Semi-quantitative thigh magnetic resonance imaging scores in assessing disease activity and determining long-term clinical outcome in idiopathic inflammatory myopathies: a causal mediation analysis

To evaluate the relationship of thigh MRI (t-MRI) with manual muscle testing-8 (MMT-8), muscle enzymes and autoantibodies. To determine the causal and mediating factors resulting in poor recovery of MMT-8 in inflammatory myositis (IIM). This was a single-centre retrospective study in IIM patients. t-MRI was semi-quantitatively scored for muscle oedema, fascial oedema, muscle atrophy and fatty infiltration. Spearman correlation of t-MRI scores was done with muscle enzymes at baseline, and MMT-8 at baseline and on follow-up. Causal mediation analysis was performed with age, sex, symptom duration, autoantibodies, diabetes and BMI as independent variables, follow-up MMT-8 as dependent and t-MRI scores as mediating variables. Baseline evaluation was done on 59 and follow-up on 38 patients. Median follow-up of the cohort was 31 (10-57) months. Baseline MMT-8 negatively correlated with muscle oedema (r = -0755), fascial oedema (r = -0.443) and muscle atrophy (r = -0.343). Creatinine kinase (r = 0.422) and aspartate transaminase (r = 0.480) positively correlated with muscle oedema. Follow-up MMT-8 correlated negatively with baseline atrophy (r = -0.497) and fatty infiltration (r = -0.531). On follow-up, MMT-8 males had positive total effect (estimate (95%CI)) via atrophy [2.93 (0.44, 4.89)] and fatty infiltration [2.08 (0.54, 3.71)]. Antisynthetase antibody had a positive total effect via fatty infiltration [4.50 (0.37, 7.59)]. Age had a negative total effect via atrophy [-0.09 (0.19, -0.01)] and fatty infiltration [-0.07 (-0.15, -0.01)]. Disease duration had a negative total effect via fatty infiltration [-0.18 (-0.27, -0.02)]. Baseline fatty infiltration and muscle atrophy resulting from older age, female sex, longer disease duration and absent anti-synthetase antibodies, partly mediate muscle recovery in IIM.