Myotonic Dystrophy type 1 (DM1) is an inherited muscular dystrophy with a prevalence of approximately 10 per 100.000. On MR images of DM1 affected muscles fatty infiltration and edema-like processes can be identified^1,2,3,4. These studies used only a semi-quantitative analysis^1,2,3 or evaluated only a single muscle in a quantitative way⁴. In other muscular dystrophies the use of quantitative MRI to evaluate fatty infiltration and edema-like processes have been found to be extremely useful in the understanding of the disease^5,6,7. Quantitative MRI is also essential in the proper evaluation of the natural progression of the disease and necessary to monitor the effect of therapeutic interventions.

The aim of this study was to evaluate the extent of fatty infiltration and edema-like processes in muscles of DM1 patients by quantitative MR.

Patients: 10 genetically proven DM1 patients (7 male, age 50±8 years, MIRS score⁸ 3.1±0.9).

MR protocol: Patients were examined with a 3T Siemens TIM Trio and a ¹H spine coil combined with phased array coils placed around the lower leg. To evaluate the extent of fatty infiltration, fat fractions (FF) were determined using a 2pt-DIXON sequence (TR/TE1/TE2: 10/2.45/3.675 ms, FA: 3°, voxel size: 1x1x5 mm, slices: 32) or a 3pt-DIXON (TE1/TE2/TE3: 2.31/3.68/5.07 ms). To quantify edema-like processes multi spin-echo (MSE) images were obtained (TR: 3720 ms, ETL: 17, ES: 8 ms, voxel size: 1.5x1.5x10 mm, slice gap: 20 mm, slices: 5).

Data analysis: Firstly, the T2 relaxation time of muscle water spins (T2_water) was mapped by fitting the MSE data on a voxel-wise basis using a bi-component extended phase graph (EPG) model⁹: SI(TE) = A×FF_EPG×EPG(T1_fat, T2_fat, ES, B1, α₁,…,α_ETL) + A×(1-FF_EPG) ×EPG(T1_water, T2_water, ES, B1, α₁,…,α_ETL), with T1_water = 1400 ms, T1_fat = 365 ms and α the refocusing angle. T2_fat was determined per patient by applying a mono-component EPG model on subcutaneous fat. Since sufficient water signal is needed to obtain a reliable T2_water, voxels with a fitted FF_EPG>50% were assumed to be unreliable and thus excluded for further analysis. Secondly, regions of interest were drawn around the calf muscles on the 5 slices of the DIXON images that corresponded in slice position with the T2map. Thirdly, the average FF and T2_water for each muscle were determined per slice and as an average of all 5 slices. For the intermediately fat infiltrated muscles (FF of at least one slice > 20% or < 50%) the FF in the most distal and proximal slice were compared (two-sided paired t-test).

The average FF distribution plot of the calf muscles has a pyramid shape (Fig. 1A), demonstrating that muscles with a higher FF have a lower prevalence than muscles with a lower FF. The FF is largest in the gastronemicus medialis and soleus, and lowest in the tibialis posterior (Fig. 1B). Furthermore, it is observed that the FF is significantly higher in the distal slice compared to the proximal slice (p<0.01) (Fig. 2). Finally, we investigated the relation between T2_water and FF for each muscle (Fig. 3). Muscles with a FF below 20% (blue dots) have a T2_water comparable to that of healthy muscles (34.3 ms)⁹. Muscles who are becoming fat infiltrated (green squares, FF>20%) have a significantly elevated T2_water (38.2 ms) (two-sided t-test p<0.0001). Muscles in which more than 90% of the voxels were excluded (because FF_EPG>50%), show a large spread in T2_water.We investigated disease signatures in muscles of DM1 patients by quantitative MR. The inverse relationship observed between FF and the prevalence of muscles with that FF (pyramid-like shape) indicates that fatty infiltration in DM1 occurs in an, apparently slow, gradual manner. This is in contrast with facioscapulohumeral dystrophy (FSHD), where healthy muscles can become completely fat infiltrated within about 3 years⁵. It was observed that the dorsal muscles as the gastronemicus medialis and soleus are the most severely affected muscles, while the tibialis posterior is spared. This is in accordance with previous semi-quantitative studies^1,2,3. Furthermore, we observed that fatty infiltration is higher in the distal part of the muscles compared to the proximal part. This suggests that the disease process with fatty infiltration starts distally in DM1, which is also observed in affected leg muscles of FSHD patients⁵. To prove this hypothesis longitudinal studies are required. At last, we showed that muscles who are in an apparent active state of fatty infiltration have an elevated T2_water compared to muscles who are not in a fat infiltration process. This might indicate reactive edema, but the pathological mechanism causing this increased T2_water requires further investigation.