ORIGINAL ARTICLE Journal of Cachexia, Sarcopenia and Muscle 2022; 13: 2888–2897 Published online 4 September 2022 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/jcsm.13063 Water T2 could predict functional decline in patients with dysferlinopathy Ursula Moore1, Ericky Caldas de Almeida Araújo2,3, Harmen Reyngoudt2,3, Heather Gordish-Dressman4,5, Fiona E. Smith6, Ian Wilson6, Meredith James1, Anna Mayhew1, Laura Rufibach7, John W. Day8, Kristi J. Jones9, Diana X. Bharucha-Goebel10,11, Emmanuelle Salort-Campana12, Alan Pestronk13, Maggie C. Walter14, Carmen Paradas15, Tanya Stojkovic16, Madoka Mori-Yoshimura17, Elena Bravver18‡, Elena Pegoraro19, Jerry R. Mendell20, The Jain COS Consortium, Kate Bushby1, Andrew M. Blamire6, Volker Straub1, Pierre G. Carlier21 & Jordi Diaz-Manera1,22,23* 1 The John Walton Muscular Dystrophy Research Centre, Translational and Clinical Research Institute, Newcastle University and Newcastle Hospitals NHS Foundation Trust, Newcastle upon Tyne, UK; 2NMR Laboratory, Neuromuscular Investigation Center, Institute of Myology, Paris, France; 3NMR Laboratory, CEA/DRF/IBFJ/MIRCen, Paris, France; 4Center for Translational Science, Division of Biostatistics and Study Methodology, Children’s National Health System, Washington, DC, USA; 5Pediatrics, Epidemiology and Biostatistics, George Washington University, Washington, DC, USA; 6Magnetic Resonance Centre, Translational and Clinical Research Institute, Newcastle University, Newcastle upon Tyne, UK; 7Jain Foundation, Seattle, WA, USA; 8Department of Neurology and Neurological Sciences, Stanford University School of Medicine, Stanford, CA, USA; 9The Children’s Hospital at Westmead and The University of Sydney, Sydney, NSW, Australia; 10Department of Neurology, Children’s National Health System, Washington, DC, USA; 11National Institutes of Health (NINDS), Bethesda, MD, USA; 12Service des maladies neuromusculaire et de la SLA, Hôpital de La Timone, Marseille, France; 13Department of Neurology, Washington University School of Medicine, St. Louis, MO, USA; 14Friedrich-Baur-Institute, Department of Neurology, Ludwig- Maximilians-University of Munich, Munich, Germany; 15Neuromuscular Unit, Department of Neurology, Hospital U. Virgen del Rocío/Instituto de Biomedicina de Sevilla, Sevilla, Spain; 16Centre de référence des maladies neuromusculaires, Institut de Myologie, AP-HP, Sorbonne Université, Hôpital Pitié-Salpêtrière, Paris, France; 17Department of Neurology, National Center Hospital, National Center of Neurology and Psychiatry, Tokyo, Japan; 18Neuroscience Institute, Carolinas Neuromuscular/ALS-MDA Center, Carolinas HealthCare System, Charlotte, NC, USA; 19Department of Neuroscience, University of Padova, Padua, Italy; 20The Abigail Wexner Research Institute at Nationwide Children’s Hospital, Columbus, OH, USA; 21Université Paris-Saclay, CEA, DRF, Service Hospitalier Frederic Joliot, Orsay, France; 22Neuromuscular Disorders Unit, Neurology Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 23Centro de Investigación Biomédica en Red en Enfermedades Raras (CIBERER), Madrid, Spain Abstract Background Water T2 (T2H2O) mapping is increasingly being used in muscular dystrophies to assess active muscle damage. It has been suggested as a surrogate outcome measure for clinical trials. Here, we investigated the prognostic utility of T2H2O to identify changes in muscle function over time in limb girdle muscular dystrophies. Methods Patients with genetically confirmed dysferlinopathy were assessed as part of the Jain Foundation Clinical Outcomes Study in dysferlinopathy. The cohort included 18 patients from two sites, both equipped with 3-tesla mag- netic resonance imaging (MRI) systems from the same vendor. T2H2O value was defined as higher or lower than the median in each muscle bilaterally. The degree of deterioration on four functional tests over 3 years was assessed in a linear model against covariates of high or low T2H2O at baseline, age, disease duration, and baseline function. Results A higher T2H2O at baseline significantly correlated with a greater decline on functional tests in 21 out of 35 muscles and was never associated with slower decline. Higher baseline T2H2O in adductor magnus, vastus intermedius, vastus lateralis, and vastus medialis were the most sensitive, being associated bilaterally with greater decline in multi- ple timed tests. Patients with a higher than median baseline T2H2O (>40.6 ms) in the right vastus medialis deteriorated 11 points more on the North Star Ambulatory Assessment for Dysferlinopathy and lost an additional 86 m on the 6-min walk than those with a lower T2H2O (<40.6 ms). Optimum sensitivity and specificity thresholds for predicting decline were 39.0 ms in adductor magnus and vastus intermedius, 40.0 ms in vastus medialis, and 40.5 ms in vastus lateralis from different sites equipped with different MRI systems. Conclusions In dysferlinopathy, T2H2O did not correlate with current functional ability. However, T2H2O at baseline was higher in patients who worsened more rapidly on functional tests. This suggests that inter-patient differences in functional decline over time may be, in part, explained by different severities of the active muscle damage, assessed by T2H2O measure at baseline. Significant challenges remain in standardizing T2H2O values across sites to allow © 2022 The Authors. Journal of Cachexia, Sarcopenia and Muscle published by John Wiley & Sons Ltd on behalf of Society on Sarcopenia, Cachexia and Wasting Disorders. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 1353921906009, 2022, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13063 by Readcube (Labtiva Inc.), Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Water T2 is a predictor of change in muscle function 2889 determining globally applicable thresholds. The results from the present work are encouraging and suggest that T2H2O could be used to improve prognostication, patient selection, and disease modelling for clinical trials. Keywords Magnetic resonance imaging; Water T2; Limb girdle muscular dystrophy; Limb girdle muscular dystrophy R2; Limb girdle muscular dystrophy 2B Received: 4 February 2022; Revised: 18 May 2022; Accepted: 4 July 2022 *Correspondence to: Professor Jordi Diaz-Manera, The John Walton Muscular Dystrophy Research Centre, Translational and Clinical Research Institute, Faculty of Medical Sciences, Newcastle University, Centre for Life, Newcastle, UK. Email: jordi.diaz-manera@newcastle.ac.uk Ursula Moore, Ericky Caldas de Almeida Araujo, Pierre G. Carlier, and Jordi Diaz-Manera equally contributed to the work. † Deceased. Introduction methods such as Dixon-based fat-fraction (FF) mapping, which provides an objective measurement of the amount of Predicting functional decline in muscular dystrophies is a dif- fat present in the skeletal muscles, have been demonstrated ficult task. There are many paediatric and adult onset forms to correlate with muscle function and to capture changes in of muscular dystrophy and they display highly variable rates muscle structure over short periods of time. Such methods of disease progression, yet few clues to the cause of the var- are being implemented in natural history studies and also iability have been identified.1 In some diseases, particularly clinical trials.15–17 However, its role as a predictor of changes the more rapidly progressive Duchenne muscular dystrophy in muscle function has not been demonstrated so far. (DMD), identifying current functional ability may suggest T2 mapping sequences have also been used to study the next function to be lost, leading to a predictable set of changes in muscle structure in several neuromuscular dis- disease milestones.2,3 Biomarkers that correlate with current eases (NMDs). Skeletal muscle T2 is elevated in the presence function can therefore predict the next step in disease pro- of oedema, inflammation, and necrosis, as a consequence of gression in DMD.4 However, in the slowly progressive limb increased water mobility and disrupted tissue ultrastructure. girdle muscular dystrophies (LGMD), predicting short-term However, fatty replacement is a common pathway in the dis- functional changes presents more of a challenge. ease progression of most NMDs, which also results in in- Dysferlinopathy is a form of LGMD, most commonly de- creased T2 values, because the T2 of fat is much longer than scribed as LGMDR2 or Miyoshi myopathy (MMD1).5,6 In this that of muscle.15,18 In order to assess the current status of disease, from the same functional starting point, one patient the residual skeletal muscle tissue, independently from the ir- may remain stable for 3 years, while another deteriorates sig- reversible end-stage fatty replacement, water T2 (T2H2O) nificantly over only 1 year.7,8 mapping methods have been developed, allowing the detec- Being able to predict progression of muscular dystrophies tion of inflammatory, oedematous, and necrotic features in has several advantages. Patients may benefit from clearer ex- fatty-infiltrated muscles.19,20 pectations about the future and more tailored care. More- T2H2O is increased in several NMDs and its capacity to pre- over, for clinical trials aiming to demonstrate an effective in- dict changes in muscle structure, mainly the fat replacement tervention, it is important to have a well-matched cohort, or that follows muscle fibre loss, has been reported in several at least to understand the differences in anticipated disease diseases.21–24 Recent data from the Clinical Outcomes Study progression without intervention.9,10 As interventional thera- (COS) baseline visit suggest that T2H2O correlates with fatty pies become a reality for the muscular dystrophies, the need replacement over time in dysferlinopathy patients, confirm- to identify biomarkers able to predict upcoming functional ing the capacity of T2H2O to identify active damage leading decline has intensified.11,12 to muscle fibre loss and expansion of fat tissue.25 However, Magnetic resonance imaging (MRI) has been proposed as the capacity of T2H2O to identify patients who will experience one such biomarker.4,11–13 The most used sequences in clinics a quicker and more severe clinical progression over a short are both qualitative and include T1-weighted, which detects period of time remains to be demonstrated. If it could be fatty replacement of the muscles, and fat-suppressed T2- shown to predict progression, T2H2O could be used both in weighted imaging, which signal is increased in several dis- clinics to identify patients at risk of progression and also in eases being usually related to oedema and inflammation, al- clinical trials to select patients who may deteriorate over a though these can be masked by the presence of fatty infiltra- short period of time and whose results could be more infor- tions due to the fat suppression.14 Although these sequences mative about the effectivity of experimental therapies. have been demonstrated to be useful for the diagnosis of pa- We hypothesized that inter-patient variation in active mus- tients with neuromuscular diseases, their interpretation can cle damage, assessed by T2H2O, may underlie and predict sub- be biased and they lack reproducibility, which hampers their sequent differences in disease progression from the same application in longitudinal follow-up studies designed over functional starting point in patients with dysferlinopathy. In short periods of time.15 In contrast, quantitative MRI this paper, we assess a series of quantitative MRI parameters, Journal of Cachexia, Sarcopenia and Muscle 2022; 13: 2888–2897 DOI: 10.1002/jcsm.13063 1353921906009, 2022, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13063 by Readcube (Labtiva Inc.), Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2890 U. Moore et al. including FF, contractile cross-sectional area (cCSA), and ambulant patients who completed a set of four functional as- T2H2O value, in each of the lower limb muscles and how they sessments including the North Star Assessment for limb relate to subsequent disease progression in a range of func- girdle-type muscular dystrophies [North Star Ambulatory As- tional and timed tests in patients with dysferlinopathy, to de- sessment for Dysferlinopathy (NSAD)], the 6-min walk test termine whether any of these parameters can be effectively (smwt), the timed up and go (TUG) test, and the 10-m walk used to predict muscle function loss. test (10MWT) at baseline, Year 1, and Year 3. The cohort in- cluded a total of 18 patients who had an NSAD score of more than 15 points (Figure 1). Methods Patients Quantitative muscle magnetic resonance imaging: acquisition and processing of Dixon and T2H2O Patients were selected from the Jain Foundation COS in imaging data dysferlinopathy.26 COS is an international, multicentric, pro- spective natural history study involving 15 centres in the Patients were imaged using 3.0-tesla MRI clinical scanners USA, Europe, Asia, and Oceania. COS recruited and assessed from two different vendors: Newcastle (Philips) and Paris 182 patients over a 3-year period. Assessments performed in- (Siemens). MRI acquisition parameters were standardized cluded medical, physical therapist, biochemical, cardiovascu- across sites before the start of the study. Patients were posi- lar, and respiratory and MRI that were performed at baseline tioned feet-first supine, and all MRI sequences were centred visit and then every year until Year 3. All patients had a diag- at one-third of the femur from the superior border of the pa- nosis of dysferlinopathy, confirmed by genetic testing, with tella and at the widest part of the calf. The acquisition proto- two pathogenic mutations or one pathogenic mutation and col for FF and T2H2O mapping in the COS study has been re- evidence of reduced dysferlin expression on a western blot cently reported.25 of skeletal muscle and/or monocytes. The baseline, 1-year Water and fat images were reconstructed using in-house data, and 3-year functional progression and quantitative Matlab code (MathWorks, Natick, MA, USA), which incorpo- muscle MRI analysis have previously been published.7,8,25 rated hierarchical IDEAL (iterative decomposition of water We selected a cohort of patients seen at Newcastle and and fat with echo asymmetry and least-squares estimation) Paris to limit the impact of inter-site variability. These two and the Tsao–Jiang algorithm for separating multiple chemi- sites had the largest number of patients with MRI results cal species by hierarchical decomposition and direct estima- and have previously been shown to have high inter-site reli- tion of phase.27 Using the mean-square error (MSE) images, ability. In order to identify if quantitative MRI measurement quantitative T2H2O maps were reconstructed based on a could predict changes in muscle function, we only included tri-exponential fitting procedure.19,21 Regions of interest Figure 1 Patients included in this study from the original COS study. COS, Clinical Outcomes Study. Journal of Cachexia, Sarcopenia and Muscle 2022; 13: 2888–2897 DOI: 10.1002/jcsm.13063 1353921906009, 2022, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13063 by Readcube (Labtiva Inc.), Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Water T2 is a predictor of change in muscle function 2891 (ROIs) were drawn manually using a free software tool each side in turn (4 functional assessments × 32 muscles— (https://www.itksnap.org) in the same five central slices on left and right). This method was repeated with FF and the shortest echo time (TE) image of the MSE series images cCSA data. For each muscle, the change in FF and cCSA by a single investigator as has been already described.25 ROIs over 3 years was modelled against the baseline T2H2O value, were drawn on both the left and right sides for seven leg with disease duration, age, and baseline FF and cCSA as muscles (extensor digitorum, tibialis anterior, tibialis poste- covariates. rior, peroneus longus, soleus, gastrocnemius medialis, and Differences were considered significant if the P-value was gastrocnemius lateralis) and in nine thigh muscles (vastus lower than 0.05. The P-values were adjusted using the lateralis, vastus intermedius, vastus medialis, gracilis, sarto- Holm–Bonferroni correction to assess statistical significance rius, adductor magnus, biceps femoris long head, in the presence of multiple comparisons. semimembranosus, and semitendinosus). For the determina- tion of FF and cross-sectional area (CSA), the boundaries of the ROIs were drawn following individual muscle delineation, Identifying a T2H2O threshold that identifies higher changes in avoiding inclusion of other muscles, subcutaneous and North Star Ambulatory Assessment for Dysferlinopathy than intermuscular fat, tendons, and major blood vessels. cCSA expected was calculated for each muscle using cCSA = (1 FFmean) * CSA. Previous studies modelling the COS cohort of patients over For the assessment of T2H2O, ROIs delineated the interior of time have demonstrated that NSAD score declines by 1.68 the muscle, avoiding visible fasciae and blood vessels. ROIs points per year, and therefore, over 3 years, a decline of more that included <10 pixels were excluded for analysis. than 5.04 points identifies those with faster than average After quality control, FF and cCSA calculation was not pos- functional decline.8 Each patient was classed as having a de- sible in all slices for all patients. This resulted in some slices cline of more than 5 points in NSAD score (positive result) or for which there was a T2H2O value but not an FF and a cCSA <5 points (negative result). We calculated the cut-off point in value. Assessment of correlations between T2H2O and FF T2H2O with the higher sensitivity and specificity to distinguish and cCSA used only those data sets where all variables were between patients progressing more and less than 5 points in available (n numbers shown in Table 1). NSAD using receiver operating characteristic curves (ROC curves) using R software (https://www.r-project.org). We in- cluded in this analysis only those muscles whose baseline Data analysis T2H2O correlated with changes in muscle function tests from baseline to the last visit assessment. Higher or lower than median T2H2O The median T2H2O of each muscle in each patient was catego- rized as either (a) greater than or equal to the cohort median T2H2O value for that muscle (high T2H2O) or (b) less than the cohort median T2H2O value for that muscle (low T2H2O). The Results change in functional score (ΔNSAD) or change in velocity in the timed tests (Δsmwt, ΔTUG, and Δ10MWT) over 3 years Demographics was calculated using assessments from baseline and Year 3. Demographic information was reviewed, including age, The cohort consisted of 18 ambulant patients (7 male) disease duration, and baseline functional assessment score. assessed in Newcastle (12 patients) or Paris (6 patients). Pa- Demographic data and baseline functional assessment results tients had a median age of 32.5 years (range 19–71 years) were compared between high and low T2H2O values in each and had had symptoms for a median of 11 years (range 3– muscle. Median values were compared between groups using 22 years). Demographic information split by T2H2O value per a Mann–Whitney test. muscle is listed in Supporting Information, Table S1. For each functional assessment, a linear model of the change in functional score over 3 years was performed with high or low baseline T2H2O as a predictor and disease dura- tion, age, and baseline functional score as covariates. Timed T2H2O values vary between muscles tests were assessed as velocities to give normally distributed values. This model produces an estimate of the additional Median T2H2O values varied from a minimum of 35.9 ms in change in functional score seen in those with a high T2H2O gastrocnemius medialis (left) to a maximum of 42.7 ms in compared with those with a low T2H2O value, when disease vastus lateralis (right) (Table 1). Within each muscle, T2H2O duration, age, and baseline functional score are held con- was not significantly different between left and right stant. The analysis was performed for each functional assess- (Wilcoxon’s test). T2H2O did not correlate with functional ment using the T2H2O value from each of the 16 muscles on score, FF, or cCSA at baseline (Table S2). Journal of Cachexia, Sarcopenia and Muscle 2022; 13: 2888–2897 DOI: 10.1002/jcsm.13063 1353921906009, 2022, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13063 by Readcube (Labtiva Inc.), Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2892 U. Moore et al. Table 1 Correlation between T2H2O and changes in muscle function, fat fraction, and contractile cross-sectional area over 3 years Δ6MWT Δ6MWT ΔTUG Δ10MWT N with Muscle [left (L)/ Median T2H2O velocity distance rate velocity Dixon ΔFF ΔcCSA 2 right (R)] value (range) ΔNSAD in m/s in m (event/s) (m/s) data (%) (mm ) Vastus intermedius (L) 40.0 (34.3–46.4 8* 0.18* 64.8* 0.04** 0.06* 16 9.4 54.7 Vastus intermedius (R) 40.6 (35.8–47.2) 11** 0.14 50.4 0.04** 0.06** 16 7.8 181.1 Vastus lateralis (L) 41.4 (27.3–48.0) 9** 0.10 36.0 0.03* 0.05* 16 8.4 551.0 Vastus lateralis (R) 42.7 (35.2–51.4) 6 0.08 28.8 0.03* 0.05* 16 10.3* 606.1 Vastus medialis (L) 39.5 (26.9–50.1) 8* 0.20* 72.0* 0.02 0.05* 16 9.8** 315.9 Vastus medialis (R) 42.0 (33.1–48.8) 11** 0.24** 86.4** 0.04*** 0.07** 15 9.7** 441.4 Adductor magnus (L) 39.3 (33.1–45.8) 7* 0.14 50.4 0.03** 0.04 16 15.5 40.7 Adductor magnus (R) 39.0 (33.6–47.2 10** 0.18* 64.8* 0.04*** 0.07** 16 13.3 156.9 Gracilis (L) 38.5 (33.4–43.0) 4 0.15 54.0 0.03** 0.03 16 9.1 123.6 Gracilis (R) 37.5 (33.0–44.3) 6 0.17* 61.2* 0.03** 0.03 16 3.5 141.9 Sartorius (L) 38.4 (32.0–43.9) 4 0.16* 57.6* 0.03** 0 16 12.0 142.6 Sartorius (R) 39.3 (34.5–44.8) 8** 0.11 39.6 0.03** 0.03 16 7.9 88.1 Biceps femoris (L) 39.6 (26.4–51.5) 2 0.05 18.0 0.02 0.02 15 5.9 162.9 Biceps femoris (R) 38.7 (27.7–50.4) 9** 0.16 57.6 0.03* 0.02 14 13.6 181.4 Semimembranosus (L) 38.3 (25.7–47.8) 7* 0.13 46.8 0.03 0.02 13 13.1 113.1 Semimembranosus (R) 38.1 (20.3–46.2) 3 0.08 28.8 0.01 0.02 13 10.0* 168.7 Semitendinosus (L) 36.7 (28.8–44.6) 5 0.01 3.6 0.01 0.01 16 7.1 57.8 Semitendinosus (R) 34.8 (22.6–47) 4 0.09 32.4 0.01 0.01 16 9.4 35.8 Tibialis anterior (L) 39.4 (23.9–49.4) 7* 0.14 50.4 0.02 0 15 4.0 41.1 Tibialis anterior (R) 38.6 (31.1–53.3) 2 0.04 14.4 0.01 0 15 2.4 21.0 Extensor digitorum (L) 40.5 (34.2–50) 8* 0.11 39.6 0.03** 0.01 15 7.7* 50.1 Extensor digitorum (R) 39.6 (34.8–48) 2 0.03 10.8 0.02 0.01 15 11.3 95.3 Peroneus (L) 37.1 (23.9–50) 2 0.12 43.2 0.01 0.01 15 14.5 159.6 Peroneus (R) 36.4 (25.2–48.2) 3 0.11 39.6 0.02 0.02 14 13.2 184.5 Tibialis posterior (L) 40.2 (34.1–50.3) 2 0.11 39.6 0.03* 0.01 15 6.9 133.8 Tibialis posterior (R) 39.9 (34.8–48.3) 4 0.06 21.6 0.02 0.02 15 8.4 129.9 Soleus (L) 37.3 (23.6–50.4) 1 0.09 32.4 0.02 0.01 14 6.1 14.9 Soleus (R) 38.1 (26.9–49.4) 9* 0.26** 93.6** 0.04** 0.03 15 10.9 40.5 Gastrocnemius lateralis (L) 36.9 (25.8–46.2) 5 0.03 10.8 0.01 0.02 14 14.1 151.1 Gastrocnemius lateralis (R) 37.7 (23.1–47.6) 1 0.06 21.6 0 0 13 15.8 34.0 Gastrocnemius medialis (L) 35.9 (26.2–57) 3 0.07 25.2 0.02 0.05* 14 3.8 95.1 Gastrocnemius medialis (R) 38.5 (26.1–60.7) 4 0.06 21.6 0.02 0 14 0.5 73.8 N, number; Δ6MWT, change in 6-min walk test between Year 3 and baseline; Δ10MWT, change in 10-m walk test between Year 3 and baseline; ΔcCSA, change in contractile cross-sectional area between Year 3 and baseline; ΔFF, change in fat fraction between Year 3 and baseline; ΔNSAD, change in North Star Ambulatory Assessment for Dysferlinopathy between Year 3 and baseline; ΔTUG, change in timed up and go test between Year 3 and baseline. The table shows mean changes in muscle function tests and their association with high T2H2O. Regression coefficients in bold were sta- tistically significant at the 0.05 level. * P < 0.05. ** P < 0.01. *** P < 0.001. High T2H2O value was significantly correlated with a change in cCSA bilaterally in any muscle after 3 years of functional deterioration follow-up (Table 1). Median T2H2O values at baseline significantly correlated with a worsening for at least one functional assessment over the Determining a T2H2O value for prediction next 3 years in 19 of the 32 muscles studied (Table 1). Mus- cles with high T2H2O were always associated with faster func- We studied the T2H2O value thresholds that maximized sensi- tional deterioration. tivity and specificity for predicting decline in function using High T2H2O in adductor magnus and vastus intermedius/ ROC curves of the adductor magnus, vastus intermedius, lateralis/medialis muscles stood out as being consistently vastus lateralis, and vastus medialis. The T2H2O value thresh- and bilaterally significantly correlated with functional pro- olds obtained were 39.0 ms [88% sensitive and 75% specific, gression, demonstrating a greater deterioration in at least area under the curve (AUC) 0.847 (0.634–1)] in adductor two functional assessments bilaterally (Table 1). magnus, 39.4 ms [100% sensitive, 60% specific, AUC 0.917 In vastus lateralis (bilateral), vastus medialis (right), (0.782–1)] in vastus intermedius, 40.5 ms [94% sensitive, semimembranosus (left), and extensor digitorum (left), me- 70% specific, AUC 0.889 (0.734–1)] in vastus lateralis, and dian T2H2O values significantly correlated with an increase in 40.1 ms [94% sensitive, 40% specific, AUC 0.875 (0.699–1)] FF over 3 years. Median T2H2O values did not correlate with in vastus medialis (Figure 2). Journal of Cachexia, Sarcopenia and Muscle 2022; 13: 2888–2897 DOI: 10.1002/jcsm.13063 1353921906009, 2022, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13063 by Readcube (Labtiva Inc.), Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Water T2 is a predictor of change in muscle function 2893 Figure 2 Receiver operating characteristic curve of the sensitivity and specificity achieved by all possible T2H2O relaxation times. The area under the curve is listed for each plot. In each curve, the threshold with the highest specificity (with zero false positives) and the threshold with optimum sen- sitivity and specificity are listed. Table 2 Sensitivity and specificity of a T2H2O threshold in predicting decline in North Star Ambulatory Assessment for Dysferlinopathy (NSAD) score Decline of >1 point Decline of >5 points Muscle Threshold Analysis on NSAD in 1 year on NSAD in 3 years Adductor magnus 39.0 ms Sensitivity 57% 63% Specificity 71% 78% Vastus intermedius 39.4 ms Sensitivity 86% 100% Specificity 57% 67% Vastus lateralis 40.5 ms Sensitivity 71% 88% Specificity 71% 78% Vastus medialis 40.1 ms Sensitivity 86% 88% Specificity 57% 67% Muscle grouping All four muscles over Respective threshold Sensitivity 43% 63% threshold bilaterally for each muscle as Specificity 71% 89% At least one muscle over above Sensitivity 86% 100% threshold bilaterally Specificity 57% 56% Validating T2H2O thresholds values weere not improved by requiring all four muscles to be over their indiviudal thresholds or requring only one Seventeen of the 18 patients had bilateral T2H2O results for of the four muscles to be over their individual theshold adductor magnus, vastus intermedius, vastus lateralis, and (Table 2). vastus medialis and could be included in this analysis. The sensitivity of predicting decline on the NSAD score with the T2H2O thresholds over both 1 and 3 years is shown in Table Discussion 2 and Figure 3. As shown in Table 2, sensitivity and specificity of the measures suggested that those with high T2H2O values We have shown that a higher T2H2O in some muscles of the were generally rapid progressors. Sensitivity and specificity lower limbs might be associated with greater functional dete- Journal of Cachexia, Sarcopenia and Muscle 2022; 13: 2888–2897 DOI: 10.1002/jcsm.13063 1353921906009, 2022, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13063 by Readcube (Labtiva Inc.), Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2894 U. Moore et al. rioration in patients with dysferlinopathy. These findings However, not all the muscles imaged showed an associa- could be particularly helpful in advising patients about prog- tion between higher T2H2O value and risk of more severe dis- nosis, in selecting patients for clinical trials, and, potentially, ease progression. The muscles in which T2H2O did not corre- as an outcome measure for interventional trials. late with functional decline were those of the posterior There is a large body of evidence already published dem- compartment of the thigh and muscles of the leg, apart from onstrating that the amount of fat present in the skeletal mus- the soleus. There are several possible reasons for this: cle, which can be quantified using quantitative MRI (Dixon method) or proton magnetic resonance spectroscopy (1H 1. Due to the functional outcome measures chosen: The dis- MRS), correlates with muscle function tests and is sensitive tal muscle that did show involvement was the soleus, to change over short periods of time.17,28 However, little is which is heavily involved in walking, while the gastrocne- known about the relation of T2H2O, which may be an indicator mius and peroneus are less involved when walking on an of active muscle damage and muscle function. It has been ob- even surface.31 Regarding the posterior compartment of served that in dysferlinopathy, the T2H2O does not correlate the thigh, walking can be maintained, even in the pres- with baseline functional scores.25 T2H2O has been shown to ence of significant weakness, by using compensatory be impacted by high levels of FF (i.e. >60%) as also described mechanisms, which may mean that active muscle damage in dysferlinopathy.25 The data described in the current study, in these muscles does not translate to impairment on however, were acquired in exclusively ambulant patients, these specific outcome measures.32 where FF values were predominantly lower than 60%. This 2. Due to a lower degree of active muscle damage in these suggests that T2H2O could be compared across patients at dif- muscles at the time of assessment: The pattern of muscle ferent stages of muscle loss to offer a snapshot of the current involvement in dysferlinopathy includes an early and se- level of disease activity. vere involvement of the soleus, both the gastrocnemius Higher than threshold baseline T2H2O value predicted and the peroneus.13 The first functional ability to be lost greater decline over both 1 and 3 years, being consistently is usually the ability to stand on tiptoes, a function highly more accurate over the 3-year window. This suggests that reliant on the gastrocnemius and soleus muscles.33 The change in function is a delayed downstream consequence T2H2O values seen in these muscles were lower than those of higher active muscle damage as quantified by the T2H2O. in the vasti muscles, and these muscles showed a higher This also reflects the slowly progressive nature of the disease, fat replacement by the time of assessment as they are af- which results generally in small changes in function over fected very early during the disease’s progression.13 1 year but more consistently over a longer period of 3 years 3. Due to the higher FF in these muscles: FFs in the posterior of follow-up. Despite this, even over a 1-year period, T2H2O compartment of the thigh and the gastrocnemius thresholds were able to predict functional decline, which were among the highest of all the muscles assessed (Table could make T2H2O particularly useful for detecting patients S2). T2H2O is much more heterogeneous in more that will probably progress significantly during a clinical trial fatty-replaced muscles, as has been shown.34 In this with a shorter running time. sense, investigation in nine patients with more advanced We found a correlation between T2H2O at baseline and dysferlinopathy and higher FFs found lower T2H2O in pa- changes in FF in some of the investigated muscles, tients compared with healthy controls.35 These data sug- including the vastus medialis and the vastus intermedius. gest that at those stages, T2H2O would probably not be Correlation between muscle histopathology and quantita- useful to identify active muscle damage due to the mas- tive MRI in patients with muscular dystrophy demonstrates sive replacement by fat.34 histopathological features of inflammation appearing in the muscle, detectable by T2H2O, before fat replacement The association of T2H2O with functional test outcomes in begins.29 In our population, it appears that active muscle specific muscles demonstrates the importance of selecting damage, detected by higher T2H2O, correlated with a func- an appropriate ‘reporter’ muscle that is related to the func- tional decline in some of the muscles affected. Given the tional outcome of interest and stage of disease progression. progressive nature of dysferlinopathy, we may anticipate In this analysis, the muscles most associated with the timed that with a longer period of follow-up, this increased dis- tests and NSAD score were the vasti and the adductor ease activity would ultimately translate into irreversibly in- magnus muscles. However, we may also anticipate that glu- creased FF and loss of cCSA. These findings suggest that teal muscles, hip extensors, and abductors, which are impor- T2H2O is a useful measure to demonstrate active muscle tant in rising from the floor and walking, may constitute damage in remaining muscle in dysferlinopathy and to pro- useful reporter muscles.31 Unfortunately, the COS study did vide an accurate quantification of potentially salvageable not include T2H2O data from the gluteal muscles and so we tissue. This could be useful for monitoring response to were unable to investigate this further. As disease progresses, future treatments, with T2H2O forming a potential early useful reporter muscles may change to those with lower FFs. biomarker of treatment efficacy.30 In non-ambulant patients, where outcome measures focus Journal of Cachexia, Sarcopenia and Muscle 2022; 13: 2888–2897 DOI: 10.1002/jcsm.13063 1353921906009, 2022, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13063 by Readcube (Labtiva Inc.), Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Water T2 is a predictor of change in muscle function 2895 more on the upper limbs, which are generally preserved for acquired on 1.5-tesla scanners.36 However, this scaling factor longer in dysferlinopathy, one of the arm muscles may be a is only an approximation that has been obtained in a few better reporter muscle.10,13 We hope to address this in fu- number of healthy controls and likely still leaves significant ture, after completion of the clinical outcome extension inter-scanner variability reinforcing the need of further study, which includes upper limb MRI. research in this specific topic. In addition, besides We felt that it would be clinically useful for prognostica- hardware-related differences that affect T2, variations in tion or trial cohort selection if a specific threshold could be patient status, such as recent exercise activity, are known to defined above which T2H2O is considered ‘high’ and patients impact on the observed T2H2O.15 In this regard, patients may be expected to progress more rapidly. We tried to vali- participating in the COS study were imaged before physio date these thresholds in a second cohort of dysferlin patients assessments were performed and were asked to avoid doing participating in COS study that were scanned in 3- and 1.5- any exercise the week before the visit. Most of the patients tesla scanners at six other sites, but even though higher used a taxi to go to the hospital or imaging centre, in order T2H2O levels were associated with faster progression, the ac- to reduce the impact that the exercise could have in the curacy, sensitivity, and specificity were lower as it is shown MRI values, and were also asked to lie down before being in the supporting information. There are many reasons that scanned. The data obtained in our extension cohort could explain these results including, among others, the dif- (supporting information results) unfortunately do not allow ferent field strengths, which is known to affect the T2; the us to take any final conclusion about what could be the different vendor-specific sequence details such as radio fre- causes of the lack of validation of the thresholds obtained, in- quency pulses and crusher-gradient schemes, which impact dicating that more research is needed in this topic. In this the measured magnetic resonance (MR) signal and hence sense, we think that clinical or clinical trial-based applications the observed T2; and the transmitter and receiver coils that attempting to use T2H2O for predicting disease progression vary between vendors and even between system versions across multiple sites would require additional research to from the same vendor, which also impact the measured MR identify the optimum standardization of scan protocol, signal. While the scan protocols were standardized and the T2H2O mapping, and normalization of data between sites be- same post-processing methodology was used for all T2H2O fore starting the trial.37 data and was conducted by the same team, normalization This study was conducted in patients with dysferlinopathy. across sites by means of control data from the same volun- However, the findings may be applicable to other slowly pro- teers in all sites was not practicable.15,16 We also attempted gressive forms of muscular dystrophy, which show periods of to correct for the difference in field strength by applying both deterioration and stability. Muscular dystrophies share a a correction factor based on published data to the results common pathomechanism of repeated cycles of muscle dam- Figure 3 Disease progression in the cohort analysed, grouped by T2H2O threshold value identified for (A) adductor magnus (AM), (B) vastus interme- dius (VI), (C) vastus lateralis (VL), and (D) vastus medialis (VM). Dots showing the North Star Ambulatory Assessment for Dysferlinopathy (NSAD) score at each time point are grouped by dashed lines to illustrate individual patient trajectories on the NSAD score over 3 years. Those deteriorating more than 5 points over 3 years are coloured blue, and those deteriorating 5 or less points are coloured red. Box plots represent the median, interquartile range, and range of NSAD score for patients at baseline (left) and Year 3 (right). Journal of Cachexia, Sarcopenia and Muscle 2022; 13: 2888–2897 DOI: 10.1002/jcsm.13063 1353921906009, 2022, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13063 by Readcube (Labtiva Inc.), Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2896 U. Moore et al. age and attempted repair, with release of pro-inflammatory treatments for dysferlinopathy. Please visit www.jain-founda- cytokines and immune cell influx.38–41 Such patterns of in- tion.org for more information about the foundation, and if flammation and oedema are detected by T2H2O, suggesting you are a patient suffering from dysferlinopathy, please con- that our finding may be extrapolated to other muscular dys- sider enrolling into their interactive dysferlinopathy registry trophies, although this needs to be confirmed in future that seeks to build a strong, engaged, and supportive commu- studies.29,37 nity (patients@jain-foundation.org). We also acknowledge The results presented in this paper suggest that faster the help of Pierre-Yves Baudin (NMR Laboratory, Institute of functional disease progression in dysferlinopathy can be pre- Myology, Paris, France) for assistance with MRI data dicted by higher T2H2O values in muscles of the thigh, over 1 processing. and 3 years. With further research, we predict that similar patterns would be demonstrated in other slowly progressive forms of muscular dystrophy. However, it is important to take Conflict of interest into account that an effort in standardizing the measure- ments between centres is crucial if these observations are Ursula Moore, Ericky Caldas de Almeida Araujo, Harmen to be employed in clinical trial design and prognostication Reyngoudt, Heather Gordish-Dressman, Fiona E. Smith, Ian for patients with muscular dystrophy. Wilson, Meredith James, Anna Mayhew, Jown W. Day, Kristi J. Jones, Diana X. Bhraucha Goebel, Emmanuelle Salort Campana, Alan Pestronk, Maggie Walter, Carmen Paradas, Acknowledgements Tanya Stojkovic, Madoka Mori-Yoshimura, Elene Bravver, Elena Pegoraro, Jerry R. Mendel, Kate Bushby, Andrew M. This study has only been possible thanks to the international Blamire, Volker Straub, Pierre Carlier, and Jordi Díaz-Manera collaboration of several specialized centres promoted by the have received funding for research from the Jain Foundation Jain Foundation. The Jain COS consortium would like to thank during the conduct of the study. the study participants and their families for their invaluable Laura Rufibach is hired by the Jain Foundation. contribution and would also like to acknowledge the ongoing support the Jain Foundation provides in the development, management, and analysis of this study. The Jain Foundation, Online supplementary material based in Seattle, USA, is entirely focused on LGMD2B/ dysferlinopathy/Miyoshi myopathy. The foundation does Additional supporting information may be found online in the not solicit funding from patients but instead funds research Supporting Information section at the end of the article. and clinical studies worldwide with the goal of finding References 1. Winckler PB, da Silva AMS, Coimbra-Neto 5. Bashlr R, Strachan T, Keers S, Stephenson phies and implications for exon skipping AR, Carvalho E, Cavalcanti EBU, Sobreira A, Mahjneh I, Marconi G, et al. A gene for therapies. Acta Myol 2020;39:179–186. CFR, et al. Clinicogenetic lessons from 370 autosomal recessive limb-girdle muscular 10. Patino CM, Ferreira JC. Inclusion and exclu- patients with autosomal recessive dystrophy maps to chromosome 2p. Hum sion criteria in research studies: definitions limb-girdle muscular dystrophy. Clin Genet Mol Genet 1994;3:455–457. and why they matter. J Bras Pneumol 2018; 2019;96:341–353. 6. Miyoshi K, Kawai H, Iwasa M, Kusaka K, 44:84. 2. Brooke MH, Fenichel GM, Griggs RC, Nishino H. Autosomal recessive distal mus- 11. Dahlqvist JR, Widholm P, Leinhard OD, Mendell JR, Moxley R, Florence J, et al. cular dystrophy as a new type of progres- Vissing J. MRI in neuromuscular diseases: Duchenne muscular dystrophy: patterns sive muscular dystrophy. Seventeen cases an emerging diagnostic tool and biomarker of clinical progression and effects of sup- in eight families including an autopsied for prognosis and efficacy. Ann Neurol portive therapy. Neurology 1989;39: case. Brain 1986;109:31–54. 2020;88:669–681. 475–481. 7. Moore U, Jacobs M, James MK, Mayhew 12. Mul K, Vincenten SCC, Voermans NC, 3. Mayhew AG, Cano SJ, Scott E, Eagle M, AG, Fernandez-Torron R, Feng J, et al. As- Lemmers RJLF, van der Vliet PJ, van der Bushby K, Manzur A, et al. Detecting mean- sessment of disease progression in Maarel SM, et al. Adding quantitative mus- ingful change using the North Star Ambula- dysferlinopathy: a 1-year cohort study. cle MRI to the FSHD clinical trial toolbox. tory Assessment in Duchenne muscular Neurology 2019;92:e461–e474. Neurology 2017;89:2057–2065. dystrophy. Dev Med Child Neurol 2013;55: 8. Jacobs MB, James MK, Lowes LP, Alfano 13. Diaz-Manera J, Fernandez-Torron R, 1046–1052. LN, Eagle M, Muni Lofra R, et al. Assessing LLauger J, James MK, Mayhew A, Smith 4. Barnard AM, Willcocks RJ, Triplett WT, dysferlinopathy patients over three years FE, et al. Muscle MRI in patients with Forbes SC, Daniels MJ, Chakraborty S, with a new motor scale. Ann Neurol 2021; dysferlinopathy: pattern recognition et al. MR biomarkers predict clinical func- 89:967–978. and implications for clinical trials. tion in Duchenne muscular dystrophy. Neu- 9. Hoffman EP. Causes of clinical variability in J Neurol Neurosurg Psychiatry 2018;89: rology 2020;94:e897–e909. Duchenne and Becker muscular dystro- 1071–1081. Journal of Cachexia, Sarcopenia and Muscle 2022; 13: 2888–2897 DOI: 10.1002/jcsm.13063 1353921906009, 2022, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jcsm.13063 by Readcube (Labtiva Inc.), Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Water T2 is a predictor of change in muscle function 2897 14. Nunez-Peralta C, Alonso-Perez J, Diaz- TTJ, et al. Selection approach to identify strategies for muscle weakness during gait Manera J. The increasing role of muscle the optimal biomarker using quantitative by simulating activation deficits seen MRI to monitor changes over time in un- muscle MRI and functional assessments in post-stroke. Gait Posture 2013;38: treated and treated muscle diseases. Curr Becker muscular dystrophy. Neurology 270–275. Opin Neurol 2020;33:611–620. 2021;97:e513–e522. 33. Moore U, Gordish H, Diaz-Manera J, James 15. Carlier PG, Marty B, Scheidegger O, 24. Hooijmans MT, Niks EH, Burakiewicz J, MK, Mayhew AG, Guglieri M, et al. Miyoshi Loureiro de Sousa P, Baudin PY, Snezhko Verschuuren JJ, Webb AG, Kan HE. Elevated myopathy and limb girdle muscular dystro- E, et al. Skeletal muscle quantitative nu- phosphodiester and T2 levels can be mea- phy R2 are the same disease. Neuromuscul clear magnetic resonance imaging and sured in the absence of fat infiltration in Disord 2021;31:265–280. spectroscopy as an outcome measure for Duchenne muscular dystrophy patients. 34. Schlaeger S, Weidlich D, Klupp E, clinical trials. J Neuromuscul Dis 2016;3: NMR Biomed 2017;30:e3667. Montagnese F, Deschauer M, Schoser B, 1–28. 25. Reyngoudt H, Smith FE, Caldas de Almeida et al. Decreased water T2 in fatty infil- 16. Burakiewicz J, Sinclair CDJ, Fischer D, Araújo E, Wilson I, Fernández-Torrón R, trated skeletal muscles of patients with Walter GA, Kan HE, Hollingsworth KG. James MK, et al. Three-year quantitative neuromuscular diseases. NMR Biomed Quantifying fat replacement of muscle by magnetic resonance imaging and phospho- 2019;32:e4111. quantitative MRI in muscular dystrophy. J rus magnetic resonance spectroscopy 35. Arrigoni F, de Luca A, Velardo D, Magri F, Neurol 2017;264:2053–2067. study in lower limb muscle in Gandossini S, Russo A, et al. 17. Rooney WD, Berlow YA, Triplett WT, dysferlinopathy. J Cachexia Sarcopenia Multiparametric quantitative MRI assess- Forbes SC, Willcocks RJ, Wang DJ, et al. Muscle 2022;13:1850–1863. ment of thigh muscles in limb-girdle mus- Modeling disease trajectory in Duchenne 26. Harris E, Bladen CL, Mayhew A, James M, cular dystrophy 2A and 2B. Muscle Nerve muscular dystrophy. Neurology 2020;94: Bettinson K, Moore U, et al. The Clinical 2018;58:550–558. e1622–e1633. Outcome Study for dysferlinopathy: an in- 36. Gold GE, Han E, Stainsby J, Wright G, 18. Carlier PG. Global T2 versus water T2 ternational multicenter study. Neurol Brittain J, Beaulieu C. Musculoskeletal in NMR imaging of fatty infiltrated Genet 2016;2:e89. MRI at 3.0 T: relaxation times and image muscles: different methodology, different 27. Tsao J, Jiang Y. Hierarchical IDEAL: fast, ro- contrast. AJR Am J Roentgenol 2004;183: information and different implications. bust, and multiresolution separation of 343–351. Neuromuscul Disord 2014;24:390–392. multiple chemical species from multiple 37. Araujo ECA, Marty B, Carlier PG, Baudin PY, 19. Azzabou N, Loureiro de Sousa P, Caldas E, echo times. Magn Reson Med 2013;70: Reyngoudt H. Multiexponential analysis of Carlier PG. Validation of a generic 155–159. the water T2-relaxation in the skeletal approach to muscle water T2 determina- 28. Spanish Pompe group, Figueroa-Bonaparte muscle provides distinct markers of disease tion at 3T in fat-infiltrated skeletal S, Llauger J, Segovia S, Belmonte I, activity between inflammatory and dystro- muscle. J Magn Reson Imaging 2015;41: Pedrosa I, et al. Quantitative muscle phic myopathies. J Magn Reson Imaging 645–653. MRI to follow up late onset Pompe 2021;53:181–189. 20. Keene KR, Beenakker JM, Hooijmans MT, patients: a prospective study. Sci Rep 38. Fanin M, Angelini C. Muscle pathology in Naarding KJ, Niks EH, Otto LAM, et al. T2 2018;8:10898. dysferlin deficiency. Neuropathol Appl relaxation-time mapping in healthy and 29. Lassche S, Küsters B, Heerschap A, Schyns Neurobiol 2002;28:461–470. diseased skeletal muscle using extended MVP, Ottenheijm CAC, Voermans NC, 39. Rosenberg AS, Puig M, Nagaraju K, phase graph algorithms. Magn Reson Med et al. Correlation between quantitative Hoffman EP, Villalta SA, Rao VA, et al. 2020;84:2656–2670. MRI and muscle histopathology in muscle Immune-mediated pathology in Duchenne 21. Carlier PG, Azzabou N, de Sousa PL, Hicks biopsies from healthy controls and patients muscular dystrophy. Sci Transl Med 2015; A, Boisserie JM, Amadon A, et al. Skeletal with IBM, FSHD and OPMD. J Neuromuscul 7:299rv294. muscle quantitative nuclear magnetic reso- Dis 2020;7:495–504. 40. Angelini C, Nardetto L, Borsato C, Padoan nance imaging follow-up of adult Pompe 30. Arpan I, Willcocks RJ, Forbes SC, Finkel RS, R, Fanin M, Nascimbeni AC, et al. The clin- patients. J Inherit Metab Dis 2015;38: Lott DJ, Rooney WD, et al. Examination of ical course of calpainopathy (LGMD2A) and 565–572. effects of corticosteroids on skeletal dysferlinopathy (LGMD2B). Neurol Res 22. Morrow JM, Sinclair CD, Fischmann A, muscles of boys with DMD using MRI and 2010;32:41–46. Machado PM, Reilly MM, Yousry TA, et al. MRS. Neurology 2014;83:974–980. 41. Tasca G, Monforte M, Corbi M, Granata G, MRI biomarker assessment of neuromus- 31. Bogey RA, Barnes LA. Estimates of individ- Lucchetti D, Sgambato A, et al. Muscle cular disease progression: a prospective ual muscle power production in normal microdialysis to investigate inflammatory observational cohort study. Lancet Neurol adult walking. J Neuroeng Rehabil 2017; biomarkers in facioscapulohumeral muscu- 2016;15:65–77. 14:92. lar dystrophy. Mol Neurobiol 2018;55: 23. van de Velde NM, Hooijmans MT, Sardjoe 32. Knarr BA, Reisman DS, Binder-Macleod SA, 2959–2966. Mishre ASD, Keene KR, Koeks Z, Veeger Higginson JS. Understanding compensatory Journal of Cachexia, Sarcopenia and Muscle 2022; 13: 2888–2897 DOI: 10.1002/jcsm.13063