Session Information
Session Type: Abstract Submissions (ACR)
Background/Purpose: Limitation of x-ray joint damage is a key indicator of therapeutic efficacy in rheumatoid arthritis (RA). Although magnetic resonance imaging (MRI) is increasingly used due to its greater sensitivity vs radiographs, summary evidence validating MRI features of RA joint inflammation and damage is lacking. By systematic literature review (SLR), we examined the validity of MRI for assessment of RA wrist/hand features according to Outcome Measures in Rheumatology Clinical Trials (OMERACT) criteria.
Methods: An SLR in PubMed with Cochrane hedge was conducted using search terms: RA, AND MRI, AND specific terms i.e., synovitis, joint space narrowing (JSN), erosions or bone marrow edema (BME) AND humans AND randomized controlled trials [RCTs], clinical studies; 1970 to Aug 2011; English. Pairs of authors evaluated titles/abstracts, selecting articles for extraction according to these criteria: adults with RA, MRI of hands and/or wrists assessing at least one of the following: synovitis, BME/osteitis, tenosynovitis, erosions, JSN; RCT, observational study, or case series ≥10 patients. To achieve ≥ 95% consistent data extraction, authors evaluated the same 5 articles, ensuring consensus on data extraction methods. Each author then extracted a proportion of the articles. Data extracted included MRI, field strength, measurement methods, and validity criteria: criterion, content, construct, reliability, responsiveness, and discrimination. Level of evidence was assessed using Cochrane Handbook criteria, adapted for imaging research.
Results: Of 575 MRI titles/abstracts, 180 met criteria with 81 having at least 1 type of validation. Although MRI measurement methods were developed using several approaches, OMERACT RA MRI scoring (RAMRIS) was determined to be the dominant method. As 43 articles utilized 1.5T, and 7 used other field strengths; only 1.5T articles were analyzed. (Table) 19 articles including 6 RCTs using 1.5T and RAMRIS were extracted seeking validation of MRI measures. The table summarizes the number of articles with data for validation. Histologic or radiographic evidence for criterion validity was extracted from 5 other articles.
Conclusion: Using a rigorous PRISMA-compliant method for quality assurance and uniform article extraction in a SLR, we found that synovitis, osteitis/BME, and erosions are the best-validated features of joint inflammation and damage for RA wrist and hand. Assessment of other measurement methods, field strengths and joints requires further work.
Table: Validation status for measurement with 1.5T MRI images of synovitis, tenosynovitis, osteitis, erosions, joint space narrowing (JSN) — number of articles providing data
|
Type of validity
|
Criterion
|
Content
|
Construct (Discriminant)
|
Construct (Convergent)
|
Responsiveness to change
|
Intra-rater reliability
|
Inter-rater reliability
|
Summary
|
|
|
Definition
|
Extent to which measure agrees with “gold standard” |
Describes joint features regardless of age, gender, RA duration, treatment |
Correlation with clinical assessment of joint status |
Compares correlation coefficients between scores on the same health component, as measured by two different instruments |
Sensitivity to change |
Same reader scores similarly (ICCs )
|
Two or more readers score similarly (ICCs )
|
# validated of 7 required validity criteria
|
|
|
Synovitis (RAMRIS) |
1 |
2 |
12 |
4 |
1 |
7 |
6 |
5 |
7 |
|
0 |
1 |
3 |
|
|
1 |
|
|
||
|
Tenosynovitis |
1 |
|
5 |
1 |
*(a)
|
3 |
2 |
1 |
5 |
|
0 |
1 |
|
|
|
1 |
|
|
|
|
|
Erosions (RAMRIS) |
1 |
1 |
13 |
|
2 |
3 |
7 |
7 |
6 |
|
0 |
1 |
5 |
1 |
|
7 |
|
|
||
|
BME/Osteitis (RAMRIS) |
1 |
3 |
9 |
2 |
*(b)
|
5 |
6 |
5 |
6 |
|
0 |
|
4 |
|
|
2 |
|
|
|
|
|
JSN |
1 |
|
1 |
|
|
|
|
|
1 |
|
0 |
|
|
|
|
1 |
|
|
|
|
1=statistically significant validation; 0=non-significant data; ICC = intra/inter-class correlation coefficient
* Requires histology: a) Jain A, et al. Arthritis Rheum 2001;44:1754Ð60; b) Jimenez-Boj E, et al. Arthritis Rheum 2007;56:1118Ð24.
Disclosure:
T. Woodworth,
None;
O. Morgacheva,
None;
O. Troum,
Genentech,
2,
Genentech,
5;
O. Pimienta,
None;
P. Maranian,
None;
V. K. Ranganath,
UCB, BMS, Celgene,
2,
UCB,
5;
D. Furst,
BMS, Centocor, UCB, Genentech, Amgen,
2,
BMS, Centocor, UCB, Amgen, Abbott,
5.
« Back to 2012 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/assessment-of-validity-of-magnetic-resonance-imaging-measures-of-joint-inflammation-and-damage-in-rheumatoid-arthritis-wristhand-a-systematic-literature-review/