Many neuromuscular disorders involve loss and fat replacement of skeletal muscle. Volumetric MRI-based analysis of muscle volume and muscle fat has emerged as a promising technique for monitoring disease progression and evaluating treatment response. However, the experience of using the approach in sporadic inclusion body myositis (sIBM) remains limited.
This work aimed at evaluating the within-scanner repeatability of MRI-based muscle composition analysis in such a patient population. N = 10 sIBM patients underwent two MRI examinations in a test-retest setup on a 3T Siemens Verio. A T1-weighted Dixon acquisition with coverage of the full legs was employed. Subsequently, the following anatomical regions were segmented: quadriceps, hamstrings, adductors, gastrocnemius medialis and tibialis anterior. The segmentations were used for measuring the lean muscle volume (LMV), the muscle fat fraction (MFF), and the muscle fat infiltration (MFI, defined as the MFF of the contractile muscle tissue) of each region bilaterally. Lastly, the cohort average as well as the within-subject standard deviation (sw) and the within-subject coefficient of variation (CVw) were computed for each region and measurement type.
The average LMV of the different regions within the cohort ranged from 6.9 cl (tibialis anterior) to 72.0 cl (quadriceps), the average MFF from 21.7% (adductors) to 43.5% (gastrocnemius medialis), and the average MFI from 13.0% (adductors) to 22.5% (gastrocnemius medialis). The measurements showed a high repeatability although the precision varied among regions; sw ranged from 0.27 – 1.80 cl for LMV, from 0.68 – 1.32 percentage points for MFF, and from 0.19 – 0.71 percentage points for MFI. Similarly, CVw varied between 1.48% – 9.34% for LMV, 2.95% – 7.14% for MFF, and 0.91% – 4.53% for MFI.
Volumetric MRI-based muscle composition analysis of sIBM patients proved to be highly repeatable, which is a prerequisite for usage in natural history studies as well as clinical trials.