Session Information
Session Type: Abstract Submissions (ACR)
Background/Purpose: Mechanical back pain (MBP) patients and healthy subjects may show spinal MRI lesions suggestive for spondyloarthritis (SpA) such as corner inflammatory lesions (CIL) or corner fat lesions (CFL). There are few data about their frequency and whether they may result in a false positive classification of controls as having SpA. We aimed to assess the frequency of false positive classification as SpA in controls consisting of MBP patients and healthy volunteers, and to evaluate which MRI lesion type had the highest impact on misclassification.
Methods: The study sample comprised 2 independent cohorts A/B of 130 consecutive patients with back pain ≤50 years newly referred to 2 university clinics, and 20 healthy controls (HC). Patients were classified according to clinical examination and pelvic radiography as having non-radiographic SpA (nr-axSpA; n=50), ankylosing spondylitis (AS; n=33), or MBP (n=47). Age-matched healthy controls in cohort A were recruited according to the Nordic Questionnaire from hospital staff of the same clinic that also recruited the SpA patients. Spinal MRI scans were assessed by 4 blinded readers according to the Canada-Denmark MRI definitions for spinal lesions [1, 2]. Bone marrow edema (BME) and fat infiltration (FI) were reported in the central and lateral compartment of 23 discovertebral units. Readers recorded presence/absence of SpA and their level of confidence in this conclusion by global evaluation of the MRI scans (T1SE and STIR sequences) on a 0-10 scale (0 = definitely not SpA; 10 = definite SpA). The mean number (percentage) of controls misclassified as having SpA by ≥1 of 4 readers and the principal spinal lesions indicative of SpA were analysed descriptively.
Results: 33.9%/28.0% of MBP patients in cohorts A/B, and 26.3% of healthy subjects in cohort A were misclassified as having SpA by global assessment of MRI of the entire spine. Both BME and FI in varying percentages were the most important MRI lesions leading to this misclassification. The mean number of CIL and CFL observed in controls was lower than in SpA patients.
Mean percentage of false positive controls, principal MRI lesions responsible for misclassification, and mean number for CIL and CFL on spinal MRI in cohort A/B
|
Cohort |
Cohort A (n=62) |
Cohort B (n=88) |
|||||
|
Group |
nr-axSpA |
AS |
MBP |
HC |
nr-axSpA |
AS |
MBP |
|
Number of subjects |
19 |
9 |
14 |
20 |
31 |
24 |
33 |
|
Classification as SpA (%) |
51.3 |
72.2 |
33.9 |
26.3 |
48.4 |
72.9 |
28.0 |
|
BME most important (%) |
79.5 |
76.9 |
52.6 |
52.4 |
23.3 |
47.1 |
16.2 |
|
FI most important (%) |
17.9 |
23.1 |
26.3 |
23.8 |
66.7 |
45.7 |
62.2 |
|
CIL (mean) |
3.4 |
6.4 |
1.3 |
1.4 |
2.2 |
4.1 |
0.9 |
|
CFL (mean) |
3.9 |
6.8 |
3.2 |
2.8 |
6.5 |
14.3 |
3.0 |
|
Classification as SpA (%) |
35.5 |
55.6 |
14.3 |
7.5 |
31.5 |
62.5 |
10.6 |
Conclusion: 26% to 34% of healthy controls and patients with MBP were misclassified as having SpA by evaluation of MRI of the spine alone. Misclassifications with high confidence ranged from 8% to 14%. Caution is warranted if a classification of SpA is based on MRI of the spine alone.
References. [1] Lambert R et al. J Rheumatol 2009;36 suppl 84:3. [2] Ostergaard M et al. J Rheumatol 2009;36 suppl 84:18.
Disclosure:
U. Weber,
None;
Z. Zhao,
None;
V. Zubler,
None;
S. Chan,
None;
R. G. Lambert,
None;
M. Ostergaard,
None;
S. J. Pedersen,
None;
K. Rufibach,
None;
W. P. Maksymowych,
None.
« Back to 2013 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/frequent-detection-of-inflammation-and-fat-infiltration-suggestive-of-spondyloarthritis-on-mri-of-the-entire-spine-in-healthy-subjects-and-patients-with-mechanical-back-pain/