Posted at the Institutional Resources for Unique Collection and Academic Archives at Tokyo Dental College,

Available from http://ir.tdc.ac.jp/

Title

Potential of fluid-attenuated inversion recovery

(FLAIR) in identification of temporomandibular

joint effusion compared with T2-weighted images.

Author(s)

Alternative

Imoto, K; Otonari-Yamamoto, M; Nishikawa, K; Sano,

T; Yamamoto, A

JournalOral surgery, oral medicine, oral pathology, oral

radiology, and endodontics, 112(2): 243-248

URL http://hdl.handle.net/10130/2532

Right

1

Potential of FLAIR in identification of temporomandibular joint effusion in

comparison with T2-weighted images

Kenichi Imoto,DDS, a,b Mika Otonari-Yamamoto, DDS, PhD,b

Keiichi Nishikawa, PhD,b Tsukasa Sano, DDS, PhD,b

Aya Yamamoto, DDS, PhD,b

a Oral Health Science Center HRC7,

Tokyo Dental College

b Department of Oral and Maxillofacial Radiology,

Tokyo Dental College

Address for Correspondence:

Kenichi Imoto, D.D.S.

Department of Oral and Maxillofacial Radiology

Tokyo Dental College

1-2-2, Masago, Mihama-ku, Chiba, 261-8502 Japan

Business Phone: 81-(0)43-270-3961

Home Phone: 81-(0)43-211-0167

Fax: 81-(0)43-270-3963

E-mail: imotokenichi@tdc.ac.jp

2

Acknowledgements

This research was supported by Oral Health Science Center Grant HRC7 from Tokyo

Dental College, and by a “High-Tech Research Center” Project for Private Universities:

matching fund subsidy from MEXT (Ministry of Education, Culture, Sports, Science

and Technology) of Japan, 2006-2010.

We would like to thank Associate Professor Jeremy Williams, Tokyo Dental College,

for his assistance with the English of this manuscript.

1

Abstract

Objective. The purpose of this study was to determine the potential of fluid attenuated

inversion recovery (FLAIR) sequence images in the identification of joint effusion (JE)

in comparison with T2-weighted images.

Study design. A total of 31 joints (28 patients) with JE were investigated by magnetic

resonance imaging. Regions of interest were placed over JE, cerebrospinal fluid (CSF)

and gray matter (GM) on T2-weighted and FLAIR images and their signal intensities

compared. The signal intensity ratios (SIRs) of JE and CSF were calculated used GM as

the reference point. The Pearson product-moment correlation coefficient was used for

the statistical analysis.

Result. The SIR of JE showed a strong correlation between T2-weighted and FLAIR

images. However, no correlation was observed for CSF. The average suppression ratio

for JE was lower than that for CSF.

Conclusion. MRI using FLAIR sequences revealed that JE was not only water content,

but a fluid accumulation containing elements, such as protein. Further studies are

needed and FLAIR sequences could be useful for the diagnosis of pain and symptoms

of the TMJ.

2

Magnetic resonance imaging (MRI) can be used for visualization of the

temporomandibular joint (TMJ). Earlier studies have used it to investigate disk

morphology and position,1 osseous change in the condyle2 and joint effusion (JE).3,4

High signal intensity from the upper and lower joint spaces on T2-weighted images is

considered to indicate JE.4-8 Joint effusion in the TMJ on MRI was first investigated by

Harms et al.5, followed by a number of subsequent studies aimed at clarifying its

association with TMJ pain and symptoms6 and intra-articular pathosis.7 While some

studies have suggested a correlation between JE and TMJ pain,8-10 others have found no

such correlation.11,12 From a biochemical perspective, it has been suggested that JE is

related to inflammatory changes seen in patients with disorders of the TMJ.13

Furthermore, microscopic investigation revealed that JE may reflect synovitis.14,15

Gynther et al.16 found that the inflammatory changes brought about by synovitis in the

TMJ induced hyperplasia of the synovial cell layer. Segami et al.17 suggested that there

was a strong correlation between hyperplastic synovial tissue and amount of JE.

Inflammation of synovial tissue may result in an increase in its signal intensity to a

level almost equal to that obtained with JE on T2-weighted images. It is possible that JE

indicates inflammation of the synovial membrane tissue itself. Although much research

has been carried out on JE, its detection by MRI remains problematic, and identification

by T2-weighted images alone is difficult. In this study, we investigated the potential of

fluid attenuated inversion recovery (FLAIR) sequence images to demonstrate JE in

comparison with T2-weighted images. Employing T2-weighted images, FLAIR is often

used to suppress cerebrospinal fluid (CSF), allowing differentiation between CSF and

3

brain lesions.18 We hypothesized that the signal intensity of JE would be lower on

FLAIR images, provided that JE comprised mainly fluid, as does CSF. A signal

intensity on FLAIR may be depending on amount of protein in the JE. Our another

hypothesis is that more TMJ pain a patient has, more protein JE has. As a result, a TMJ

with more pain would show increased signal on FLAIR images. Based on this

hypothesis, we compared the signal intensities of JE obtained from T2-weighted and

FLAIR images and investigated the relationship between them retrospectively.

4

Materials and methods

Magnetic resonance images were obtained from patients referred to our department at

Tokyo Dental College Chiba Hospital in 2004, 2005 and 2008. Informed consent was

obtained from all patients and the study protocol was approved by the Ethical Review

Board of our institution (No.139). This study was based on 232 joints of 116 patients.

According to the inclusion and exclusion criteria (Table I), 201 joints were excluded

from this study. A total of 31 joints of 28 patients were studied. A total of 31 joints of 28

patients were studied. There were 4 males and 24 females. The mean age of the patients

was 39.9 years with an age range of 13 to 76 years.

All images were obtained with the 1.5 Tesla MR Imager (Magnetom Symphony,

Siemens, Erlangen, Germany), using a double loop array coil. Using a fast-spin echo

sequence, T2-weighted and FLAIR sagittal images were obtained at the closed-mouth

position (Figure 1). Table II shows the parameters employed for both sets of images .

Echo times (TE) of 122 ms and 168 ms were used for the FLAIR images. During our

study, TE was changed to TE 168 ms to improve attempt the contrast as a result of the

examination of FLAIR images with TE 122 ms images obtained with 2004-05. Images

were obtained with a TE of 122 and 168 ms in 15 and 13 patients, respectively.

Amount of TMJ fluid on the T2-weighted images was determined by the method of

Larheim et al.19 Accordingly, fluid content in the joints was categorized as moderate,

marked, or extensive.

Images were saved as DICOM files. Signal intensity was determined using the public

domain image analysis software ImageJ 1.37j (NIH, USA). Measurements of signal

intensity were performed on both the T2-weighted and FLAIR images by one of the

5

authors. The decision as to which images were suitable for ROI selection for each area

by two sequences was made by a consensus of two oral and maxillofacial radiologists.

The signal intensity of gray matter (GM) was taken as the reference point. Regions of

interest (ROI) were placed over the GM, CSF and JE. The size of the ROI and its

position on the CSF and GM were determined according to the method of Yajima et al.

20 A 6.25 mm2 for TE of 122ms or a 6.86 mm2 for TE of 168 ms as ROI were defined. 20

The maximal area of JE on the T2-weighted images was defined as the ROI for JE

(Figure2). An average area were 3.45 mm2 (ranging from 1.29 – 9.09 mm2 ) for TE of

122ms and 7.15 mm2 (ranging from 1.80 – 18.45 mm2) for TE of 168ms. A 0.25 mm2

for TE of 122ms or a 0.34 mm2 for TE of 168 ms as ROI were defined and placed over

CSF. They were placed closest to the mandibular condyle on a line perpendicular to the

top of the condyle.20

The signal intensity ratio (SIR)s of JE and CSF were calculated as follows: SIR =

the signal intensity of JE or CSF/the signal intensity of GM. In the FLAIR sequence

images, the patients were divided into Group A (TE, 122 ms) and Group B (TE, 168

ms) for the statistical analysis. The Pearson product-moment correlation coefficient was

used to assess the correlation between the SIRs on the T2-weighted and FLAIR images

of both JE and CSF. A probability of less than 0.05 was considered statistically

significant. The suppression ratios of the signal intensities in FLAIR imaging were also

compared according to the following equation: (SIR of T2-weighted images­SIR of

FLAIR images)/SIR of T2-weighted images × 100 (Figure 3).

6

Results

The SIRs obtained for JE and CSF on T2-weighted and FLAIR images are shown in

Table III. The average SIR of JE was higher than that of CSF on both T2-weighted and

FLAIR images in both Groups A and B.

A strong correlation between T2-weighted images and FLAIR images was obtained

for JE in both groups. The correlation coefficient was 0.81 in Group A and 0.94 in

Group B (Figures 4 and 5). On the other hand, no correlation was found between the

two sets of images for CSF (0.09 in Group A、-0.33 in Group B).

The average suppression ratio for the signal intensity of JE by FLAIR imaging was

36.9% in Group A and 16.3% in Group B. The ratio for CSF was 75.6% in Group A and

71.7% in Group B (Table IV). The signal intensity of CSF was suppressed more

markedly than that of JE.

7

Discussion

Joint effusion is considered to be a pathological collection of joint fluid that can only

be observed on T2-weighted images. Hitherto, the only way to determine the presence

of JE has been by amount of fluid present in the joint. We hypothesized that the FLAIR

technique would provide additional information for clarification of JE.

In this study, a strong correlation was found between T2-weighted images and

FLAIR images in both groups for JE. On the other hand, no such correlation was found

for CSF.

A reduction in signal intensity was observed for JE on FLAIR sequence images.

However, the average suppression ratio for JE was lower than that for CSF. If JE

comprised liquid components similar to those of CSF, the suppression ratio for JE

would not have differed from that for CSF. This indicates that the liquid content of JE

differs somewhat to that of CSF, and that some element in JE induces a shortened T1

relaxation time in MRI.

Joint effusion is believed to represent not only exudation from inflamed tissue, but

also hypertrophic synovium in the joint.21-23 Hypertrophic synovium is a characteristic

of synovitis seen in the patients with internal derangement.16 Synovitis has been

observed arthroscopically and studied to clarify histological changes in disorders of the

TMJ.24,25 Synovial hyperplasia, vasucularity and an increase in inflammatory cells are

all characteristic changes brought about by synovitis.14,15 In an arthroscopic study,

Segami et al.17 investigated whether JE indicated the presence of synovitis by

comparing amounts of JE and degree of synovitis. They found that joints with effusion

had more pronounced synovitis than joints with no effusion, and, moreover, amount of

8

JE increased according to severity of synovitis. It is possible that JE indicates

morphological changes brought about by synovitis such as hyperplastic synovium, as

the signal intensity for such inflammatory synovial tissues would not be suppressed,

unlike that for CSF. However, synovitis was predominantly observed in the posterior

disk attachment in arthroscopic studies.16, 26-28 Joint effusion is usually located in the

anterior recess of the upper joint space. In this study, too, the maximal area of JE was

always observed in the anterior recess of the upper joint space, and this is where the

ROI was placed. To our knowledge, this is the first study to report that hyperplastic

synovium was observed predominantly in the anterior recess of the upper joint space.

These results strongly suggest that JE indicates morphological change in synovial

tissue.

There are several different contents in JE, such as protein or polysaccaride.29 In a

study on the relationship between joint effusion, joint pain and protein levels in joint

lavage fluid obtained from patients with internal derangement or osteoarthritis of the

TMJ, Takahashi et al.7 found that painful joints were more likely to show joint effusion

on MRI and that protein levels in joint lavage fluid recovered from those joints was

higher than that obtained from pain-free joints. In another study on the relationship

between amount of JE and levels of total protein and proinflammatory cytokines in

synovial fluid from patients with internal derangement or osteoarthrosis of the TMJ,

Segami et al.6 found that synovial fluid in TMJs with internal derangement or

osteoarthrosis and JE contained higher concentrations of total protein and

proinflammatory cytokines IL-6 and IL-8 than did synovial fluid in TMJs without JE.

These two earlier studies indicated that JE contains a high level of protein derived

9

from inflammatory synovial tissues. Protein can induce a shortening of T1 relaxation

time.30 This suggests that the average suppression ratio of signal intensity for JE was

lower than that for CSF in this study because higher amounts of proteins were present

in the JE, which would be consistent with the results of these earlier studies.

If JE contains secondary products produced by inflamed synovium, the signal

intensity of joint fluid in asymptomatic TMJs and its suppression ratio on FLAIR

images would differ from those for JE. Further study investigating joint fluid from

volunteers with no TMJ symptoms is needed to clarify this point, however.

Takahashi et al.7 found that mean protein concentration in pain-free joints with JE

was higher than that in pain-free joints without JE, although the difference was not

statistically significant. On the other hand, mean protein concentration in painful,

JE-free joints was significantly higher than that in JE-free joints that were not painful.

Therefore, they suggested that elevated protein levels may be correlated with joint pain

rather than JE. Joint pain, a representative symptom of TMJ disorders, may be

correlated with signal intensity in joint fluid, including JE, on FLAIR images. Further

study is needed to clarify the clinical significance of JE in joint pain, however.

Although the average suppression ratio of signal intensity for JE by FLAIR imaging

with a TE of 122 ms was higher than that with a TE of 168 ms in this study, this may be

considered to be within the margin of error, as the standard deviation of the average

suppression ratio of signal intensity for JE was relatively high (Table IV).

In conclusion, MRI using FLAIR sequences revealed that JE was not only water

content, but a fluid accumulation containing protein elements. However, we do not

know whether this fluid accumulation containing protein elements can be specialized in

10

only JE yet. The further study, which investigates the content of small amount of joint

fluid using FLAIR image is needed. Moreover, the correlation between the suppression

ratio of signal intensity for JE by FLAIR imaging and clinical symptom should be

analyzed. Depending on further studies, FLAIR sequences could be useful for the

diagnosis of pain and symptoms of the TMJ.

11

References

1. Katzberg RW, Schenck J, Roberts D, Tallents RH, Manzione JV, Hart HR, Foster TH,

Wayne WS, Bessette RW. Magnetic resonance imaging of the temporomandibular joint

meniscus. Oral Surg Oral Med Oral Pathol 1985;59:332-5.

2. Kirk WS Jr. A comparative study of axial corrected tomography with magnetic

resonance imagery in 35 joints. Oral Surg Oral Med Oral Pathol 1989;68:646-52.

3. Schellhas KP, Wilkes CH, Omlie MR, Peterson CM, Johnson SD, Keck RJ, Block JC,

Fritts HM, Heithoff KB. The diagnosis of temporomandibular joint disease:

two-compartment arthrography and MR. AJR Am J Roentgenol 1988;151:341-50.

4. Westesson PL, Brooks SL. Temporomandibular joint: relationship between MR

evidence of effusion and the presence of pain and disk displacement. Am J Roentgenol

1992;159:559-63.

5. Harms SE, Wilk RM, Wolford LM, Chiles DG, Milam SB. The temporomandibular

joint: magnetic resonance imaging using surface coils. Radiology 1985;157:133-6.

6. Segami N, Miyamaru M, Nishimura M, Suzuki T, Kaneyama K, Murakami K. Does

joint effusion on T2 magnetic resonance images reflect synovitis? Part 2.Comparison of

concentration levels of proinflammatory cytokines and total protein in synovial fluid of

the temporomandibular joint with internal derangements and osteoarthrosis. Oral Surg

Oral Med Oral Pathol Oral Radiol Endod 2002;94:515-21.

7. Takahashi T, Nagai H, Seki H, Fukuda M. Relationship between joint effusion, joint

pain, and protein levels in joint lavage fluid of patients with internal derangement and

osteoarthritis of the temporomandibular joint. J Oral Maxillofac Surg 1999;57:1187-93.

8. Schellhas KP, Wilkes CH. Temporomandibular joint inflammation: comparison of

12

MR fast scanning with T1- and T2-weighted imaging techniques. Am J Roentgenol

1989;153:93-8.

9. Larheim TA, Westesson PL, Sano T. MR grading of temporomandibular joint fluid:

association with disk displacement categories, condyle marrow abnormalities and pain.

Int J Oral Maxillofac Surg 2001;30:104-12.

10. Yamamoto M, Sano T, Okano T. Magnetic resonance evidence of joint fluid with

temporomandibular joint disorders. J Comput Assist Tomogr 2003;27:694-8.

11. Murakami K, Nishida M, Bessho K, Iizuka T, Tsuda Y, Konishi J. MRI evidence of

high signal intensity and temporomandibular arthralgia and relating pain. Does the high

signal correlate to the pain? Br J Oral Maxillofac Surg 1996;34:220-4.

12. Adame CG, Monje F, Offnoz M, Martin-Granizo R. Effusion in magnetic resonance

imaging of the temporomandibular joint: a study of 123 joints. J Oral Maxillofac Surg

1998;56:314-8.

13. Seki H, Fukuda M, Iino M, Takahashi T, Yoshioka N. Immunohistochemical

localization of cyclooxygenase-1 and -2 in synovial tissues from patients with internal

derangement or osteoarthritis of the temporomandibular joint. Int J Oral Maxillofac

Surg. 2004:33:687-92.

14. Gynther GW, Dijkgraaf LC, Reinbolt FP, Holmlund AB, Liem RS, de Bont LG.:

Synovial inflammation in arthroscopically obtained biopsy specimens from the

temporomandibular joint: A review of the literature and a proposed histologic grading

system. J Oral Maxillofac Surg 1998;56:1281-6.

15. Segami N, Nishimura M, Kaneyama K, Miyamaru M, Sato J, Murakami K. Does

joint effusion on T2 magnetic resonance images reflect synovitis? Comparison of

13

arthroscopic findings in internal derangements of the temporomandibular joint. Oral

Surg Oral Med Oral Pathol Oral Radiol Endod 2001;92:341-5.

16. Gynther GW, Holmlund AB, Reinholt FP. Synovitis in internal derangement of the

temporomandibular joint: correlation between arthroscopic and histologic findings. J

Oral Maxillofac Surg 1994;52:913-8.

17. Segami N, Suzuki T, Sato J, Miyamaru M, Nishimura M, Yoshimura H. Does joint

effusion on T2 magnetic resonance images reflect synovitis? Part 3. Comparison of

histologic findings of arthroscopically obtained synovium in internal derangements of

the temporomandibular joint. Oral Surg Oral Med Oral Pathol Oral Radiol Endod

2003;95:761-6.

18. Coene BD, Hajnal JV, Pennock JM, Bydder GM. MRI of the brain stem using fluid

attenuated inversion recovery pulse sequences. Neuroradiology 1993;35:327-331.

19. Larheim TA, Westesson PL, Sano T. MR grading of temporomandibular

joint fluid: association with disk displacement categories, condyle marrow

abnormalities and pain. Int J Oral Maxillofac Surg 2001;30:104-12.

20. Yajima A, Sano T, Otonari-Yamamoto M, Otonari T, Ohkubo M, Harada T, Wakoh

M. MR evidence of characteristics in symptomatic osteoarthritis of the

temporomandibular joint: increased signal intensity ratio on proton density-weighted

images of bone marrow in the mandibular condyle. Cranio 2007;25:250-6.

21. Smith HJ, Larheim TA, Aspestrand F. Rheumatic and nonrheumatic disease in the

the temporomandibular joint: Gadolinium-enhanced MR imaging. Radiology

1992;185:229-34.

22. Ostergaard M, Stoltenberg M, Lovgreen-NielsenP, Volck B, Jensen CJ, Lorenzen I.

14

Magnetic resonance imaging-determined synovial membrane and joint effusion

volumes in rheumatoid arthritis and osteoarthritis: comparison with the macroscopic

anc microscopic appearance of the synovium. Arthritis Rheum 1997;40:1856-67.

23. Beltran J, Noto AM, Hrman LJ, Mosure JC, Burk JM, Christoforidis AJ. Joint

effusions: MR imaging. Radiology 1986;158:133-7.

24. Merrill RG, Yih WY, Langan MJ. A histologic evaluation of the accuracy of TMJ

diagnostic arthroscopy. Oral Surg Oral Med Oral Pathol Oral Radiol Endod

1990;70:393-8.

25. Holmlund AB, Gynther GW, Reinholt FP. Disk derangement and inflammatory

changes in the posterior disk attachment of the temporomandibular joint: a histologic

study. Oral Surg Oral Med Oral Pathol 1992;73:9-12.

26. Murakami K, Segami N, Okamoto M, Yamamura I, Takahashi K, Tsuboi Y.

Outcome of arthroscopic surgery for internal derangement of the

temporomandibular joint: long-term results covering 10 years. J Craniomaxillofac Surg

2000;28:264-71.

27. Holmlund A, Hellsing G. Arthroscopy of the temporomandibular joint: occurrence

and location of osteoarthrosis and synovitis. in a patient material. Int. Oral Maxillofac.

Surg 1988;17:36-40.

28. Varol A, Basa S, Topsakal A, Akpinar I. Assessment of synovial vascularization by

power Doppler ultrasonography in TMJ internal derangements treated arthroscopically.

Br J Oral Maxillofac Surg 2008;46:625-30.

29. Bouloux GF. Temporomandibular joint pain and synovial fluid analysis: a review of

the literature. J Oral Maxillofac Surg. 2009;67:2497-504.

15

30. Mishra AM, Reddy SJ, Husain M, Behari S, Husain N, Prasad KN, Kumar S, Gupta

RK. Comparison of the magnetization transfer ratio and fluid-attenuated inversion

recovery imaging signal intensity in differentiation of various cystic intracranial mass

lesions and its correlation with biological parameters. J Magn Reson Imaging.

2006;24:52-6.

16

Legends

Figure 1. T2-weighted (a) and FLAIR (b) sagittal MR images of the TMJ. Joint effusion

(arrow) is demonstrated in upper joint space on both images.

Figure 2. Measurement of region of interest (ROI) on sagittal T2-weighted images in

closed-mouth position. ROIs were placed over gray matter (GM) (black arrow) and

cerebrospinal fluid (CSF) (white arrow) close to top of condyle. Joint effusion (JE)

(arrow heads) in this case shows extensive fluid, and ROI for JE was placed over

maximal area in upper joint space.

Figure 3. Schematic drawing of equation for the suppression ratio.

Figure 4. Signal intensity ratio (SIR) for joint effusion (JE) in group A. Strong

correlation observed between T2-weighted and FLAIR images for JE. Correlation

coefficient was 0.81 (P < .001) with Pearson product-moment correlation.

Figure 5. Signal intensity ratio (SIR) for joint effusion (JE) in Group B. A strong

correlation was observed between T2-weighted images and FLAIR images. Correlation

coefficient was 0.94 (P < .001) with Pearson product-moment correlation.

Table I. Inclusion and exclusion criteria.

Table II. MR imaging parameters.

Table III. Average signal intensity ratios from joint effusion (JE) and cerebrospinal

fluid (CSF).

Table IV. Average suppression ratio for signal intensity from joint effusion (JE) and

cerebrospinal fluid (CSF) by FLAIR.

Figure. 1. A Figure 1. B

Figure. 2

Figure. 3

Figure. 4 Figure. 5

Table I. Inclusion and exclusion criteria

Inclusion Criteria:

Temporomandibular joint with evidence of disk derangement.

Temporomandibular joint with amount of joint fluid categorized as “moderate”,

“marked” or “extensive” by Larheim and the others19.

Exclusion Criteria:

History of trauma or surgery, systematic inflammatory disease, mandibular

growth disturbances, or TMJ tumors.

Temporomandibular joint with amount of joint fluid categorized as “no fluid” or

“minimal” by Larheim and the others19.

Table II. MR imaging parameters.

Index T2-weighted images FLAIR images

TI (ms） 2,500

TR (ms） 3,300 9,000

TE (ms） 85 122 or 168

ETL 5 21

FOV（mm） 150×150 150×150

Section thickness（mm） 3 3

Slice gap（mm） 0.3 0.3

Matrix 256×256 or 512×512 256×256

Table III. Average signal intensity ratios from joint effusion (JE) and cerebrospinal

fluid (CSF).

Mean ± SD (%)

JE CSF

Group A

T2-weighted images

1.71 ± 0.64

1.44 ± 0.23

FLAIR images 1.15 ± 0.86 0.34 ± 0.22

Group B

T2-weighted images

1.81 ± 0.67

1.36 ± 0.33

FLAIR images 1.52 ± 0.61 0.35 ± 0.16

Table IV. Average suppression ratio for signal intensity from joint effusion (JE) and

cerebrospinal fluid (CSF) by FLAIR.

Mean ± SD (%)

Echo Time (ms)

JE 122 36.9 ± 25.3

168 16.3 ± 13.5

CSF 122 75.6 ± 17.1

168 71.7 ± 14.5