Longitudinal Follow-Up of Patients With Duchenne Muscular Dystrophy Using Quantitative 23Na and 1H MRI

Abstract

Background: Quantitative muscle MRI commonly evaluates disease activity and muscle wasting in Duchenne muscular dystrophy (DMD). Disturbances in ion homeostasis contribute to DMD pathophysiology, but their relationships with disease progression is unclear. ²³Na MRI may provide insights into the disease course and treatment response. This longitudinal study assessed whether sodium levels are elevated in DMD patients regardless of fat fraction (FF) and whether baseline sodium levels influence FF changes over time. Additionally, we quantified the effect of slice selection on measured sodium values.

Methods: Thirteen DMD boys (age 7.8 ± 2.4 years) underwent MRI of lower leg muscles at 3T at three visits, spaced 6 months apart. We assessed FF for disease progression and water T₂, pH, apparent tissue sodium concentration (aTSC), and intracellular-weighted ²³Na signal (ICwS) for disease activity. Fourteen healthy boys (age 9.5 ± 1.7 years) underwent the same MRI protocol once. Linear regression and mixed-effect modelling were used to examine sodium level increases and their impact on FF changes.

Results: In DMD, muscles with FF < 10% exhibited significantly elevated aTSC (24.8 ± 4.6 mM vs. 14.5 ± 2.1 mM in controls, p < 0.001) and higher ICwS (23.6 ± 2.5 a.u. vs. 14.1 ± 2.1 a.u., p < 0.001). At Visit 1, FF values showed a significant negative association with aTSC (β = -17.30, p = 0.016) and ICwS (β = -21.02, p < 0.001). The first mixed-effect model, which assessed aTSC alone, showed no significant effect on FF progression but indicated a weak trend (p = 0.098). The second, more comprehensive model-incorporating also ICwS and water T₂-revealed that FF changes were positively associated with aTSC (p = 0.0023) and negatively associated with ICwS and wT₂ (p < 0.001 and p = 0.025, respectively), with ICwS showing a significant interaction with time (p = 0.0033). Varying slice positioning and slice number demonstrated minimal impact on aTSC and ICwS, with low CV (2%-4%) in the mid-belly region.

Conclusions: The study demonstrates significant MRI-based changes related to dystrophic alterations in DMD. We identified early alterations in sodium homeostasis, independent of FF. Our findings suggest that the relationship between sodium levels and FF progression is complex and may not be fully explained by total sodium measurements alone. Given the small sample size, further validation in larger cohorts is needed. Combined ¹H and ²³Na-MRI may offer deeper insights into how metabolic and ionic changes interact with FF progression and overall disease activity.

Keywords: Duchenne muscular dystrophy; fat fraction; muscle imaging; quantitative MRI; skeletal muscle; sodium.

FIGURE 1

Flow diagram depicting study design, enrollment and dropouts. Patients were recruited through the local Children's Hospital and through newsletters distributed by the patient registry of the Ludwig‐Maximilians‐Universität Munich. Eligible patients who were willing underwent up to three MRI scans with a 6‐month interval between scans. MD, muscle dystrophy.

FIGURE 2

Quantitative MRI‐derived muscle quality maps of the lower leg of an 11‐year‐old patient at Visit 1 (first row) and Visit 3 (second row) and an age‐matched control (third row). Dystrophic muscles showed increased fat content, water T₂, apparent tissue sodium concentration (aTSC), and intracellular‐weighted sodium signal (ICwS) compared with healthy muscle tissue. Active disease activity, as indicated by higher water T₂, aTSC, and ICwS, was also observed in spared muscles with low fat content, such as the tibialis anterior and tibialis posterior muscles. The legs are positioned on top of reference phantoms that are utilized for ²³Na signal quantification (from left to right: 40 mM NaCl, 40 mM NaCl + 5% Agarose, 20 mM NaCl, 20 mM NaCl + 5% Agarose). It is worth noting that the inversion‐recovery pulse effectively suppresses the ²³Na signal from the saline solution (without agarose) in the ICwS images.

FIGURE 3

Changes over time for quantitative MRI derived measures. Each line represents a single muscle over age, with different colours indicating the four different muscles. Dystrophic muscles showed an overall increase in fat fraction, water T₂, aTSC, and ICwS compared with healthy muscle tissue (median across all four muscles shown with a solid grey line and the corresponding 0.9 quantile with a dashed grey line), but no difference in pH. CTL, control; GM, gastrocnemius medialis; Sol, soleus; TA, tibialis anterior; TP, tibialis posterior.

FIGURE 4

Water T₂, pH, aTSC, and ICwS plotted against fat fraction for the three visits. Each line represents a single muscle over time, with different colours indicating the four different muscles. Dystrophic muscles showed increased water T₂, aTSC, and ICwS compared with controls, with their median across all four muscles shown as a solid grey line and the 0.9 quantile as a dashed line. These increases in water T₂, aTSC, and ICwS were also observed in muscles with low fat fraction. ICwS demonstrated a negative correlation with fat fraction (Spearman rho = −0.22, p = 0.034). Water T₂ values are missing for one patient (high fat fraction) for Visit 3, as the leg was too stiff to correctly position the calf in the isocenter of the coil. CTL, control; GM, gastrocnemius medialis; Sol, soleus; TA, tibialis anterior; TP, tibialis posterior.