Background/Purpose:

Magnetic resonance imaging (MRI) has been used as a non-invasive tool to aid diagnosis and monitor disease activity in the idiopathic inflammatory myopathies (IIMs). Recent research has focused on whether radiological features may help subtype a patient with IIM, as this can be difficult to determine clinically and false negative results may occur with muscle biopsy. Herein, we describe the MRI findings in patients with IIM and identify radiological patterns that may discriminate between the subtypes of polymyositis (PM), dermatomyositis (DM), inclusion body myositis (IBM) and necrotising autoimmune myopathy (NAM).

Methods:

We retrospectively reviewed 114 lower-limb MRIs performed on 66 patients with IIM and 10 patients with non-inflammatory muscle histology (non-IIM) between 2009 and 2017 at the Royal Adelaide Hospital. Three patients did not satisfy EULAR/ACR criteria for IIM due to either normal creatinine kinase (CK) or strength, but had convincing clinical and histological features. A number of patients (n = 28) had serial images performed. Two musculoskeletal radiologists (SP, NB) independently quantified the degree of muscle oedema (MO), fatty replacement (MFR) and atrophy (MA). Clinical information was prospectively recorded. Ordinal logistic regression was performed analysing radiological grades for each IIM subtype. Spearman correlations were performed to evaluate associations between radiological grades and clinical parameters. Fisher’s exact test was used to analyse categorical data. Grades were compared between time points using the Wilcoxon signed rank test.

Results:

There was low inter-rater reliability (kappa < 0.60) between radiologists for numerous muscular compartments and, as such, a finding was only considered convincing if it was statistically significant for both radiologists. Pelvic and adductor MO, MFR and MA and posterior thigh MA significantly increased the likelihood of NAM. Anterior thigh MA and MFR significantly increased the likelihood of IBM. Pelvic MO decreased the likelihood of PM. Radiological changes (MO, MFR or MA) in adductors, anterior thighs or posterior thighs significantly reduced the likelihood of a non-IIM diagnosis. Measures of disease activity by visual analogue scale, muscle strength and CK did not correlate with radiological muscle oedema. Follow up scans were performed at a median of 370 days (IQR 217-528 days) and the most striking observation was of stability in radiological grades in many muscular compartments. A significant increase in MA and MFR grades were observed in the anterior thigh and anterior lower limbs, due to progression in a minority of patients.

Conclusion:

MRI patterns help discriminate between IIM subtypes. The observed stability over time in MA and MFR in multiple muscle compartments is reassuring and may reflect that these patients were treated intensively with suppression of inflammation, such that further muscle damage was not detected. Clinical disease activity assessments do not seem to reflect the degree of radiological muscle inflammation and, as such, MRI may be a useful adjunct for monitoring disease activity in clinical practice.

« Back to 2018 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/patterns-of-muscle-oedema-atrophy-and-fatty-replacement-in-the-idiopathic-inflammatory-myopathies-a-single-centre-retrospective-review-of-magnetic-resonance-imaging-data/