DOES DEDICATED EXTREMITY MRI RELIABLY DETECT RA BONE EROSIONS? A COMPARISON WITH HIGH RESOLUTION CT
[2007] [SAT0439] DOES DEDICATED EXTREMITY
MRI RELIABLY DETECT RA BONE EROSIONS? A COMPARISON WITH HIGH RESOLUTION CT
A. Duer 1, B.
Ejbjerg 2, E. Albrecht-Beste 3, A. Vestergaard 3,
U. Møller Døhn 4, M.L. Hetland 4, M. Østergaard 5
1Rheumatology, Copenhagen University Hospital and King Christian Xth
Hospital of Rheumatic Diseases, Hvidovre and Graasten, 2Rheumatology,
Copenhagen University Hospital, Herlev, 3Radiology, 4Rheumatology,
Copenhagen University Hospital, Hvidovre, 5Rheumatology, Copenhagen
University Hospital, Hvidovre and Herlev, Denmark
Objectives: With computed tomography (CT) as the reference method, to
determine whether bone erosions in rheumatoid arthritis (RA) wrist joints
detected with two different dedicated extremity magnetic resonance units
(E-MRI) and conventional radiography (CR), represent true erosive changes.
Methods: 20 RA patients (17 females, median 55 years, range 32-74) and 5
healthy controls (3 females, median 35 years, range 25-59) underwent, within 2
weeks, two MRI, a CR and a CT examination of the wrist of one hand. The 2 E-MRI
units were 1) A 0.2 T MRI unit (Artoscan; Esaote Biomedica, Genoa, Italy) using
a 3D gradient echo sequence with subsequent multiplanar recontruction (voxel
size 0.55 x 0.55 X 1 mm) and 2) A 0.2 T portable MRI system (MagneVu MV1000, Carlsbad,
USA) using a 3D spin echo sequence (16 slices, voxel size 1 x 1 x 0.94 cm). CT
was done on a Siemens, Somatom Sensation 10 (voxel size 0.6 x 0.6 X 0.6 mm). CR
was done in posterior-anterior and Nørgaard projections. In all modalities each
bone was blindedly evaluated for erosion and, furthermore, were the MagneVu
images assessed for the proportion (0-33, 34-66, 67-99 and 100%) of each bone
being visualized.
Results: 350 joint areas were examined. CT, Artoscan, MagneVu and CR
detected 99, 87, 33 and 27 bones with erosions respectively. These were located
as follows: radius and ulna: 15, 14, 2 and 10; carpal bones: 71, 60, 29 and 14
and metacarpal bases: 13, 13, 2 and 3. With CT as the reference, the
sensitivities for Artoscan, MagneVu and CR for detecting erosions in the wrist
were 0.60, 0.24 and 0.18 respectively (see table).
The MagneVu allowed visualization of 1.5 cm of the ventral-dorsal diameter
(slab thickness = 1 cm) of the bone. It was found that 31.6% of bones were
visualised entirely and 37.9% of bones were 67-99% visualized (see table).
|
|
Radius |
Carpal |
Phalangeal |
All |
|
|
and ulna |
bones |
bases |
wrist bones |
|
Sensitivity* |
|
|
|
|
|
Artoscan |
0.73 |
0.55 |
0.69 |
0.60 |
|
MagneVu |
0.08 |
0.27 |
0.22 |
0.24 |
|
CR |
0.53 |
0.11 |
0.15 |
0.18 |
|
Specificity* |
|
|
|
|
|
Artoscan |
0.91 |
0.84 |
0.95 |
0.89 |
|
MagneVu |
0.97 |
0.91 |
1.0 |
0.95 |
|
CR |
0.94 |
0.95 |
0.99 |
0.96 |
|
Accuracy* |
|
|
|
|
|
Artoscan |
0.86 |
0.74 |
0.92 |
0.81 |
|
MagneVu |
0.73 |
0.70 |
0.93 |
0.77 |
|
CR |
0.82 |
0.66 |
0.88 |
0.74 |
|
Bone-coverage by MagneVu |
|
|
|
|
|
0% |
12.5 |
25.5 |
13.1 |
19.7 |
|
1-33% |
14.6 |
1.6 |
3.6 |
3.9 |
|
34-66% |
10.4 |
7.3 |
4.8 |
6.7 |
|
67-99% |
45.8 |
24.5 |
56.0 |
37.9 |
|
100% |
16.7 |
41.1 |
22.6 |
31.6 |
*CT as the standard reference.
Conclusion: When applying CT as the reference method for detecting
erosions in the wrist, the Artoscan showed higher sensitivity (0.60) than the
MagneVu (0.24) and conventional radiography (0.18). The field of view of the
MagneVu is too small to vizualise all bones in the wrist in one examination,
which contributes to the low sensitivity.
Diagnostics and imaging
procedures
Citation: Ann Rheum Dis 2007;66(Suppl II):563