Background/Purpose: in patients with active myositis, magnetic resonance imaging (MRI) often demonstrates muscle edema thought to represent active inflammation. An MRI-based muscle inflammatory scoring system has been validated for juvenile myositis (Rheumatology 2011; 50:2237), but to the best of our knowledge there is no such a scoring system for adult myositis. We aimed to validate an MRI-based edema score for patients with adult myositis.

Methods: twenty-one patients (9 polymyositis [PM], 12 dermatomyositis diagnosed according to Bohan and Peter criteria) were included in the study. In 8/9 patients with PM the diagnosis was confirmed by consistent histology. MRI of the thigh and pelvic floor muscles was performed on a 1.0 or 1.5T scanner using a surface coil. Edema (1= present, 0= absent) was assessed by fat-suppressed sequences in 17 thigh and pelvic floor muscles and a score (0-17) was calculated by adding the separate muscle scores by three blinded independent observers. Sensitivity to change was assessed by sign test. Muscle strength was evaluated in 12 muscle groups by manual muscle test (MMT) and graded according to the extended Medical Research Council scale (0-5). Serum creatine kinase (CK) levels were measured (U/l) and MMT performed simultaneously or within a week from the MRI study in each patient.

Results:  38 MRI scans were available from 21 patients. 10 patients had new onset of myositis. Two patients had 4 scans, 1 had 3 scans, 9 had 2 scans, and 9 had 1 scan each. The median MRI edema score was 2.7 and the interquartile range (IQR) 8.96. Inter-rater Cronbach’s alpha was 0.984, while single-measure intraclass correlation coefficient (ICC) was 0.954 (95% confidence interval [CI] 0.922-0.974). Bland-Altman’s plot showed no significant differences between raters. Two raters evaluated the same MRI exams on 2 different time points to determine intra-rater variability. Cronbach’s alpha was 0.903 and 0.921, while single-measure ICC was 0.815 (95% CI 0.667-0.901) and 0.851 (95% CI 0.729-0.921) for rater 1 and 2, respectively. In 7 patients for whom a repeat MRI after treatment onset was available median edema score decreased from a median of 8.83 (IQR 9.18) to 0.71 (IQR 12.3) (p 0.07). MRI edema score did not correlate with serum CK levels (median 384, IQR 1317) or MMT (median 4.39, IQR 0.45) (p>0.05 Spearman’s rho).

Conclusion:  the MRI edema score presented herein is highly reproducible with good inter- and intra-rater variability and appears to be sensitive to change. The significant lack of correlation of MRI edema score with serum CK levels and MMT suggests that these measures of myositis activity may be at least partially uncoupled, suggesting that MRI score adds information on disease activity above the CK and MMT, although the small sample size hampers a confident conclusion in this regard.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/preliminary-validation-of-a-magnetic-resonance-imaging-based-inflammatory-scoring-system-in-adult-myositis/