0344

Quantitative MRI of skeletal muscle in a cross-sectional cohort of spinal muscular atrophy type 2 and type 3
Louise A.M. Otto1, Ludo W.L. van der Pol1, Lara Schlaffke 2, Camiel A. Wijngaarde1, Marloes Stam1, Renske I. Wadman1, Inge Cuppen3, Ruben P.A. van Eijk1,4, Fay-Lynn Asselman1, Bart Bartels5, Danny van der Woude5, Jeroen Hendrikse6, and Martijn Froeling6
1Neurology, UMC Utrecht Brain Center, University Medical Center, Utrecht, Utrecht, Netherlands, 2Neurology, BG-University Hospital Bergmannsheil, Ruhr-University Bochum, Bochum, Germany, 3Neurology and Child Neurology, UMC Utrecht Brain Center, University Medical Center, Utrecht, Utrecht, Netherlands, 4Biostatistics & Research Support, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, Netherlands, 5Child Development and Exercise Center, UMC Utrecht Brain Center, University Medical Center, Utrecht, Utrecht, Netherlands, 6Radiology, UMC Utrecht Brain Center, University Medical Center, Utrecht, Utrecht, Netherlands

Synopsis

qMRI of skeletal muscle has shown promising results in other neuromuscular diseases, but multi-parametric imaging has not been executed in Spinal Muscular Atrophy. We investigated a cohort of 31 patients and 20 controls with protocol consisting of DIXON, T2 mapping and DTI on a 3T MR scanner. All parameters differed significantly between patients and controls. DTI elucidates distinct properties of the muscle, suggesting atrophy by a lowered MD and increased FA. DTI shows correlation with muscle strength and motor function. This suggests the potential of diffusion tensor imaging of muscle in monitoring disease progression in SMA.

Background

Hereditary proximal Spinal Muscular Atrophy (SMA) is caused by Survival Motor Neuron (SMN) protein deficiency due to homozygous loss of function of the SMN1 gene.1 Treatment strategies have been developed to skew splicing of the remaining, paralogous SMN2 gene towards full length SMN protein, which is critical to alpha-motor neurons. In the absence of therapy, SMA exhibits a slowly progressive decline in muscle strength and motor function in later-onset patients.2 Outcome measures that can appreciate treatment effects of existing and future therapies are still warranted. Quantitative MRI (qMRI) of muscle has shown promising results in other neuromuscular disorders, such as Duchenne muscular dystrophy (DMD), limb girdle muscular dystrophies (LGMD), glycogen storage disorders, i.e. McArdle’s and Pompe’s disease.3–6 Data on various qMRI measures of skeletal muscle in a large cohort of SMA patients encompassing a broad range of disease severity and disease duration are still lacking.

Methods

We included 20 control subjects and 31 patients with SMA type 2 (classification when highest acquired motor milestone was independent sitting) and type 3 (highest acquired motor milestone was walking independently, with symptom onset before 3 years – 3a or after – 3b), demographics are presented in figure 1.
MR acquisition
MR datasets of both legs were acquired on a 3T MR scanner (Philips Ingenia, Philips Medical Systems, the Netherlands) in supine position, with a 12-channel posterior and 16 channel anterior body coil. Images were acquired with a FOV of 15cm and aligned with the femur, starting 17,5cm from the femoral head or centered mid-femoral in case of severe contractures. The MR scanning protocol was ~10 minutes and has shown reproducibility in a previous multicenter study7 and comprised: i. 4-point DIXON (TR/TE/ 210/2.6/3.36/4.12/4.88 ms; flip angle 10°; voxel size 1x1.5x1.5; no gap; 25 slices) ii. T2 mapping (17 echoes TR/TE/ΔTE 4598/17/7.6; flip angle 90/180°; voxel size 1x3x3; slice gap 6mm; 13 slices, no fat suppression) iii. SE-EPI (TR/TE 5000/57 ms; b-value 0 s/mm2; voxel size 1x3x3; no gap; 25 slices, 8 dynamics, SPAIR and SPIR fat suppression).

MR processing
All MR data were processed using QMRITools for Mathematica (mfroeling.github.io/QMRITools)8. The processing steps for each method are summarized in figure 2. Before processing all data was visually inspected for artefacts and data quality. Manual segmentation was done based on the DIXON out-phase and fat images with open-source software (ITK-SNAP9), and transferred to the T2 and DTI data with rigid and b-spline image registration (Elastix10). Muscles with volume smaller than 10 voxels (=67.5mm3) were omitted from analyses. Since the T2 and the DTI parameters have a bias with increasing fat contribution4,11, simulations were performed to estimate this effect (figure 4).4

Statistics
Differences between patients and controls were assessed by means of an independent Student’s t-test with Welch’s correction. MR parameters were averaged across muscles per subject. Due to non-normality of the data, associations between clinical and imaging parameters were evaluated using Kendall’s tau correlation coefficient. Significance was set at p<0.05.

Results

Two datasets were excluded because of image quality, resulting in 49 datasets for analysis.
Mean fat fraction of all upper leg muscles was 7.6±1.5% in controls and 47.6±17.4% in patients (p<.001). Mean T2, averaged over all muscles was 27.3±1.5 in patients and 28.9±0.4 in controls (p<.001). MD is lowered in patients (1.13±.28) versus controls (1.47±.10, p<.001). FA is higher (.41±.09) in patients compared to controls (.24±.03, p<.001). The decrease of MD and increase of FA are greater than the predicted effect of an increase of partial volume effects of fat, as illustrated by the simulation experiments. Our data shows that T2 of fat infiltrated muscles follow the predicted T2 decrease due to fat partial volume effects. Overall, we observed that the estimated EPG fat fraction is different from estimated DIXON fat fraction. Fat fraction, FA and MD strongly correlate with muscle strength and motor function: fractional anisotropy is negatively associated with score on the Hammersmith Functional Motor Scale, Expanded and Medical Research Council sum score (τ =-.56 and -.59; both p<.001) whereas for fat fraction values are τ =-.50 and .58, respectively (both p<.001).

Discussion

All qMRI measures differed significantly between patients and controls. FA and MD manifest stronger effects than can be accounted for the effect of fatty replacement partial volume effect. DTI findings indirectly indicate cell atrophy and act as a measure independently of fat fraction which is in accordance with known pathology. Lastly, DTI parameters show correlation with muscle strength and motor function. This suggests the potential of diffusion tensor imaging of muscle in monitoring disease progression and to study pathogenesis of muscle in SMA. From this ongoing study longitudinal data of one-year follow-up in a subset of 10 patients without treatment and 6 children on treatment will determine the potential of qMRI as a biomarker for disease progression or treatment effect. This data is currently being acquired and under analysis.

Conclusion

qMRI measures of skeletal muscle in a cohort of patients with SMA varying in type, age and disease severity exhibit potential in monitoring disease progression.

Acknowledgements

We thank all SMA patients, their families and control subjects for their participation. We thank Christa van Ekris for her assistance. This work was supported by the Prinses Beatrix Spierfonds (Grant no. W.OR16-06). The Dutch SMA register is supported by stichting Spieren voor Spieren.

References

1. Lefebvre S, Bürglen L, Reboullet S, et al. Identification and characterization of a spinal muscular atrophy-determining gene. Cell. 1995;80:155-165. doi:10.1016/0092-8674(95)90460-3

2. Mercuri E, Finkel R, Montes J, et al. Patterns of disease progression in type 2 and 3 SMA: Implications for clinical trials. Neuromuscul Disord. 2016;26(2):126-131. doi:10.1016/j.nmd.2015.10.006

3. Rehmann R, Schlaffke L, Froeling M, et al. Muscle diffusion tensor imaging in glycogen storage disease V (McArdle disease). Eur Radiol. 2019;29:3224-3232. doi:10.1007/s00330-018-5885-1

4. Hooijmans MT, Damon BM, Froeling M, et al. Evaluation of skeletal muscle DTI in patients with duchenne muscular dystrophy. NMR Biomed. 2015;28:1589-1597. doi:10.1002/nbm.3427

5. Willis TA, Hollingsworth KG, Coombs A, et al. Quantitative Muscle MRI as an Assessment Tool for Monitoring Disease Progression in LGMD2I: A Multicentre Longitudinal Study. PLoS One. 2013;8. doi:10.1371/journal.pone.0070993

6. Carlier PG, Azzabou N, de Sousa PL, et al. Skeletal muscle quantitative nuclear magnetic resonance imaging follow-up of adult Pompe patients. J Inherit Metab Dis. 2015;38:565-572. doi:10.1007/s10545-015-9825-9

7. Schlaffke L, Rehmann R, Rohm M, et al. Multicenter evaluation of stability and reproducibility of quantitative MRI measures in healthy calf muscles. NMR Biomed. 2019:1-14. doi:10.1002/nbm.4119

8. Froeling M. QMRTools: a Mathematica toolbox for quantitative MRI analysis. J Open Source Softw. 2019. doi:10.21105/joss.01204

9. Yushkevich PA, Piven J, Hazlett HC, et al. User-guided 3D active contour segmentation of anatomical structures: Significantly improved efficiency and reliability. Neuroimage. 2006;31(3):1116-1128. doi:10.1016/j.neuroimage.2006.01.015

10. Klein S, Staring M, Murphy K, Viergever MA, Pluim JPW. Elastix: A toolbox for intensity-based medical image registration. IEEE Trans Med Imaging. 2010;29:196-205. doi:10.1109/TMI.2009.2035616

11. Williams SE, Heemskerk AM, Welch EB, Li K, Damon BM, Park JH. Quantitative effects of inclusion of fat on muscle diffusion tensor MRI measurements. J Magn Reson Imaging. 2013;38:1292-1297. doi:10.1002/jmri.24045

Figures

Clinical characteristics of study participants. n= number, M=male, F=female, SD= standard deviation, MRC = Medical Research Council, HFMSE = Hammersmith Functional Motor Scale, Expanded

Overview of the various steps involved the acquisition and processing pipeline. A) example data for a healthy volunteer and a SMA patient for all three qMRI methods. B) A summary of the processing steps and example parameters obtained for each of the qMRI methods. C) Overview of the muscle segmentation. Muscle are manually segmented using the out-phase Dixon image. Then, segmentation is transferred to the DTI and T2 mapping data using b-spline registration.

Overview of fat infiltration of the thigh stratified for SMA type and age. Images of the thigh in the transversal plane of the middle section of the image stack. The images are categorized into age category for distinct SMA types. The heat bar beneath indicates a scale from 0 to 100%. The right column presents a magnification of the left leg of type 3b patients. The red arrows point at the annotated muscle, that seem relatively spared. NA= non-ambulant, A= ambulant, RF= rectus femoris; GR= gracilis; BFS= biceps femoris, short head.

Left: boxplots for qMRI parameters for the distinct muscle groups for healthy controls, SMA type 2 and type 3. Middle: qMRI parameters plotted against fat fraction; each data point represents a measurement of individual muscles of all subjects with SMA (grey) and control subjects (green). Right: datapoints reduced to an average using local regression and 95% CI (shaded area). Fat fraction is simulated from 0 to 100% fat with T2 of fat set at 180ms and T2 of water in fat at 20ms. For MD and FA, the water component of fat is assumed to have isotropic diffusion (FA=0) with an MD of 0.8 mm2/s.

Each of the qMRI parameters (columns) is plotted against the following clinical outcome measures(columns); disease duration (in months), HFMSE score, MRC sum score and MRC sum score of the upper leg. The correlation formula, Kendall’s tau correlation coefficient and the p-value (significance level set at <.05) are shown per correlation plot. HFMSE= Hammersmith Motor Function Scale Expanded, MRC = Medical Research Council.

Proc. Intl. Soc. Mag. Reson. Med. 28 (2020)
0344