DOI:10.2214/AJR.05.0098
AJR 2006; 187:358-363
© American Roentgen Ray Society
Periarticular Bone Findings in Rheumatoid Arthritis: T2-Weighted Versus Contrast-Enhanced T1-Weighted MRI
Lawrence Yao1,
Michael Magalnick2,
Mildred Wilson2,
Peter Lipsky2 and
Raphaela Goldbach-Mansky2
1 Diagnostic Radiology Department, National Institutes of Health, 10 Center
Drive, Room 1C640, Bethesda, MD 20892-1182.
2 National Institute of Arthritis, Musculoskeletal, and Skin Diseases, National
Institutes of Health, Bethesda, MD.
Received January 19, 2005;
accepted after revision July 7, 2005.
Address correspondence to L. Yao.
Abstract
OBJECTIVE. The purpose of this study was to establish the relative
predictive value of T2-weighted and contrast-enhanced T1-weighted MRI
techniques for bone erosions that are evident on CT. Because it is known that
MRI depicts abnormalities in the periarticular bone of patients with
rheumatoid arthritis, we wanted to compare the outcomes of T2-weighted versus
contrast-enhanced T1-weighted MRI techniques.
MATERIALS AND METHODS. Eleven patients with rheumatoid arthritis
underwent CT imaging of their most affected wrist. Fast spin-echo T2-weighted
MR images were then acquired with spectral fat saturation. Enhanced
T1-weighted spin-echo images acquired before and after IV administration of
gadopentetate dimeglumine were used to determine the percent enhancement.
Imaging examinations were scored for 15 anatomic zones. The CT score was based
on cortical bone erosion. The MR score was based on periarticular bone marrow
signal alteration.
RESULTS. Both T2-weighted MR images with spectral fat saturation and
enhanced T1-weighted images were concordant for the presence or absence of
bone abnormalities in 122 of 165 zones (74%). Of the 43 zones that were
discordant for an abnormality by the two MR techniques, the T2-weighted images
were positive in five zones, and enhanced T1-weighted images were positive in
38 zones (p < 0.001). Of the 43 zones that were discordant by the
two MR techniques, enhanced T1-weighted images were concordant with CT in 20
zones, whereas the T2-weighted images were concordant with CT in 23 zones
(p = 0.76). A greater proportion of lesions detected by the
T2-weighted images were "edema-like" signal patterns.
CONCLUSION. In rheumatoid arthritis, contrast-enhanced T1-weighted
MRI depicts more periarticular bone abnormalities than fat-suppressed
T2-weighted MRI. These MR techniques are equally predictive of frank, erosive
disease that is evident on CT.
Keywords: arthritis • MRI • MR technique • musculoskeletal system • rheumatoid arthritis • wrist
Introduction
MRI depicts many important articular manifestations of rheumatoid
arthritis. MRI extends visualization of disease effects beyond cortical bone
to include processes in synovium, periarticular soft tissues, and the marrow
space. While the role of MRI in the early diagnosis of rheumatoid arthritis is
still evolving [1], the breadth
of information provided by MRI in the assessment of diseased joints may bring
greater power to clinical trials that study disease-modifying therapies
[2-4].
Early clinical MRI studies emphasized the efficacy of MRI for detection and
quantification of synovial proliferation in rheumatoid arthritis
[5-7].
MRI of rheumatoid joints in these applications was performed with an IV
contrast agent. Unenhanced T1-weighted MRI, however, is useful for depicting
signal changes in cortical and subcortical bone that correspond to erosive
disease [8]. Fat-suppressed MRI
techniques, particularly with T2 weighting, may reveal more ill-defined signal
change in the marrow space around rheumatoid joints that has been described as
an "edema pattern"
[1,
9]. These findings in the
marrow space are conspicuous even on lower-resolution MRI scans or on STIR
sequences.
T2-weighted imaging adds scanning time to already lengthy assessments of
rheumatoid joints when high-resolution T1-weighted sequences, before and after
contrast administration, may also be needed. In this study, we examine the
contribution of fat-suppressed T2-weighted MRI, compared with
contrast-enhanced T1-weighted MRI, in the assessment of periarticular bone in
rheumatoid arthritis patients. This study addresses, in part, the question of
whether contrast-enhanced MRI alone is a sufficient diagnostic test for
rheumatoid disease processes in periarticular bone.
View larger version (130K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —43-year-old woman with rheumatoid arthritis. Fat-suppressed
T2-weighted MR image depicts large osseous lesion at proximal aspect of lunate
(arrow). Other less hyperintense osseous abnormalities are present
elsewhere in carpus.
|
|
View larger version (86K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —43-year-old woman with rheumatoid arthritis. Corresponding
enhanced T1-weighted image similarly depicts large lesion in lunate
(arrow). Mild synovitis is more clearly evident (arrowhead)
on enhanced imaging.
|
|
Materials and Methods
The study included 11 patients (9 women and 2 men). Mean patient age was 48
years (range, 32-74 years). All patients had active rheumatoid arthritis with
an incomplete response to methotrexate therapy (administrated at 15 mg/wk for
more than 6 weeks). The most severely involved wrist, as determined by
clinical examination, underwent CT and MRI. The time interval between CT and
MRI examinations was less than 1 week.
CT scans were acquired on a 10-cm field of view using 1-mm scan thickness.
MRI was performed on a 1.5-T scanner (Signa, GE Healthcare) using a receiver
coil designed for the wrist. The imaging field of view was 10 cm. T1-weighted
spin-echo imaging was performed in the axial and coronal planes without fat
saturation, before and after administration of gadopentetate dimeglumine at a
dose of 0.1 mmol/kg. Images were acquired at a 1-mm slice thickness in an
interleaved, contiguous (no gap) fashion (TR range/TE, 600-650/11; matrix =
256 x 256; average = 1). Fast spin-echo T2-weighted imaging was
performed in axial and coronal planes using frequency-selective presaturation
to suppress signal from fat (TR range/TEeff, 2,500-2,800/68;
echotrain length, 4; matrix, 256 x 256; average, 2). Images were
acquired with a 3-mm slice thickness and a 1-mm interslice gap.
Unenhanced and contrast-enhanced T1-weighted spin-echo images were
postprocessed using custom plug-ins for imageJ, an open source software
platform written in Java (Sun Microsystems). Contrast-enhanced images were
registered to motion-corrected unenhanced images using an unsupervised
intensity-based algorithm
[10]. Images reflecting
percentage enhancement were generated by subtracting unenhanced images from
contrast-enhanced images, adjusting for the noise threshold. These images will
be referred to as enhanced T1-weighted images. Percentage enhancement values
were also used to generate pseudo color images that were fused with unenhanced
T1-weighted images to facilitate image interpretation.
The CT, T1-weighted MR images, and fat-suppressed T2-weighted MRI images
were scored in independent reviewing sessions, separated in time by more than
1 month. For purposes of scoring, the wrist examinations were divided into 15
anatomic zones, consisting of the eight carpal bones, the distal radius,
distal ulna, and the five metacarpal bases. An experienced musculoskeletal
radiologist assigned semiquantitative scores to each anatomic zone.
The CT scoring was based on alterations in the contour of cortical bone.
The CT scoring system was defined as: 0 = normal, 1 = single definite erosion,
2 = multiple but noncontiguous discrete erosions, 3 = confluent erosive
disease, 4 = confluent erosive disease involving multiple joint compartments
or with associated bone deformity.
The MR scoring was based on signal abnormalities within osseous zones,
either within the marrow space or within subcortical or cortical regions. The
MR scoring system was defined as follows: 0 = normal, 1 = single focal signal
abnormality, 2 = multiple noncontiguous focal signal abnormalities, 3 =
confluent abnormalities that are ill-defined (occupying less than 50% of the
regional bone volume), 4 = confluent abnormality involving greater than 50% of
the regional bone volume. By the defined scoring system, MR scores of 3 and 4
correspond to a so-called marrow "edema pattern," whereas MR
scores of 1 and 2 correspond to focal osseous findings.
Statistical testing for differences in the incidence of positive findings
and for differences in agreement was done with the McNemar test. Differences
in CT scores for zones grouped by MR scores were tested with a Mann-Whitney
test.
Results
MRI: Contrast-Enhanced Versus T2-Weighted Imaging
The two imaging techniques were concordant for the presence or absence of
abnormalities in 74% (122/165) of zones. Figures
1A,
1B, and
1C illustrates concordance
between the two techniques for a focal bone lesion. Enhanced T1-weighted
imaging depicted abnormalities in more zones than fat-suppressed T2-weighted
fast spin-echo imaging. Figures
2A,
2B, and
2C illustrates a small bone
lesion detected by enhanced T1-weighted imaging, but not prospectively
identified on T2-weighted imaging. The incidence of positive findings on
enhanced T1-weighted imaging was 71% (117/165) versus 51% for T2-weighted
imaging (84/165). The difference in the incidence of positive findings between
the two MR techniques was significant (p < 0.001).
View larger version (125K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —37-year-old woman with rheumatoid arthritis. Fat-suppressed
T2-weighted coronal MRI shows prominent, ill-defined marrow signal
abnormality, or edema pattern, in third metacarpal base (arrow).
|
|
View larger version (79K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3D —37-year-old woman with rheumatoid arthritis. Adjacent and
more distal CT image shows potential deepening of enthesis at ulnar aspect of
third metacarpal base (arrow), which was not interpreted as
erosion.
|
|
MRI: Edema-Like Signal
As discussed in the Materials and Methods section, scores of either 3 or 4
correspond to edema-like signal patterns, whereas scores of 1 or 2 correspond
to focal abnormalities. Of the 84 abnormal zones, 24 (29%) on T2-weighted
images were scored as edema-like, compared with 15 of 117 (13%) abnormal zones
on enhanced T1-weighted imaging. The two MR techniques were concordant for
edema-like findings in 13 zones. T2-weighted imaging depicted edema-like
findings in 11 of these 13 discordant zones. This higher incidence of apparent
marrow "edema" on T2-weighted imaging is significant (p
< 0.05). Figures 3A,
3B,
3C, and
3D illustrates the potential
difference in the appearance of a marrow finding with edema-like features on
the two MR techniques.
MRI Findings: Correlation with Bone Erosion Determined by CT
CT depicted bone erosion in 55% (90/165) of anatomic zones. Enhanced
T1-weighted imaging was positive in 97% (87/90) of zones that were positive
for bone erosion on CT, whereas T2-weighted imaging was positive in 83%
(72/87). Figures 1A,
1B, and
1C illustrates a zone that was
scored positive by all techniques.
T2-weighted imaging was not positive in any zones that exhibited bone
erosion on CT and were negative on enhanced T1-weighted imaging. Conversely,
enhanced T1-weighted imaging was positive in 40% (30/75) of zones that were
negative on CT, whereas T2-weighted was positive in 16% of these zones
(12/75). Figures 3A,
3B,
3C, and
3D illustrates a region scored
positive on the two MR techniques and scored negative on CT.
Overall, enhanced T1-weighted imaging concurred with CT for the presence or
absence of abnormalities in 80% (132/165) of zones, whereas T2-weighted MRI
concurred with CT in 82% (135/165) of zones. This difference in agreement is
not significant (p =0.76).
Discussion
In this study, contrast-enhanced MRI depicted more abnormalities within the
osseous structures of the rheumatoid wrist than corresponding fat-suppressed
T2-weighted fast spin-echo imaging. The contrast-enhanced imaging depicted a
preponderance of small or more focal osseous abnormalities. Although the
detection rate of fat-suppressed T2-weighted fast spin-echo imaging was lower,
this technique tended to depict a greater percentage of marrow abnormalities
that were edema-like. This observation gains relevance in light of studies
suggesting that marrow edema patterns detected by MRI in patients with
rheumatoid arthritis may be equally or more predictive of future erosive
disease than the presence or severity of synovitis
[2,
9].
Although the two MR techniques differed in detection rate and in the
characteristics of bone findings, both techniques were similarly predictive of
periarticular bone erosion as shown by CT. Our study results suggest that
contrast-enhanced T1-weighted MRI may be more sensitive but less specific than
T2-weighted MRI for erosive disease shown by CT. In other words,
contrast-enhanced MRI reveals many small osseous findings that are either
absent or inconspicuous on CT. However, CT is not a gold standard for bone
damage in rheumatoid arthritis. The prognostic value and natural history of
radiographically occult MRI findings in periarticular bone is an important
topic for future investigation.
Are T2-weighted images necessary in MRI assessment of rheumatoid arthritis?
Certainly, T2-weighted images are indispensable if an IV contrast agent is not
used. Published and clinical experiences suggest that synovial proliferation
is less well depicted by unenhanced imaging
[11,
12]. For this reason, MR
protocols for the assessment of rheumatoid arthritis routinely include IV
contrast administration. An official scoring method for assessing rheumatoid
arthritis using MRI has been established
[13]. This standard includes
the use of IV-administered contrast material, but it also advocates the use of
T2-weighted sequences for identifying edema-like lesions in bone. Our study
results support the inclusion of T2-weighted sequences if a scoring system
incorporates the presence of edema-like marrow space findings, as it probably
should. Similarly, our study results suggest these scores would be influenced
by whether they are based on T2-weighted or contrast-enhanced MRI.
Our study contains some technical aspects that create potential limitations
and warrant further discussion. Contrast-enhanced imaging was analyzed in this
study using digital subtraction and image registration techniques. This
approach yields more precise information about contrast enhancement by
normalizing baseline signal values and avoiding the pitfalls of inhomogeneous
fat suppression. To increase study efficiency in clinical settings,
contrast-enhanced imaging is typically performed using spectral fat
saturation. Fat-suppressed contrast-enhanced images may be interpreted in
these settings without direct reference to unenhanced signal values. Whether
these disparate assessments of contrast enhancement spawn systematic
differences in interpretation is open to question. In reality, the time saved
by a fat-saturation strategy can be parlayed into improved image resolution
with separate, attendant advantages.
A systematic difference, or bias, between the deployment of
contrast-enhanced and T2-weighted MRI in our study is a difference in spatial
resolution. The T1-weighted images were generated from thinner image sections,
which may partially explain the higher detection rate for small or more focal
lesions with this technique. This bias is inherent to clinical imaging,
because T2-weighted and STIR techniques are relatively noise limited and
therefore warrant larger voxel sizes. These physical differences in
acquisition technique, however, are unlikely to explain the greater incidence
of apparent marrow edema patterns shown by T2-weighted imaging in our
study.
This study has addressed osseous abnormalities in the rheumatoid wrist as
detected by MRI. By incorporating small abnormalities in our scoring system,
particularly those in cortical or subcortical locations, we have likely
included disease that may not be strictly osseous. Enhancement or signal
alterations may arise from pannus or other reactive tissue occupying an
erosion crater, for example. The actual interface between erosion and the
marrow space may be difficult to define by MRI. This challenge is especially
important in light of efforts to use MRI to quantify erosion volumes as an
index of disease [14].
Finally, the use of a single observer is a structural weakness of our
study. Although potential observer bias may influence the absolute scores for
bone involvement, the relative differences in scores between imaging
techniques are hopefully still meaningful. This study highlights the
interaction of MRI methods and standardized scoring of disease. MRI protocols
and scoring methods both have important effects on interpreter variance, which
was not addressed in this study.
In summary, our study illustrates that contrast-enhanced MRI detects more
periarticular osseous abnormalities in rheumatoid arthritis than
fat-suppressed T2-weighted MRI or CT and that these tend to be smaller, focal
osseous lesions. The two MRI techniques are equally predictive of erosive
disease as revealed by CT. Despite the lower detection rate of fat-suppressed
T2-weighted imaging for osseous lesions, the potentially better depiction of
marrow edema as a feature of individual bone lesions on fat-suppressed
T2-weighted MRI may warrant its continued application in MRI assessments of
rheumatoid arthritis.
References
- McQueen FM. Magnetic resonance imaging in early inflammatory
arthritis: what is its role? Rheumatology (Oxford)2000; 39:700
-706
- McQueen FM, Stewart N, Crabbe J, et al. Magnetic resonance imaging
of the wrist in early rheumatoid arthritis reveals a high prevalence of
erosions at four months after symptom onset. Ann Rheum
Dis 1998; 57:350
-356[Abstract/Free Full Text]
- Ostergaard M, Hansen M, Stoltenberg M, et al. New radiographic bone
erosions in the wrists of patients with rheumatoid arthritis are detectable
with magnetic resonance imaging a median of two years earlier.
Arthritis Rheum 2003;48
: 2128-2131[CrossRef][Medline]
- McQueen FM, Benton N, Crabbe J, et al. What is the fate of erosions
in early rheumatoid arthritis? Tracking individual lesions using X-rays and
magnetic resonance imaging over the first two years of disease. Ann
Rheum Dis 2001; 60:859
-868[Abstract/Free Full Text]
- Sugimoto H, Takeda A, Masuyama J, Furuse M. Early-stage rheumatoid
arthritis: diagnostic accuracy of MR imaging.
Radiology 1996;198
: 185-192[Abstract/Free Full Text]
- Sugimoto H, Takeda A, Kano S. Assessment of disease activity in
rheumatoid arthritis using magnetic resonance imaging: quantification of
pannus volume in the hands. Br J Rheumatol1998; 37:854
-861[Abstract/Free Full Text]
- Ostergaard M, Hansen M, Stoltenberg M, et al. Magnetic resonance
imaging-determined synovial membrane volume as a marker of disease activity
and a predictor of progressive joint destruction in the wrists of patients
with rheumatoid arthritis. Arthritis Rheum1999; 42:918
-929[CrossRef][Medline]
- McQueen FM, Stewart N, Crabbe J, et al. Magnetic resonance imaging
of the wrist in early rheumatoid arthritis reveals progression of erosions
despite clinical improvement. Ann Rheum Dis1999; 58:156
-163[Abstract/Free Full Text]
- McQueen FM, Benton N, Perry D, et al. Bone edema scored on magnetic
resonance imaging scans of the dominant carpus at presentation predicts
radiographic joint damage of the hands and feet six years later in patients
with rheumatoid arthritis. Arthritis Rheum2003; 48:1814
-1827[CrossRef][Medline]
- Thevenaz P, Ruttimann UE, Unser M. A pyramid approach to subpixel
registration based on intensity. IEEE Transactions on Image
Processing 1998; 7:27
-41
- Nakahara N, Uetani M, Hayashi K, Kawahara Y, Matsumoto T, Oda J.
Gadolinium-enhanced MR imaging of the wrist in rheumatoid arthritis: value of
fat suppression pulse sequences. Skeletal Radiol1996; 25:639
-647[CrossRef][Medline]
- Konig H, Sieper J, Wolf KJ. Rheumatoid arthritis: evaluation of
hypervascular and fibrous pannus with dynamic MR imaging enhanced with
Gd-DTPA. Radiology 1990;176
: 473-477[Abstract/Free Full Text]
- Ostergaard M, Peterfy C, Conaghan P, et al. OMERACT Rheumatoid
Arthritis Magnetic Resonance Imaging Studies: core set of MRI acquisitions,
joint pathology definitions, and the OMERACT RA-MRI scoring system.
J Rheumatol 2003;30
: 1385-1386[Medline]
- Bird P, Lassere M, Shnier R, Edmonds J. Computerized measurement of
magnetic resonance imaging erosion volumes in patients with rheumatoid
arthritis: a comparison with existing magnetic resonance imaging scoring
systems and standard clinical outcome measures. Arthritis
Rheum 2003; 48:614
-624[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
