Three-year quantitative magnetic resonance imaging and phosphorus magnetic resonance spectroscopy study in lower limb muscle in dysferlinopathy

Abstract

Background: Natural history studies in neuromuscular disorders are vital to understand the disease evolution and to find sensitive outcome measures. We performed a longitudinal assessment of quantitative magnetic resonance imaging (MRI) and phosphorus magnetic resonance spectroscopy (31 P MRS) outcome measures and evaluated their relationship with function in lower limb skeletal muscle of dysferlinopathy patients.

Methods: Quantitative MRI/31 P MRS data were obtained at 3 T in two different sites in 54 patients and 12 controls, at baseline, and three annual follow-up visits. Fat fraction (FF), contractile cross-sectional area (cCSA), and muscle water T2 in both global leg and thigh segments and individual muscles and 31 P MRS indices in the anterior leg compartment were assessed. Analysis included comparisons between patients and controls, assessments of annual changes using a linear mixed model, standardized response means (SRM), and correlations between MRI and 31 P MRS markers and functional markers.

Results: Posterior muscles in thigh and leg showed the highest FF values. FF at baseline was highly heterogeneous across patients. In ambulant patients, median annual increases in global thigh and leg segment FF values were 4.1% and 3.0%, respectively (P < 0.001). After 3 years, global thigh and leg FF increases were 9.6% and 8.4%, respectively (P < 0.001). SRM values for global thigh FF were over 0.8 for all years. Vastus lateralis muscle showed the highest SRM values across all time points. cCSA decreased significantly after 3 years with median values of 11.0% and 12.8% in global thigh and global leg, respectively (P < 0.001). Water T2 values in ambulant patients were significantly increased, as compared with control values (P < 0.001). The highest water T2 values were found in the anterior part of thigh and leg. Almost all 31 P MRS indices were significantly different in patients as compared with controls (P < 0.006), except for pHw , and remained, similar as to water T2 , abnormal for the whole study duration. Global thigh water T2 at baseline was significantly correlated to the change in FF after 3 years (ρ = 0.52, P < 0.001). There was also a significant relationship between the change in functional score and change in FF after 3 years in ambulant patients (ρ = -0.55, P = 0.010).

Conclusions: This multi-centre study has shown that quantitative MRI/31 P MRS measurements in a heterogeneous group of dysferlinopathy patients can measure significant changes over the course of 3 years. These data can be used as reference values in view of future clinical trials in dysferlinopathy or comparisons with quantitative MRI/S data obtained in other limb-girdle muscular dystrophy subtypes.

Keywords: 31P MRS; Dysferlinopathy; Longitudinal; Outcome measures; Quantitative MRI.

Conflict of interest statement

A.M.B., A.M., B.M., E.C.A.A., F.E.S., H.R., I.W., J.M.B., J.L.L., K.B., L.R., M.J., M.K.J., N.A., P.G.C., U.M., and V.S. report grants from JAIN Foundation, during the conduct of the study. A.C., D.W., J.Y.H., J.D.M., L.W., R.F.T., T.S., and T.H. report no competing interests.

© 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.

Figures

Figure 1

Consort diagram. Missing MRI and 31P MRS data sets were due to medical reasons (x = 6), equipment failure (x = 3), study drop‐out following BL (x = 12), or Y2 (x = 5). Additionally, 31P MRS data were not acquired in those muscles that were highly or completely replaced by fat (x = 36), based on visual assessment of the images. Following processing of the image data, 9% of the acquired MRI data were omitted for final analysis due to failed image reconstructions (x = 18 of total of 190 data sets). More than half of the 31P MRS data were omitted for final analysis due to low SNR, which in most of the cases was due to high FF and/or low residual muscle in the anterior part of the leg. A median FF of 8% corresponded to the 31P MRS data that were kept for final analysis, in contrast to the median FF of 43% for those data sets that were omitted. ‘Validated’ means ‘included for final analysis’ following quality control (image series complete, artefact‐free, successful image reconstruction, sufficient signal‐to‐noise ratio). BL, baseline; n, number of patients; x, number of exams; Y1, Year 1; Y2, Year 2; Y3, Year 3.

Figure 2

Baseline FF in individual muscles and global segments. (A) Thigh baseline FF. (B) Leg baseline FF. For all individual muscles and global segments, there were significant differences between controls and ambulant patients, between controls and non‐ambulant patients, and between ambulant and non‐ambulant patients (all P < 0.001). More data on baseline FF can be found in TablesS2–S4. A, ambulant; AM, adductor magnus; BF, biceps femoris long head; ED, extensor digitorum longus; FF, fat fraction (in %); GL, gastrocnemius lateralis; GM, gastrocnemius medialis; GRA, gracilis; NA, non‐ambulant; PER, peroneus longus; SAR, sartorius; SM, semimembranosus; SOL, soleus; ST, semitendinosus; TA, tibialis anterior; TP, tibialis posterior; VI, vastus intermedius; VL, vastus lateralis; VM, vastus medialis.

Figure 3

Global FF and cCSA changes over time. (A) Global thigh FF. (B) Global thigh cCSA. (C) Global leg FF. (D) Global leg cCSA. (E) Global thigh FF trajectories for all individual patients as a function of age. (F) Global leg FF trajectories for all individual patients as a function of age. (G) Global thigh cCSA trajectories for all individual patients as a function of age. (H) Global leg cCSA trajectories for all individual patients as a function of age. Data for both controls, and ambulant and non‐ambulant patients are shown. The symbols in the trajectory plots represent the different phenotypes. The SRM values for annual and three‐year changes in FF and cCSA are depicted beneath the box‐and‐whisker plots (A) to (D). The horizontal dashed lines depict the 90th percentile value for FF (thigh: 10.4%, leg: 7.8%) or the 10th percentile value for cCSA (thigh: 87.6 cm2, leg: 48.7 cm2), as determined in controls. More data can be found in TablesS2–S4 for longitudinal FF and cCSA, and in Table1 and TableS5 for the LMM analyses. *P < 0.008, **P < 0.001 (between visits); #P < 0.008, ##P < 0.001 (between controls and patients); §P < 0.008, §§P < 0.001 (between ambulant and non‐ambulant patients). A, ambulant; BL, baseline; cCSA, contractile cross‐sectional area (in cm2); CTRL, controls; FF, fat fraction (in %); LGMD R2, limb‐girdle muscular dystrophy type R2; MM, Miyoshi myopathy; NA, non‐ambulant; SRM, standardized response mean; Y1, Year 1; Y2, Year 2; Y3, Year 3.

Figure 4

Plot depicting the increase in FF per individual muscle and per segment (thigh and leg) across the 3 years. Every FF value reflects the median in the ambulant study population. AM, adductor magnus; BF, biceps femoris long head; BL, baseline; ED, extensor digitorum longus; FF, fat fraction (in %); GL, gastrocnemius lateralis; GM, gastrocnemius medialis; GRA, gracilis; PER, peroneus longus; SAR, sartorius; SM, semimembranosus; SOL, soleus; ST, semitendinosus; TA, tibialis anterior; TP, tibialis posterior; VI, vastus intermedius; VL, vastus lateralis; VM, vastus medialis; Y1, Year 1; Y2, Year 2; Y3, Year 3.

Figure 5

Global water T2 and 31P MRS changes over time. (A) Global thigh water T2. (B) Global leg water T2. (C) Anterior leg pHw. (D) Anterior leg Pi,b/Pi,tot. (E) Anterior leg Pi,tot/PCr. (F) Anterior leg Pi,tot/γATP. (G) Anterior leg PCr/γATP. (H) Anterior leg PDE/γATP. (I) Anterior leg PME/γATP. Data for both controls, and ambulant and non‐ambulant patients are shown. The SDM and SRM values for water T2 and 31P MRS indices are depicted beneath the box‐and‐whisker plots. More data can be found in TablesS6–S8 for longitudinal water T2 and TableS10 for 31P MRS, and and TablesS9 and S11 for the LMM analyses. *P < 0.008 (water T2)/0.006 (31P MRS), **P < 0.001 (between visits); #P < 0.008/0.006, ##P < 0.001 (between controls and patients); §P < 0.008/0.006, §§P < 0.001 (between ambulant and non‐ambulant patients). A, ambulant; BL, baseline; CTRL, controls; NA, non‐ambulant; PCr, phosphocreatine; PDE, phosphodiesters; pHw, weighted pH; Pi,b, alkaline inorganic phosphate; Pi,tot, total inorganic phosphate; PME, phosphomonoesters; SDM, standardized difference mean; SRM, standardized response mean; Y1, Year 1; Y2, Year 2; Y3, Year 3; γATP, adenosine diphosphate (γ‐resonance in 31P MR spectrum).

Figure 6

Example of an FF map, a water T2 map (for thigh and leg), and a 31P MR spectrum (leg anterior part) in a patient across the 3 years (four visits) and in a healthy control subject. Notice the significant increase in FF in both thigh and leg (clearly visible in the anterior part of both segments) and the concomitant elevated water T2 values (bright as opposed to the control water T2 map). Values for annual global FF and global water T2 are also presented. Global FF values exceeding 10.4% (thigh) and 7.8% (leg), and a global water T2 exceeding 35.5 ms (thigh) and 37.9 ms (leg) are considered as abnormal (i.e. 90th percentiles determined in control subjects). In the four‐patient 31P MR spectra, obtained in the anterior part of the right leg, we notice a clear splitting of the inorganic phosphate resonance (black arrows), an increased PDE resonance (white arrows), and a decreased PCr resonance (grey arrows), which are all quasi‐unchanged across the 3 years, as compared with the healthy control spectrum (where we observe no significant splitting of the inorganic phosphate resonance and a smaller PDE resonance). FF, fat fraction (in %); PCr, phosphocreatine; PDE, phosphodiesters; Y1, Year 1; Y2, Year 2; Y3, Year 3.

Figure 7

Correlations between MRI, 31P MRS, and clinical variables. (A) Relationship between baseline global thigh water T2 and global change in thigh FF after 3 years: the vertical dashed lines depict the 90th percentile value for global thigh water T2 (i.e. 35.5 ms). (B) Relationship between years since symptom onset (at baseline) and baseline Pi,tot/PCr value: the horizontal dashed lines depict the 90th percentile value for anterior leg Pi,tot/PCr (i.e. 0.13). Investigating mean water T2, mean CK values, and the physical activity level during adolescence, three subgroups could be observed. (C) Total NSAD trajectories over 3 years for individual patients as a function of global thigh FF: the vertical dashed lines depict the 90th percentile value for global thigh FF (i.e. 10.4%). (D) Total NSAD trajectories over 3 years for individual patients as a function of global thigh cCSA: the vertical dashed lines depict the 10th percentile value for global thigh cCSA (i.e. 87.6 cm2). (E) Relationship between change in global thigh FF and the total NSAD after 3 years (in ambulant patients). (F) Relationship between change in global thigh FF and the change in total NSAD after 3 years (in ambulant patients). Spearman rho (ρ) correlation values and corresponding P‐values are also indicated [the correlation values in figures (C) and (D) are based on the baseline values]. Physical activity (phys. act.) in teenage years has been investigated in Jain COS patients in a publication by Moore et al., with 0 = reported no physical activity, 1 = reported vigorous activity occasionally/monthly or moderate activity weekly, 2 = reported moderate activity multiple times a week or vigorous activity once weekly, 3 = reported vigorous activity multiple times a week. [CK], creatine kinase concentration (in U/L); A, ambulant; cCSA, contractile cross‐sectional area (in cm2); FF, fat fraction (in %); LGMD R2, limb‐girdle muscular dystrophy type R2; MM, Miyoshi myopathy; NA, non‐ambulant; NSAD, North Star Assessment for limb‐girdle muscular Dystrophies; PCr, phosphocreatine; phys. ΔFF, change in FF (in %, absolute); Pi,tot, total inorganic phosphate; ΔNSAD, change in total NSAD value.