Quantitative muscle MRI captures early muscle degeneration in calpainopathy

Study population

In total, 19 individuals with genetically and/or histologically confirmed calpainopathy (LGMD R1/D4) (10 females, aged 25–69 years; mean age 39.8 years, 9 males; aged 26–56 years; mean age 36.2 years) and 19 age- and gender-matched healthy volunteers (10 females, aged 26–60 years; mean age 39.2 ± 12.6 years) participated in this study. Clinical information of calpainopathy patients is given in Table 1. In all individuals with only a mono-allelic mutation in the genetic testing, diagnosis of calpainopathy was proven by myopathic changes in muscle biopsy specimen and reduced calpain-protein abundance (investigated by immunoblotting). Exclusion criteria for healthy volunteers were medical history of neuromuscular diseases (NMD) and injuries in lower extremity within 12 months prior to study enrolment.

Table 1 Demographic and clinical data of patients with calpainopathy.

The institutional and committee for approving the experiments was the ethics committee of the Faculty for Medicine of the Ruhr University Bochum (Ruhr University Bochum No. 15-5281). The study was performed in accordance with the guidelines of the ethics committee and line with the Declaration of Helsinki (DVH), the relevant national and international recommendations according to Good Clinical Practice (GCP), §7 Health Professions Act North Rhine-Westphalia and §15 of the Professional Code of the Medical Association of Westphalia-Lip. Informed consent was obtained from all participants.

Clinical assessments

Muscle strength was evaluated using both, the Medical Research Council (MRC) scale and Quick Motor Function Test (QMFT) by an experienced clinician (JF: 5 years of experience, RR: 8 years of experience, AKG: 16 years of experience)24. Hip flexion, knee flexion, knee extension, ankle dorsiflexion, and ankle plantarflexion were assessed with a Chatillon Dynamometer® (Chatillon DFE II Dynamometer, Chatillon Force Measurement, AMETEK, USA) for force measurements. Daily life activities were enquired using the ACTIVLIM and the NSS24,25. An experienced medical technical assistant performed the following tests in ambulant individuals: the 6-MWT, time to walk 10 m, and timed up- and-go test.

MRI acquisition and sequences

The participants laid in a feet-first supine position. Cushions were used to support participants’ knees and sandbags placed around their feet to prevent motion. Scans were obtained using a Philips 3.0T Achieva MR system and a 16CH Torso XL coil. The thigh region from hip to knee was split into two axial fields of view (FOV) of 480 × 276 × 150 mm3 along the z-axis with a 30 mm overlap and the proximal edge positioned in the crotch. The calf region was scanned with one axial FOV of 480 × 276 × 150 mm3. The proximal edge of the FOV was positioned 60 mm below the tibial plateau perpendicular to the tibial bone. The protocol consisted of a 4-point Dixon sequence (voxel size 1.5 × 1.5 × 6.0 mm3; TR/TE 210/2.6, 3.36, 4.12, 4.88 ms; flip angle 8°, SENSE: 2), a multi‐echo spin‐echo (MESE) sequence for quantitative water mapping including 17 echoes and Cartesian k‐space sampling (voxel size 3.0 × 3.0 × 6.0 mm3; TR/TE 4598/17 × ∆7.6 ms; flip angle 90/180°, SENSE: 2), and a diffusion-weighted spin-echo EPI (voxel size 3.0 × 3.0 × 6.0 mm3; TR/TE 5000/57 ms; SPAIR/SPIR fat suppression; SENSE: 1.9; 42 gradient directions with eight different b-values (0–600 s/mm2)26. A noise scan was obtained using the same imaging parameters as the DWI, but without RF power and gradients (only acquisition channels open). Scanning time per stack (each FOV) was approximately 12 min.

Data pre-processing

Data were pre-processed as described before using QMRITools ( In brief, the diffusion data were denoised using a PCA method27. To correct for subject motion and eddy currents both legs were registered separately. Then the tensors were calculated by taking IVIM into account and using an iWLLS algorithm27,28. The IDEAL method was used for the Dixon data considering a single T2* decay and resulting in a separated water and fat map30. The derived water maps were used for the manual segmentation. The T2‐mapping data were analysed using an extended phase graph (EPG) fitting approach31.

Muscle segmentation and tractography

Eight thigh muscles (vastus lateralis, vastus medialis, rectus femoris, semimembranosus, semitendinosus, biceps femoris, sartorius, and gracilis) and seven calf muscles (gastrocnemius medialis and lateralis, soleus, tibialis anterior, peroneus, extensor digitorum and tibialis posterior) were first segmented in patients and controls using an automated segmentation tool and subsequently optimized by an experienced rater (JF) in both legs32. The rater checked the automated segmentation results and manually corrected the muscle shape if necessary. Automated segmentation was not precise enough in patients due to loss of muscle structure in fatty infiltrated muscles. Patient muscles were segmented manually on all 25 acquired slices of the Dixon water images (3D-slicer 4.4.0,

The segmentations were then registered to T2 and DTI data to correct for small motions between sequences and image distortions using sequential rigid and b-spline transformations (elastix, https://elastix.lumc.nl34. Average values over all slices of water-T2 time and proton density fat fraction (FF) were obtained. SNR was calculated as the local average signal divided by the local noise sigma35. For analysis of diffusion data, the pre-processed diffusion data were divided based on the muscle segmentation. Secondly, whole muscle tractography was performed within each diffusion muscle volume. (MRIToolkit)36. The following fiber tracking stop parameters were used: maximum angle 15°, step size 1.5 mm, FA range 0.1–0.635,36. The DTI parameters fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (λ1) and radial diffusivity (RD) were extracted for each individual muscle using tract-based sampling.

Statistical analysis

FF was compared between LGMD patients and healthy controls in a general linear model with patient/control, body side, and muscle (to control degrees of freedom for multiple test points per subject) as fixed factors for all leg muscles.

For the analysis of the diffusion metrics and T2 times, we chose a disease-specific approach in terms of muscle involvement, based on the degree of muscle fat infiltration shown by individual muscle FF in this study and literature review18. We chose to analyse qMRI changes in different groups of muscles that have a different risk of degeneration and not in a patient/control approach.

We identified the three following muscle groups with different risks of fat infiltration in calpainopathies20:

  1. I.

    high risk: biceps femoris, semimembranosus, semitendinosus, gastrocnemius medialis and soleus

  2. II.

    intermediate risk: vastus medialis, vastus lateralis, rectus femoris, sartorius, gracilis, gastrocnemius lateralis and peroneal group

  3. III.

    low risk: extensor digitorum, tibialis anterior and tibialis posterior

To assess differences in diffusion metrics and T2 time between low-fat muscles of controls and LGMD patients, the cut-off value for low-fat muscles was defined as: highest mean fat fraction in healthy controls + 1 SD (~ 8%)17.

Additionally, muscles with low diffusion data quality, which was defined by a SNR of lower than 10 before denoising, were excluded for analysis of diffusion parameters39. After performing Levene’s test for equality of variances, two-sided t-tests for independent samples between LGMD and control group were conducted to evaluate the following hypotheses for muscles with < 8% FF:

  1. I.

    In high-risk muscles we hypothesize that the known fiber atrophy and myocellular damage in calpainopathies40 could lead to an increase of FA, a decrease of MD and RD and an increase in T2.

  2. II.

    In intermediate-risk muscles, there may be significant differences in T2 and diffusion metrics between the patient and control group.

  3. III.

    In low-risk muscles no significant changes of T2 and diffusion metrics between study groups can be detected.

To evaluate correlations between clinical assessments and qMRI values mean FF, FA, MD, and T2 of all thigh and calf muscles were correlated to the 6-MWT, time to walk 10 m, and timed up-and-go test using Pearson’s correlation coefficients. Furthermore, mean qMRI values of thighs and calves were correlated to the results of ACTIVLIM, NSS, and QMFM using Spearman’s rank correlation coefficients. Additionally, the qMRI values of the corresponding muscle groups were correlated to the force measurements by dynamometry and MRC scale, i.e., quadriceps muscle and knee extension.

All statistical analyses were performed using IBM SPSS V28. The significance level for all tests was set at p < 0.05.

Source link

Back to top button