Refining MRI pattern in sarcoglycanopathies: upper body pattern and new approaches to assess disease progression

Sarcoglycanopathies are autosomal recessive limb-girdle muscular dystrophies caused by mutations in γ, α, β, and δ sarcoglycan genes. The lower limb (LL) MRI pattern has already been reported and gives diagnostic clues. Instead, the MRI pattern of the upper body (UB) has not been described and could be valuable to understand the imaging phenotype and progression of muscle involvement in sarcoglycanopathies. To describe the UB MRI pattern in sarcoglycanopathies and its correlation with pattern of the LL, disease duration and motor status. MRI from UB and LL of 64 patients with LGMDR3-6 were included in an international collaborative study. Muscle fatty replacement (FR) was evaluated semiquantitatively on T1-weighted images according to Lamminen score, and represented in a heatmap. We correlated FR of each muscle with motor status. We assessed variability of the FR score within each muscle using the homogeneity index (HI). We also tested the relationship between ambulatory status and FR along the different body regions with random forests. Partial correlation analysis and proportional odds logistic regressions were used to assess the relationship between FR in each muscle, disease duration and age at onset. The involvement of the UB muscles started early during the ambulant period. Some muscles of the scapular girdle (latissimus dorsi, subscapularis and serratus anterior) are early involved while muscles from head, neck and forearms are mostly spared, even in advanced disease stages. A caudocranial gradient of involvement in axial muscles (mainly, paravertebral muscles and trapezius) was consistently found in most of the patients. FRs of most scapular girdle and LL muscles are correlated with motor status. Ambulant patients show lower HI than non-ambulant patients. Random forests trained with information from different body parts predicts well ambulatory status with small differences in accuracy. Partial correlations of disease duration controlled by age at onset were high or moderate. The effect of disease duration seems higher in patients with earlier onset. UB fingerprint in sarcoglycanopathy exists and it is related with LL involvement and functional status. Disease duration is the main driver of FR in sarcoglycanopathies and early onset seems to accelerate it. Sarcoglycanopathies are autosomal recessive limb-girdle muscular dystrophies caused by mutations in γ, α, β, and δ sarcoglycan genes. The lower limb (LL) MRI pattern has already been reported and gives diagnostic clues. Instead, the MRI pattern of the upper body (UB) has not been described and could be valuable to understand the imaging phenotype and progression of muscle involvement in sarcoglycanopathies. To describe the UB MRI pattern in sarcoglycanopathies and its correlation with pattern of the LL, disease duration and motor status. MRI from UB and LL of 64 patients with LGMDR3-6 were included in an international collaborative study. Muscle fatty replacement (FR) was evaluated semiquantitatively on T1-weighted images according to Lamminen score, and represented in a heatmap. We correlated FR of each muscle with motor status. We assessed variability of the FR score within each muscle using the homogeneity index (HI). We also tested the relationship between ambulatory status and FR along the different body regions with random forests. Partial correlation analysis and proportional odds logistic regressions were used to assess the relationship between FR in each muscle, disease duration and age at onset. The involvement of the UB muscles started early during the ambulant period. Some muscles of the scapular girdle (latissimus dorsi, subscapularis and serratus anterior) are early involved while muscles from head, neck and forearms are mostly spared, even in advanced disease stages. A caudocranial gradient of involvement in axial muscles (mainly, paravertebral muscles and trapezius) was consistently found in most of the patients. FRs of most scapular girdle and LL muscles are correlated with motor status. Ambulant patients show lower HI than non-ambulant patients. Random forests trained with information from different body parts predicts well ambulatory status with small differences in accuracy. Partial correlations of disease duration controlled by age at onset were high or moderate. The effect of disease duration seems higher in patients with earlier onset. UB fingerprint in sarcoglycanopathy exists and it is related with LL involvement and functional status. Disease duration is the main driver of FR in sarcoglycanopathies and early onset seems to accelerate it.