Muscle MRI and functional outcome measures in Becker muscular dystrophy

Figure 1. Hierarchically clustered heatmap of fatty substitution Mercuri scores in 3 right upper limb and 22 right lower limb muscles, in MRIs from 51 BMD individuals. Rows, each corresponding to one individual, are hierarchically clustered (dendogram on the left) based solely on MRI data (Mercuri scores). Each column in the heatmap corresponds to one muscle, ordered left to right from cranial to caudal. A green-yellow-red gradient in the heatmap indicates increasing fatty substitution (legend to the right). The five columns left of the heatmap annotate individual clinical features (not used in the hierarchical clustering algorithm): mutation group, age, 6 Minute Walk Test (6MWT) distance, North Star Ambulatory Assessment (NSAA score), and presence of dilated cardiomyopathy (DCM), color-coded as detailed in the legend on the left.

Figure 2. Representative T1w MRI scan images. (A) Thigh MRI scan (T1w sequence) of four BMD patients (aged 40–42 years) with different genotypes. Patients with del 48 and del 48–51 mutations showed a normal signal without signs of fatty replacement. On the contrary, patients with del 45–47 and del 45–48 mutations showed severe fatty replacement, in particular of the gluteus maximus (Mercuri score 2b-4). Note the sparing of the gracilis (white arrowhead) and sartorius (white arrow) muscles. (B) Leg MRI scan (T1w sequence) of four BMD patients (aged 40–42 years) with different genotypes. Patients with del 48 and del 48–51 mutations showed a normal signal without signs of fatty replacement, while patients with del 45–47 and del 45–48 mutations showed advanced fatty replacement, in particular of the gastrocnemii (Mercuri score 2b-3), and a moderate involvement of soleus, peroneus, and tibialis anterior (score 2b). Conversely, muscle MRI showed a lesser involvement of the tibialis posterior (Mercuri score 0–1).

Figure 3. Boxplots of NSAA scores and 6MWT distances by T1w Mercuri score in 9 lower limb muscle. The boxplots illustrate the correlation between T1w MRI findings in each muscle and functional outcome measures evaluated within ± 3 month before or after performing the MRI. The nine left hand panels (A–C,G–I, and M–O) correspond to NSAA correlations, while the nine right hand panels (D,E,J,K, and P–R) correspond to 6MWT correlations. Upper row panels correspond to pelvic girdle and anterior thigh muscles: gluteus maximus (A and D), vastus lateralis (B and E), and adductors (C and F). Middle row panels correspond to posterior thigh muscles: biceps femoris (G and J), semitendinosus (H and K), and semimembranosus (I and L). Lower row panels correspond to leg muscles: tibialis anterior (M and P), soleus (M and Q), and gastrocnemius medialis (O and R). Correlation parameters (Spearman correlation coefficient “rho” and p-value) are indicated below each graph.

Figure 4. Boxplots of NSAA score changes and 6MWT distance changes after 1 year by T1w Mercuri score in 9 lower limb muscle. The boxplots illustrate the correlation between T1w MRI findings in each muscle and functional outcome measure change, twelve months after performing the MRI and obtaining baseline functional measures. The nine left hand panels (A–C,G–I, and M–O) correspond to NSAA changes, while the nine right hand panels (D,E,J,K, and P–R) correspond to 6MWT changes correlations. Upper row panels correspond to pelvic girdle and anterior thigh muscles: gluteus maximus (A and D), vastus lateralis (B and E), and adductors (C and F). Middle row panels correspond to posterior thigh muscles: biceps femoris (G and J), semitendinosus (H and K), and semimembranosus, (I and L). Lower row panels correspond to leg muscles: tibialis anterior (M and P), soleus (M and Q), and gastrocnemius medialis (O and R). Correlation parameters (Spearman correlation coefficient “rho” and p-value) are indicated below each graph. The “zero” value on the y axis, corresponding to no change in functional measures, is indicated by a dashed line.

Figure 5. Mosaic plots of functional status in subgroups with different fatty substitution of lower limb muscles. Each bar corresponds to a group of participants with a certain fatty substitution score in the muscle of interest. Bar width is proportional to group size. Bars are subdivided vertically in colored areas representing different functional status. Darker shades of grey correspond to better functional status as shown in the legend at the bottom. Upper row panels correspond to pelvic girdle and anterior thigh muscles: gluteus maximus (A), vastus lateralis (B), and adductors (C). Middle row panels correspond to posterior thigh muscles: biceps femoris (D), semimembranosus (E), and semitendinosus (F). Lower row panels correspond to leg muscles: tibialis anterior (G), soleus (H), and gastrocnemius medialis (I).

Figure 6. Beeswarm plots of baseline and 1-year functional measures. Beeswarm plots of (A) 6MWT distances and (B) NSAA scores at baseline (at the time of MRI) and after 1 year. Baseline and 1-year values for each individual participant are linked by a dashed segment. Composite T1w score is color-coded as in the heatmap in a green → yellow → red gradient from 0% to 100% (i.e. total T1w score 0 to 88 out of a possible 88, as a sum of 22 lower limb muscles).