN et al. | 271

TA B L E 1 Age, distribution of gender and diagnoses, duration of The operators were calibrated to each other, and the examination
disease and use of medication in 141 patients with various degrees was performed according to the clinical examination routine used
of temporomandibular joint (TMJ) arthritis and 15 healthy
in several previous studies.7,17-25 The examination performed in this
individuals
study comprised the following variables.
Percentiles
1. The patients were asked about pain in nine joint regions be-
Median 25th 75th n
sides the TMJ (neck, shoulders, elbows, hands, upper back,
Patients
lower back, hips, knees and feet), and the number of painful
Age
joint regions was recorded (score = 0-9).
Years 51 40 60 141 2. A 10-cm visual analogue scale (ACO, Stockholm, Sweden; score
Distribution of gender according to diagnosis 0-10) or a numerical rating scale 0-10 with end-points marked with
Rheumatoid 10/65 No pain and Worst pain ever experienced were used to assess the
arthritis (M/W) current degree of global pain intensity as well as TMJ pain intensity
Psoriatic arthritis 7/13 at rest, on maximum mouth opening and on chewing.
(M/W)
3. A score (0-4) for pain to digital palpation of the TMJ was adopted
Ankylosing 4/8 that involved evaluation of the lateral (palpation over the lateral
spondylithis
TMJ pole with the jaw in a resting position) and posterior (lateral
(M/W)
palpation posteriorly of the caput on mouth opening) aspect of
Other systemic 0/20
inflammatory the joint on each side. For each site, one unit was scored if the
diseases (M/W) patient reported pain upon palpation and two units if the palpa-
Monoarthritic 2/12 tion pain also caused a palpebral reflex, adding up to a maximum
conditions (M/W) score of 4 for each TMJ.
Duration of systemic disease 4. TMJ crepitus was assessed by digital palpation on maximum
Years 10 4 20 126 mouth opening (at least three times) and recorded as “absent” or
Duration of TMJ disease “present.” This corresponds to the definition in DC/TMD for crep-

Years 4 1 10 127 itus even though our examinations were performed long before
DC/TMD was published.
Medication (analgesic, antiinflammatory or immunomodulating)
5. TMJ pain on mandibular movements (maximum voluntary mouth
Paracetamol 19 (13%)
opening, ipsilateral and contralateral laterotrusions and protru-
NSAID 46 (33%)
sion) was recorded separately for each TMJ as present or absent.
Corticosteroid 23 (16%)
One unit was scored for each movement causing TMJ pain on
DMARD 34 (24%)
each side (score 0-4).
Anti-­TNF 0 (0%) 6. The degree of anterior open bite was used as a coarse clinical
Other biologic 0 (0%) marker of the degree of cartilage and bone destruction in the
Total 93 (66%) TMJ, and it was assessed by recording of the occlusal contacts on
Healthy individuals each side upon hard biting in intercuspid position (2 × 8 μm,
Age Occlusions-Prüf-Folie, GHM Hanel Medizinal, Nürtingen,
Years 36 31 44 15 Germany). The following scores were used on each side: 0 = oc-
Distribution of 9/6 clusal contacts including the canine, 1 = no contacts anterior to
gender (M/W) the first premolar, 2 = no contacts anterior to the second premo-

n, number of observations; M, men; W, women; NSAID, non-­steroidal lar, 3 = no contacts anterior to the first molar, 4 = no contacts
anti-­
inflammatory drugs; DMARD, disease-­modifying anti-­
rheumatic anterior to the second molar and 5 = no occlusal contact; scoring
drug; TNF, tumour necrosis factor. system adapted from Ref. 11. The sum of the scores on the right
and left side was used in the analysis as an estimation of the de-
This study followed the Standards for Reporting of Diagnostic gree of anterior open bite. None of the patients in our study was
Accuracy Studies (STARD) guidelines for reporting diagnostic accu- edentulous and the score thus ranged from 0 to 9. The presence
racy studies.16 of anterior open bite was defined as a score of ≥5.

2.2 | Assessment of subjective symptoms and 2.3 | Temporomandibular joint synovial

clinical signs fluid sampling
Each subject was clinically examined by one of two operators (PA, Immediately after the clinical examination, two experienced and
SK) immediately before the blood and TMJ synovial fluid sampling. specially trained operators (PA, SK) obtained all TMJ synovial fluid
272 | ALSTERGREN et al.

samples. A disposable needle (diameter 0.60 mm) was placed in the included.7,26 The small hydroxocobalamin interaction was com-
superior TMJ compartment after disinfecting the skin with 70% pletely compensated for by this procedure.
ethanol and 5% chlorhexidine and auriculotemporalis nerve block The median (75th/90th percentile) plasma level of TNF in healthy
with 2.0 mL Lidocaine (Xylocain® 20 mg/mL, AstraZeneca, London, individuals is 6 (12/18) pg/mL, whereas plasma concentration of IL-­
UK). TMJ synovial fluid samples were obtained with a push–pull 1β is undetectable (n = 31), with our assay. Normal serotonin levels
technique, and the amount of recovered synovial fluid in each sam- to be expected in human serum according to the manufacturer are
ple was quantified with the hydroxycobalamin method that has a 300 ± 700 nmol/L for males and 500 ± 900 nmol/L for females.
detection limit of <1% synovial fluid in the aspirate, as described
by Alstergren et al,7,26,27 The technique enables determination of
2.6 | Reference standard
the true synovial fluid concentration of the investigated mediators
in the aspirate after saline washing of the joint. Briefly, a washing TMJs with detectable concentrations of either TNF, IL-­1β or seroto-
solution consisting of 22% hydroxycobalamin (Behepan® 1 mg/mL; nin or with a TMJ synovial fluid concentration exceeding the 95th
Kabi Pharmacia, Uppsala, Sweden) in physiological saline (sodium percentile of that from healthy individuals regarding TNF soluble
chloride 9 mg/mL) was used as washing solution. Each TMJ was receptor II (TNFsRII), IL-­1 receptor antagonist (IL-­1ra), IL-­1 soluble re-
washed with a total of 4 mL washing solution through a valve con- ceptor II (IL-­1sRII) or IL-­6 were considered as the reference standard
necting the syringes for the washing solution and the aspirate. One for arthritic joints with ongoing inflammatory activity (Table 2).7,17
mL of washing solution was injected slowly, and as much as possi- This was based on data from TMJ synovial fluid from healthy indi-
ble was then aspirated back. The injection/aspiration was repeated viduals, where TNF, IL-­1β and serotonin were undetectable using the
a total of four times for each joint. To secure an intra-­articular nee- identical sampling procedure and identical assays. These mediators
dle placement, a small aspiration was initially conducted. If aspi- have been shown to be related to TMJ arthritis, TMJ pain as well as
ration of the washing solution was possible and the resistance in TMJ cartilage and bone tissue destruction.7,17,20-22,24,25
the syringe was minor during injection, then the placement of the To assess the degree of inflammatory activity, all included 219
needle tip was considered within the joint space.7,26,27 TMJs were divided into “No arthritis” and “Arthritis,” according to
The aspirates were evaluated for blood contamination (no, minor, the definition above. After that, the arthritic TMJs were further
clear or excessive) and centrifuged at 1500 g at 4°C for 10 minutes.7 divided into “Low inflammatory activity” and “High inflammatory
The supernatants were then aliquoted into tubes (specific for each activity.” High inflammatory activity was here defined as a TMJ sy-
mediator to be analysed) and frozen at −80°C. novial concentration of serotonin exceeding 37 nmol/L or a detect-
able TNF concentration.

2.4 | Blood sampling

2.7 | Statistics
Venous blood was collected and used for determination of rheuma-
toid factor level, erythrocyte sedimentation rate, serum level of C-­ For descriptive statistics, median values and 25th/75th percentiles
reactive protein as well as plasma or serum levels of inflammatory are presented. Data from patients and healthy individuals were
mediators. Rheumatoid factor titres below 15 IE/mL and C-­reactive compared with Mann-­Whitney U test. Sensitivity and specificity
protein levels below 10 mg/L were considered as zero values accord- for single clinical variables and combinations of clinical variables
ing to the standard procedures of the accredited laboratory at the (pain-­and function-­related variables) were calculated, as were the
Department of Clinical Chemistry at Karolinska University Hospital, positive and negative likelihood ratios. Logistic regression was used
Huddinge, Sweden.
TA B L E 2 Reference standard definition of temporomandibular
joint (TMJ) arthritis
2.5 | Analysis of mediators Cut-­off for
Mediator arthritis
The TMJ synovial fluid and blood plasma concentrations of TNF,
TNF soluble receptor II (TNFsRII), IL-­1β, IL-­1 receptor antagonist (IL-­ Serotonin, nmol/L >0
1ra), IL-­1 soluble receptor II (IL-­1sRII), IL-­6 and serotonin were de- Tumour necrosis factor (TNF), pg/mL >0
termined using commercially available enzyme-­linked immunoassays TNF soluble receptor II, pg/mL >2624
in which highly specific antibodies were used to detect the media- Interleukin-­1β (IL-­1β), pg/mL >0
tors (TNF, TNFsRII, IL-­1β, IL-­1ra, IL-­1sRII, IL-­6 ELISAs, R&D Systems, IL-­1 receptor antagonist, pg/mL >1168
Minneapolis, MN USA; Serotonin: Serotonin EIA-­kit, Immunotech IL-­1 soluble receptor II, pg/mL >2400
A Coulter Company, Marseille, France). The assay of synovial fluid
IL-­6, pg/mL >60
concentration of serotonin was modified to be applicable at con-
Cut-­off values are based on TMJ synovial fluid concentrations from
centrations between 1.6 and 5000 nmol/L. To compensate for
healthy individuals from our laboratory. If a mediator is detectable in
hydroxocobalamin interaction with the assay, the synovial fluid as- healthy TMJ synovial fluid, the 95th percentile value from healthy indi-
pirates were read against a standard curve with hydroxocobalamin viduals was used as cut-­off value.
ALSTERGREN et al. | 273

to calculate the influence of combinations of clinical variables on In the total material, 13 (7%) of the 139 joints that fulfilled the
the predictive value for the presence of arthritis and presented as reference standard criteria for arthritis were pain-­free. On the other
ROC curves, starting with combination of all included variables. hand, 53 (66%) of the 80 TMJs that did not fulfil the criteria for ar-
Acceptable validity for the multivariate tests was defined as an thritis showed some degree of pain (resting, movement or palpation;
area under the ROC curve (AUC) of ≥0.70. 28 After each calculation, Figure 1).
the variable with the lowest contribution to the total result was
omitted as long as the area under the ROC curve remained >0.70.
3.3 | Univariate analysis
According to this principle, the cut-­off values for each clinical vari-
able were determined and the sensitivity and specificity were cal- “Global pain intensity, TMJ resting pain and TMJ pain on maximum
culated accordingly. opening and number of jaw movements causing TMJ pain” were sig-
A separate calculation as described above was performed for nificantly higher in TMJs with arthritic than in non-­arthritic joints
joints without pain to calculate factors explaining the presence of (Table 4). Laterotrusion to the contralateral side was lower in patients
arthritis in non-­painful joints. with TMJs with arthritis than in patients with non-­arthritic TMJ ar-
In order to investigate the usefulness of the clinical variables for thritis (Table 4).
the assessment of inflammatory activity, the TMJs were grouped Crepitus was more prevalent and the laterotrusion to the contralat-
into joints with high (median or higher) or low synovial fluid con- eral side were significantly smaller in TMJs with non-­painful arthritic
centrations of the investigated mediators. The significance of the than in non-­arthritic joints (P = .005 and P < .001, respectively).
difference in findings between joints with high or low inflammatory
activity was calculated with Mann-­Whitney U test. Also, the signif-
3.4 | Multivariate analysis
icance of the correlations between TMJ synovial fluid levels of the
mediators and the TMJ pain variables was analysed using Spearman’s Logistic regression using the presence of arthritis as the dependent
ranked correlation. variable and the clinical variables “TMJ resting pain intensity, TMJ
In general, a probability level of P < .05 was considered as sig- maximum opening pain intensity, Number of jaw movements caus-
nificant. To compensate for multiple testing regarding differences ing TMJ pain and Laterotrusive movement to the contralateral side”
between patients and healthy individuals as well as differences be- as independent variables statistically explained the presence of ar-
tween joints with and without arthritis, a probability level of P < .01 thritis with an area under the ROC curve = 0.72 (n = 190; P < .001;
was considered as significant for those comparisons. Figure 2).
Logistic regression using the presence of arthritis as the depen-
dent variable among joints with no pain and the independent clinical
3 | R E S U LT S
variables Crepitus and Contralateral laterotrusion found that these
variables statistically explained the presence of arthritis with an area
3.1 | Signs and symptoms from the
under the ROC curve = 0.91 (n = 40; P < .001; Figure 3).
temporomandibular joint
Table 3 shows all investigated clinical and laboratory variables and
3.5 | Diagnostic sensitivity and specificity
the probability level of the differences between the patients and
healthy individuals. Table 5 shows the sensitivity and specificity of single clinical vari-
No healthy individual reported any global or TMJ pain. Patients ables and combinations of variables to identify arthritis in relation to
had more global and TMJ pain and reduced mouth opening capacity the reference standard.
compared to healthy individuals but the groups did not differ in de- The highest sensitivity for a single variable was found for TMJ
gree of anterior open bite. pain on mandibular movement: 0.71. Corresponding specificity was
0.48. The highest specificity for a single variable was found for
Contralateral laterotrusion < 8 mm: 0.83 (sensitivity 0.37).
3.2 | Reference standard
The highest overall sensitivity, 0.89, was found for observing
Table 3 shows the TMJ synovial fluid concentrations of TNF, one or more of the following: “TMJ resting pain or TMJ pain on
TNFsRII, IL-­1β, IL-­1ra, IL-­1sRII, IL-­6 and serotonin in the patients and mandibular movements or TMJ pain on maximum mouth opening
healthy individuals. In the patients, 71% of the TMJs compared to 4% or Contralateral laterotrusion <8 mm”. The corresponding speci-
(one joint) in the healthy individuals were considered as arthritic ac- ficity (ie, observing none of the variables in a non-­a rthritic joint)
cording to the reference standard definition used in this study. There was 0.39. The highest overall specificity was found for the com-
were no significant differences between men and women regarding bination TMJ pain on maximum mouth opening and Contralateral
TMJ synovial concentrations of TNF, TNFsRII, IL-­1β, IL-­1ra, IL-­1sRII, laterotrusion <8 mm: 0.89. The sensitivity of this combination was
IL-­6 or serotonin. Age and TMJ synovial fluid concentrations of TNF, 0.24.
TNFsRII, IL-­1β, IL-­1ra, IL-­1sRII, IL-­6 or serotonin were not significantly In the non-­painful joints, crepitus, short duration of TMJ symptoms
related. and low contralateral laterotrusion explained TMJ non-­painful arthritis
274 | ALSTERGREN et al.

TA B L E 3 Clinical and laboratory findings in 141 patients/195 temporomandibular joints (TMJ) with various degrees of TMJ arthritis and
15 healthy individuals/24 TMJs

Patients Healthy individuals

Percentiles Percentiles

Median 25th 75th % pos n Median 25th 75th % pos n P

Clinical findings
Related to the individual
Number of painful regions 0-­9 6 4 7 98 117 0 0 0 0 14 <.001
Global pain intensity NRS 0-­10 4 3 6 94 96 0 0 0 0 14 <.001
Maximum mouth opening mm 38 32 44 n.a. 128 49 48 60 n.a. 15 <.001
Anterior open bite 0-­9 0 0 2 n.a. 141 0 0 1 n.a. 15 .100
Related to the temporomandibular joint
Pain intensity, rest NRS 0-­10 4 2 7 84 167 0 0 0 0 24 <.001
Pain intensity, mouth opening NRS 0-­10 1 0 4 66 195 0 0 0 0 24 <.001
Pain on mouth opening 0/1 63 195 0 24 <.001
Pain on ipsilateral 0/1 40 182 0 24 <.001
laterotrusion
Pain on contralateral 0/1 47 182 0 24 <.001
laterotrusion
Pain on protrusion 0/1 49 182 0 24 <.001
Number of jaw movements 0-­4 2 0 3 71 195 0 0 0 0 24 <.001
causing TMJ pain
Pain score on TMJ palpation, 0-­2 0 0 2 69 182 0 0 0 17 24 <.001
lateral
Pain score on TMJ palpation, 0-­2 0 0 1 45 104 0 0 0 12 24 .003
lateral-­posterior
Pain score on TMJ palpation, 0-­4 1 0 2 71 182 0 0 0 29 24 0.003
total
Laterotrusion to the contralat- mm 9 7 10 n.a. 182 11 11 12 n.a. 24 <.001
eral side
Crepitus 0-­3 0 0 1 30 182 0 0 0 0 24 <.001
Laboratory findings
Blood
Erythrocyte sedimentation mm/h 20 10 35 46 112 4 3 6 0 12 .001
rate
C-­reactive protein mg/L 0 0 18 44 111 0 0 0 0 15 .092
Tumour necrosis factor pg/mL 14 11 20 67 85 7 6 11 0 14 .009
(plasma)
Interleukin-­1β (plasma) pg/mL 0 0 1.6 33 104 0 0 0 0 13 .431
Serotonin (serum) nmol/L 878 556 1218 15 84 700 626 932 0 15 .858
TMJ synovial fluid
Tumour necrosis factor pg/mL 0 0 71 47 109 0 0 0 20 20 .042
Tumour necrosis factor soluble pg/mL 731 149 1492 88 17 0 0 304 17 6 .049
receptor II
Interleukin-­1β pg/mL 0 0 0 19 186 0 0 0 0 22 .024
Interleukin-­1 receptor pg/mL 463 0 1366 72 80 0 0 0 0 20 <.001
antagonist
Interleukin-­1 soluble receptor II pg/mL 1580 668 3494 87 79 0 0 1522 21 18 .035
Serotonin nmol/L 37 0 3272 71 106 0 0 0 17 10 .016

n.a., not applicable; % pos, percentage of observations exceeding 0; n, number of observation; P, probability level for the difference between patients
and healthy individuals for each variable. Bold P-values denote significant differences.
ALSTERGREN et al. | 275

F I G U R E 1 Distribution of pain symptoms in a total of 219

TMJs with arthritis (n = 139) and without arthritis (n = 80)
as determined by synovial fluid sampling and analysis of
inflammatory mediators F I G U R E 2 Receiver operating characteristic (ROC) curve
showing the diagnostic sensitivity and 1-­specificity regarding
temporomandibular joint (TMJ) arthritis for the combinations of the
to a great extent. The sensitivity of Crepitus (as single variable) was
variables TMJ resting pain, Maximum opening pain intensity, Number
0.60, and the specificity was 0.93. The sensitivity of Contralateral lat- of jaw movements causing TMJ pain and Laterotrusive movement
erotrusion <8 mm was 0.64, and the specificity was 0.13. to the contralateral side. This combination significantly explained
the presence of TMJ arthritis with an area under the ROC curve
(AUC) = 0.72 (n = 190; P < .001)
3.6 | Predictive values
Table 5 shows the predictive values of single clinical variables and com- movements, TMJ pain on maximum mouth opening and Contralateral
binations of variables to identify arthritis in relation to the reference laterotrusion <8 mm. At the same time, the lowest negative likelihood
standard. Using combinations of variables, the highest positive predic- ratio (0.28) was found for not fulfilling any of the criteria TMJ rest-
tive value (0.79) was found for the combination of TMJ pain on maxi- ing pain, TMJ pain on mandibular movements, TMJ pain on maximum
mum mouth opening and Contralateral laterotrusion <8 mm (Table 5). The mouth opening or Contralateral laterotrusion <8 mm.
highest negative predictive value (0.67) was found for TMJ resting pain
or TMJ pain on mandibular movements or TMJ pain on maximum mouth
3.7 | Inflammatory activity
opening or Contralateral laterotrusion <8 mm. However, the negative
predictive value for TMJ resting pain alone was 0.66. Among the arthritic TMJs, TMJs with high inflammatory activity
The highest positive likelihood ratio (2.20) was found for the showed significantly higher TMJ pain on maximum mouth open-
combination of the variables TMJ resting pain, TMJ pain on mandibular ing (P < .001) and higher number of mandibular movements causing

TA B L E 4 Significant differences in univariate analysis between patients/temporomandibular joints (TMJ) with and without arthritis

Arthritis Not arthritis

Percentile Percentile

Variable Median 25th 75th n Median 25th 75th n P

Global pain intensity NRS 0-­10 5 3 7 76 2 0 4 54 <.001

TMJ pain
Resting pain intensity NRS 0-­10 4 2 7 120 1 0 5 71 .002
TMJ pain on maximum NRS 0-­10 2 1 5 132 0 0 1 76 <.001
mouth opening
Number of painful TMJ 0-­4 2 0 3 139 1 0 2 80 .009
movements
Laterotrusion, mm 9 6 10 131 10 8 12 65 <.001
contralateral

n, number of observation; P, probability level of the difference between TMJ arthritis in patients/TMJs for each variable; NRS, numerical rating scale
0-­10. Bold P-values denote significant differences.
276 | ALSTERGREN et al.

TMJ pain (P = .004) than TMJs with low inflammatory activity.

Patients with high TMJ inflammatory activity had higher plasma
levels of IL-­1β (P = .013) than patients with low TMJ inflammatory
activity.
There were not sufficient amounts of data in the non-­painful
TMJ arthritis group to statistically calculate a potential relation to
inflammatory activity.

3.8 | Diagnostic procedure

Table 6 and Figure 4 show the diagnostic criteria for possible, prob-
able and definite TMJ arthritis.

4 | D I S CU S S I O N

F I G U R E 3 Receiver operating characteristic (ROC) curve This study suggests diagnostic criteria for TMJ arthritis, as a base
showing the diagnostic sensitivity and 1-­specificity regarding for future research into clinical diagnostics of TMJ arthritis. Already
temporomandibular joint (TMJ) arthritis among joints with no pain today, there may very well be real clinical value in these criteria, but
for the combinations of the variables Crepitus and Contralateral at the same time the limitations of the presented data should be con-
laterotrusion <8 mm. This combination significantly explained
sidered. This study also identifies clinical variables indicating high
the presence of TMJ arthritis with an area under the ROC curve
TMJ inflammatory activity among arthritic TMJs.
(AUC) = 0.91 (n = 40; P < .001)

TA B L E 5 Sensitivity, specificity and

Diagnostic performance
predictive values for single variables and
Variable(s) Sensitivity Specificity PPV NPV combinations of variables for diagnosis of
temporomandibular joint arthritis
Single variables
TMJ resting pain 0.86 0.46 0.73 0.66
TMJ pain on maximum mouth opening 0.63 0.56 0.72 0.47
TMJ pain on maximum mouth opening >2/10 0.40 0.78 0.75 0.42
TMJ pain on mandibular movement 0.71 0.48 0.70 0.49
TMJ pain on mandibular movement (>1 movement) 0.60 0.56 0.70 0.45
TMJ pain on mandibular movement (>2 0.41 0.71 0.71 0.41
movements)
Contralateral laterotrusion <8 mm 0.37 0.83 0.79 0.43
Combinations of variables
TMJ resting pain OR 0.83 0.48 0.73 0.61
TMJ pain on mandibular movement
TMJ resting pain AND 0.55 0.61 0.71 0.44
TMJ pain on mandibular movement
TMJ pain on maximum mouth opening OR 0.75 0.50 0.72 0.53
contralateral laterotrusion <8 mm
TMJ pain on maximum mouth opening AND 0.24 0.89 0.79 0.40
contralateral laterotrusion <8 mm
TMJ resting pain OR 0.89 0.39 0.72 0.67
TMJ pain on mandibular movements OR
TMJ pain on maximum mouth opening OR
contralateral laterotrusion <8 mm
TMJ resting pain AND 0.22 0.90 0.79 0.40
TMJ pain on mandibular movements AND
TMJ pain on maximum mouth opening AND
contralateral laterotrusion <8 mm

PPV, positive predictive value; NPV, negative predictive value.

ALSTERGREN et al. | 277

TA B L E 6 Suggested diagnostic levels for TMJ arthritis

Diagnostic performance

Level Criteria Prevalence (%) Sensitivity Specificity PPV NPV

Possible TMJ pain on maximum mouth opening 71 0.63 0.56 0.72 0.47
50 0.63 0.56 0.59 0.60
Probable TMJ pain on maximum mouth opening 71 0.24 0.89 0.79 0.40
AND contralateral laterotrusion 50 0.24 0.89 0.69 0.54
<8 mm
Definite Pathological concentration of 1.00 1.00
inflammatory mediators in TMJ
synovial fluid

PPV, positive predictive value; NPV, negative predictive value.

assessment of the TMJ, indicating the need of a more thorough ex-

4.1 | Diagnostic sensitivity and specificity
amination and perhaps referral to a specialist.
Taken together, the joint-­related variables TMJ resting and maximum The higher diagnostic level of “Probable TMJ arthritis” combined
opening pain intensity, number of jaw movements causing TMJ pain and TMJ pain on maximum mouth opening with a Contralateral laterotrusion
laterotrusive movement to the contralateral side showed an acceptable <8 mm. Our results indicate that if both these criteria are fulfilled,
diagnostic performance when combined. In the post hoc analysis, the likelihood of a true TMJ arthritis is considerably higher than at
where sensitivity and specificity are inversely related, we adopted the level of “Possible TMJ arthritis.” Perhaps even more importantly:
the use of “Possible,” “Probable” and “Definite TMJ arthritis” lev- if the two criteria are not fulfilled, arthritis can be excluded in 89%
els in order to create a model for levels of diagnostic certainty and of the healthy TMJs.
accuracy. “Definite TMJ arthritis,” as suggested in the present study, in-
The inclusion criteria of this study comprised “Ongoing TMJ volves TMJ synovial fluid sampling and laboratory analyses of in-
pain, impaired TMJ function or recent development of anterior flammatory mediators in those samples. Although not particularly
open bite or event/disorder that may initiate or maintain TMJ ar- difficult to perform in itself, this procedure requires equipment,
thritis.” There was thus a selection of patients by anamnestic data procedures, experience and knowledge. It is also probably important
before the clinical data were applied to determine the diagnosis that the operator perform this procedure a number of times per year
“TMJ arthritis.” Accordingly, in our suggested criteria, we recom- in order to develop and retain the knowledge and skills required.
mend that these anamnestic data are assessed prior to applying That said, a diagnosis of “Definite TMJ arthritis” may be of great
the criteria in order to filter out cases highly unlikely to be af- clinical value in cases where there is a “silent” arthritis, that is an
flicted with TMJ arthritis. The prevalence of patients with arthritic TMJ arthritis without pain but with structural damage or growth dis-
TMJs that have anamnestic data of ongoing TMJ pain, impaired turbance that is not possible to detect by clinical examination. TMJ
TMJ function or recent development of anterior open bite or synovial fluid sampling may therefore have a great potential as a di-
event/disorder that may initiate or maintain TMJ arthritis should agnostic tool and should be considered in selected cases.
therefore not differ much between patients identified by these The sensitivity for the variable TMJ pain at rest was unexpect-
anamnestic criteria in the general population, in general dentistry edly high, especially as movement and loading pain is believed to
and in a specialist clinical with referred patients. The positive and be the most prominent pain aspects of arthritis. However, TMJ pain
negative predictive values are dependent on the prevalence of the at rest is most likely an unspecific clinical variable as it may include
investigated condition in the sample. We therefore calculated PPV pain not related to the joint proper more than movement pain, which
and NPV for the true prevalence in the sample (71%) as well as for can explain the high sensitivity but on the other hand it has a low
a prevalence of 50% representing a chance distribution. specificity. Certainly, arthritis may cause resting pain by activating
The lowest level of diagnostic certainty, “Possible TMJ arthritis,” sensitised nociceptive fibres in and around the joint or by central
was best indicated by TMJ pain on maximum opening. This clinical cri- mechanisms. The findings in the present study suggest that also
terion gave a sensitivity of 0.63 and a specificity of 0.56. Although minor pain intensity is of importance as the sensitivity increased
these values are not impressively high, on the group level this sin- when also minor pain intensity was included.
gle criterion will correctly identify a majority of the joints with ar- About 17% of the TMJs that were classified as arthritic did not
thritis as well as exclude arthritis in the majority of healthy joints. present any clinical pain findings. Chronic arthritis may very well be
Considering the suboptimal accuracy, as an only criterion, TMJ pain present without pain, although pain is common. 6,8-11 On the other
on maximum opening is not sufficient to diagnose TMJ arthritis with hand, 78% of the TMJs without arthritis showed pain at rest or on
a high enough level of certainty. The clinical description “Possible provocation by movement or palpation. This pain is most probably
TMJ arthritis” may therefore best be considered as a preliminary due to sensitisation, peripheral or central or a combination, of the
278 | ALSTERGREN et al.

destruction is a severe complication and may be very difficult to

treat. This supports the use of TMJ synovial fluid sampling and
subsequent analyses of inflammatory mediators for “Definite TMJ
arthritis” diagnosis in some cases, such as when occlusal changes
appear suddenly or are progressing rapidly, especially in the ab-
sence of pain.
The clinical examination in this study was not performed ex-
actly according to the DC/TMD protocol. DC/TMD was published
in 2014, and our examinations were performed between 1995 and
2009. The examination procedure used in the present study was
identical to most other studies on TMJ arthritis during that time.
Interestingly, most of our investigated variables were assessed in a
similar manner to the clinical examination in DC/TMD. For example,
maximum mouth opening capacity without assistance, the presence
of TMJ pain on mouth opening, laterotrusion and protrusion and
assessment of crepitus are almost identical. In future, our findings
should therefore be possible to use as a base in research into diag-
nostics of TMJ arthritis using the DC/TMD.
So far, no clinical diagnostic criteria for TMJ arthritis have
been established. In part, this may be due to a lack of an ade-
quate reference standard. In the present study, we used a refer-
ence standard based on TMJ synovial fluid concentrations of a
number of inflammatory mediators, for which there were either
known and published distinct differences between non-­a rthritic
and arthritic joints, or TMJ synovial fluid concentrations exceed-
F I G U R E 4 Proposed diagnostic classification for ing the 95th percentiles for healthy TMJs.7,10 Also, validated sam-
temporomandibular joint (TMJ) arthritis ple quality criteria were used to ensure that only high-­q uality
TMJ synovial fluid samples were included in the study. To our
articular or adjacent tissues of a non-­inflammatory nature. It may knowledge, this is the first time a laboratory test-­b ased reference
also be due to pain related to internal derangements, without an standard has been applied, for the TMJ or any other joint. Our
apparent inflammatory component. This is one likely explanation reference standard was based on determination of the true sy-
to why pain on palpation and, especially pressure-­pain thresholds, novial fluid concentration of certain mediators using washing of
over the TMJ, has been primarily related to systemic factors and the joint with the vitamin B12-­m ethod developed and published
not to local inflammatory factors.10 The frequent finding of pain by us.7,26,27 Although this reference standard may have its lim-
despite the absence of local inflammation of the joint suggests that itations in assessing arthritis, we consider it to at least as accu-
clinical criteria of TMJ arthritis should not be limited to pain vari- rate as any other available reference standard including magnetic
ables, but should also assess function. In the present study, the de- resonance imaging.9 Another previously adopted reference stan-
gree of anterior open bite and range of movements were included dard is expert opinion, based partly on clinical findings, which is
as primarily non-­pain-­related variables in an attempt to investigate problematic in a scientific context because of the circularity in
the diagnostic value of these functional variables. Indeed, impaired argument and risk of overestimation of accuracy. 29 A biomarker
range of movements is usually related to joint pain but there may test, such as synovial fluid sampling with standardised laboratory
also be other mechanical factors behind restricted movement in ar- analysis, has the advantage of being independent of the clinical
thritis, for example fibrous adhesions or internal derangements.17 findings, an important methodological requirement in studies of
Anterior open bite, especially progressive anterior open bite, is a diagnostic accuracy. 30
clinical sign of ongoing or severe/rapid TMJ cartilage and bone tis- In the non-­painful joints, crepitus and impaired contralateral lat-
sue destruction and not uncommon among patients with more se- erotrusion explained non-­painful TMJ arthritis to a great extent. This
vere forms of arthritis, for example RA.7,11,24,25 Anterior open bite condition, which is fairly common according to this study and diffi-
is usually a late sign but it may still be the first sign of TMJ arthritis cult to identify clinically, warrants more attention.
if pain is absent. In the present study, anterior open bite was used
as a crude marker for TMJ cartilage and bone tissue destruction.
4.2 | Inflammatory activity
It has obvious limitations, and anterior open bite may certainly
be caused by other conditions than arthritis. Unfortunately, an Characteristic for TMJs with high inflammatory activity was
anterior open bite due to bilateral TMJ cartilage and bone tissue TMJ pain on mandibular movements. If a TMJ fulfils the clinical
ALSTERGREN et al. | 279

diagnostic criteria for arthritis suggested above, the presence of for RA where early diagnosis was in focus. 31 However, early diag-
TMJ pain on mandibular movements may thus indicate a higher in- nosis could not be considered in the present study. Our findings
flammatory activity. TMJ pain on jaw movements has been found must therefore be considered as a major first step towards accu-
to be strongly related to an inflammatory intra-­a rticular milieu in rate and comprehensive diagnostic criteria that need to be tested
other studies.10,18-20 TMJ pain on jaw movement thus seems to be in future studies.
of relevance as a clinical sign of TMJ arthritis. It may be of im- The included patients had systemic inflammatory diseases (RA,
portance for a clinical diagnosis and subsequently for choice of PsA, ankylosing spondylitis etc.) or local TMJ arthritis. The purpose
therapy and for monitoring therapy effectivity. However, there is was to include the whole spectrum of systemic and local TMJ con-
a need for studies that focus on this aspect before we will know if ditions. No patients had been treated with intra-­articular corticoids
assessment of inflammatory activity is possible, relevant and clini- for at least three months before sampling. Because the inflammatory
cally useful. process is unspecific, the factors involved in the inflammation are
likely to be similar and unlikely to vary a lot depending on the type of
underlying disorder. However, for some of the rheumatic disorders,
4.3 | Methodological considerations
the presence of autoantibodies like anti-­citrullinated peptide anti-
This study proposes a diagnostic model for TMJ arthritis by the use bodies (ACPA) and rheumatoid factor (RF) is common and strongly
of certainty levels possible, probable and definite TMJ arthritis. The contribute to immune system activation and thereby increased in-
model is based on clinical findings (pain-­related variables and func- flammatory activity.
tional variables). Single variables and combinations of variables were The control group of healthy individuals may seem small when
compared to the presence of the inflammatory mediators in the TMJ considering the number of subjects. However, sampling from
synovial fluid. These mediators have previously been shown to be healthy TMJs may not be allowed nowadays from an ethical point
related to both TMJ pain and tissue destruction in patients with sys- of view, so this control group can be considered as unique compared
20-25,27
temic inflammatory joint diseases. to other studies, and contributes with valuable data. To clarify, all
In the present study, the prevalence of TMJ arthritis among the samples from healthy individuals and patients were obtained with
included patients can be expected to be higher than in the popula- ethical approval.
tion as the patients were referred to a specialist clinic due to orofacial In our material, one could argue that there were actually three
pain and jaw dysfunction, and many patients were in fact referred groups: (i) healthy (no arthritis), (ii) patients with TMJ arthritis and (iii)
from rheumatologists. The reference standard tells us that 65% of patients with conditions that may cause TMJ arthritis but where the
the TMJs in the sample were arthritic according to the presence of synovial fluid sample could not confirm active arthritis (ie, a patient
inflammatory mediators in the synovial fluid. The prevalence in the control group without TMJ arthritis). The inclusion of patients with-
general population according to our proposed criteria is unknown, out active TMJ arthritis for comparison with the patients with active
using the criteria in this study, but can be expected to be less than TMJ arthritis is very important, and especially relevant as the clini-
65%. This means that the positive predictive values in the population cal presentation of chronic inflammation at a given time-­point may
are likely lower. be anywhere on a continuum from no sign or symptom whatsoever
Considering the complex nature of chronic TMJ arthritis, to any combination of pain, swelling/exudate, tissue degradation or
which may or may not include pain, functional limitations, carti- growth disturbance. Just like patients with active TMJ arthritis rep-
lage and bone tissue destruction and growth inhibition (in adoles- resent one extreme of the continuum, the healthy individuals repre-
cents), the variables included in this study cannot fully encompass sent the other extreme and therefore contribute to understanding
all the above aspects of chronic TMJ arthritis. In the present of the whole spectrum. Because of the relatively small group size,
study, only degree of anterior open bite was included as an esti- our conclusions regarding findings in healthy individuals remain very
mation of TMJ cartilage and bone tissue destruction. It is probably conservative.
a coarse and late sign of TMJ bone tissue destruction. There may,
in addition, be anamnestic information of importance for diagno-
sis, for example recent occlusal changes, TMJ pain on chewing, 5 | CO N C LU S I O N
TMJ pain with after sensation, etc. These were not investigated
in this study. Chronic TMJ arthritis may not always include pain. This study proposed clinical diagnostic criteria for TMJ arthritis, as
In those cases, the inflammation may very likely be a silent type a base for future research into clinical diagnostics of TMJ arthri-
of inflammation, causing cartilage and bone tissue destruction or, tis. The clinical variables TMJ pain on maximum mouth opening and
in adolescents, growth disturbances as well but with no or very Contralateral laterotrusion of less than 8 mm appear to have high clini-
low intensity pain. In the present study, almost one-­fifth of the cal diagnostic value. Already today, there may very well be clinical
TMJs with arthritis were pain-­free, indicating that pain-­free ar- value in these criteria but at the same time the limitations in knowl-
thritis is not uncommon. In addition, diagnosis of early arthritis is edge should be considered.
important to improve treatment prognosis. In 2010, the American The study also identified clinical variables indicating high TMJ
College of Rheumatology (ACR) published new diagnostic criteria inflammatory activity among arthritic TMJs.
280 | ALSTERGREN et al.

AC K N OW L E D G M E N T S 12. Arnett FC, Edworthy SM, Bloch DA, et al. The American Rheumatism
Association 1987 revised criteria for the classification of rheuma-
The authors wish to thank the laboratory technicians Karin Trollsås, toid arthritis. Arthritis Rheum. 1988;31:315‐324.
Agneta Gustafsson and Kari Eriksson for their tremendous work, as 13. Tillett W, Costa L, Jadon D, et al. The classification for psoriatic ar-
thritis (CASPAR) criteria–a retrospective feasibility, sensitivity, and
well as the healthy individuals that participated. This research was
specificity study. J Rheumatol. 2012;39:154‐156.
funded by the Swedish Council of Medical Research, the National
14. Raychaudhuri SP, Deodhar A. The classification and diagnostic cri-
Institute of Dental and Craniofacial Research (NIDCR; R01-­DE15420), teria of ankylosing spondylitis. J Autoimmun. 2014;48–49:128‐133.
the Schering-­Plough Corporation, and the Swedish Dental Society. 15. Anonymous American College of Rheumatology [Web Page].
Dr. Alstergren reports grants from Svenska Tandläkare-­Sällskapet, Available from: http://www.rheumatology.org. Accessed November
5, 2017.
grants from Schering-­Plough, grants from NICDR, during the con-
16. Cohen JF, Korevaar DA, Gatsonis CA, et al. STARD for Abstracts:
duct of the study. Dr. Kopp reports grants from Svenska Tandläkare-­ essential items for reporting diagnostic accuracy studies in journal
Sällskapet, grants from Schering-­Plough, grants from NICDR, during or conference abstracts. BMJ. 2017;358:j3751.
the conduct of the study. The other author has stated explicitly that 17. Alstergren P, Ernberg M, Kvarnstrom M, Kopp S. Interleukin-­1beta
in synovial fluid from the arthritic temporomandibular joint and its
there are no conflicts of interest in connection with this article.
relation to pain, mobility, and anterior open bite. J Oral Maxillofac
Surg. 1998;56:1059‐1065; discussion 1066.
18. Alstergren P, Kopp S. Pain and synovial fluid concentration of sero-
ORCID tonin in arthritic temporomandibular joints. Pain. 1997;72:137‐143.
19. Alstergren P, Ernberg M, Kopp S, Lundeberg T, Theodorsson E.
P. Alstergren http://orcid.org/0000-0002-8539-7742 TMJ pain in relation to circulating neuropeptide Y, serotonin, and
M. Pigg http://orcid.org/0000-0002-7989-1541 interleukin-­1 beta in rheumatoid arthritis. J Orofac Pain. 1999;13:
49‐55.
20. Nordahl S, Alstergren P, Kopp S. Tumor necrosis factor-­alpha in sy-
novial fluid and plasma from patients with chronic connective tis-
REFERENCES sue disease and its relation to temporomandibular joint pain. J Oral
1. Dias IM, Coelho PR, Picorelli Assis NM, Pereira Leite FP, Devito Maxillofac Surg. 2000;58:525‐530.
KL. Evaluation of the correlation between disc displacements and 21. Nordahl S, Alstergren P, Eliasson S, Kopp S. Radiographic signs of
degenerative bone changes of the temporomandibular joint by bone destruction in the arthritic temporomandibular joint with spe-
means of magnetic resonance images. Int J Oral Maxillofac Surg. cial reference to markers of disease activity. A longitudinal study.
2012;41:1051‐1057. Rheumatology (Oxford). 2001;40:691‐694.
2. Dias IM, Cordeiro PC, Devito KL, Tavares ML, Leite IC, Tesch Rde S. 22. Voog U, Alstergren P, Eliasson S, Leibur E, Kallikorm R, Kopp S.
Evaluation of temporomandibular joint disc displacement as a risk Inflammatory mediators and radiographic changes in temporoman-
factor for osteoarthrosis. Int J Oral Maxillofac Surg. 2016;45:313‐317. dibular joints of patients with rheumatoid arthritis. Acta Odontol
3. Stabrun AE, Larheim TA, Hoyeraal HM. Temporomandibular joint Scand. 2003;61:57‐64.
involvement in juvenile rheumatoid arthritis. Clinical diagnostic cri- 23. Alstergren P, Benavente C, Kopp S. Interleukin-­1beta, interleukin-­1
teria. Scand J Rheumatol. 1989;18:197‐204. receptor antagonist, and interleukin-­1 soluble receptor II in tem-
4. Olson L, Eckerdal O, Hallonsten AL, Helkimo M, Koch G, Gare BA. poromandibular joint synovial fluid from patients with chronic poly-
Craniomandibular function in juvenile chronic arthritis. A clinical arthritides. J Oral Maxillofac Surg. 2003;61:1171‐1178.
and radiographic study. Swed Dent J. 1991;15:71‐83. 24. Voog U, Alstergren P, Eliasson S, Leibur E, Kallikorm R, Kopp S.
5. Kjellberg H, Fasth A, Kiliaridis S, Wenneberg B, Thilander B. Progression of radiographic changes in the temporomandibular
Craniofacial structure in children with juvenile chronic arthritis joints of patients with rheumatoid arthritis in relation to inflam-
(JCA) compared with healthy children with ideal or postnormal oc- matory markers and mediators in the blood. Acta Odontol Scand.
clusion. Am J Orthod Dentofacial Orthop. 1995;107:67‐78. 2004;62:7‐13.
6. Bakke M, Zak M, Jensen BL, Pedersen FK, Kreiborg S. Orofacial 25. Alstergren P, Kopp S. Insufficient endogenous control of tumor
pain, jaw function, and temporomandibular disorders in women necrosis factor-­ alpha contributes to temporomandibular joint
with a history of juvenile chronic arthritis or persistent juvenile pain and tissue destruction in rheumatoid arthritis. J Rheumatol.
chronic arthritis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2006;33:1734‐1739.
2001;92:406‐414. 26. Alstergren P, Appelgren A, Appelgren B, Kopp S, Lundeberg T,
7. Alstergren P, Kopp S, Theodorsson E. Synovial fluid sampling from Theodorsson E. Determination of temporomandibular joint fluid
the temporomandibular joint: sample quality criteria and levels of in- concentrations using vitamin B12 as an internal standard. Eur J Oral
terleukin-­1 beta and serotonin. Acta Odontol Scand. 1999;57:16‐22. Sci. 1995;103:214‐218.
8. Doria A, Putterman C, Sarzi-Puttini P, Szekanecz Z, Shoenfeld Y. 27. Alstergren P, Appelgren A, Appelgren B, Kopp S, Nordahl S,
Controversies in rheumatism and autoimmunity. Autoimmun Rev. Theodorsson E. Measurement of joint aspirate dilution by a spec-
2012;11:555‐557. trophotometer capillary tube system. Scand J Clin Lab Invest.
9. Peck CC, Goulet JP, Lobbezoo F, et al. Expanding the taxonomy 1996;56:415‐420.
of the diagnostic criteria for temporomandibular disorders. J Oral 28. Hajian-Tilaki K. Receiver operating characteristic (ROC) curve anal-
Rehabil. 2014;41:2‐23. ysis for medical diagnostic test evaluation Caspian. J Intern Med.
10. Alstergren P, Fredriksson L, Kopp S. Temporomandibular joint pres- 2013;4:627‐635.
sure pain threshold is systemically modulated in rheumatoid arthri- 29. Lobbezoo F, Visscher CM, Naeije M. Some remarks on the RDC/
tis. J Orofac Pain. 2008;22:231‐238. TMD Validation Project: report of an IADR/Toronto-­2008 work-
11. Tegelberg A, Kopp S. Clinical findings in the stomatognathic system shop discussion. J Oral Rehabil. 2010;37:779‐783.
for individuals with rheumatoid arthritis and osteoarthrosis. Acta 30. Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J. The de-
Odontol Scand. 1987;45:65‐75. velopment of QUADAS: a tool for the quality assessment of studies
ALSTERGREN et al. | 281

of diagnostic accuracy included in systematic reviews. BMC Med Res

Methodol. 2003;3:25. How to cite this article: Alstergren P, Pigg M, Kopp S. Clinical
31. Aletaha D, Neogi T, Silman AJ, et al. Rheumatoid arthritis classi- diagnosis of temporomandibular joint arthritis. J Oral Rehabil.
fication criteria: an American College of Rheumatology/European
2018;45:269–281. https://doi.org/10.1111/joor.12611
League against rheumatism collaborative initiative. Arthritis Rheum.
2010;2010:2569‐2581.