Multiparametric quantitative magnetic resonance imaging biomarkers associated with clinical phenotype in dystrophinopathies

Dystrophinopathies are a group of inherited progressive myopathies marked by muscle weakness caused by genetic mutations, including Duchenne muscular dystrophy (DMD), Becker muscular dystrophy (BMD), and intermediate muscular dystrophy (IMD). Existing clinical assessments demonstrate limitations in specificity and lack quantitative standards. This study evaluated the clinical value of quantitative magnetic resonance imaging (qMRI) in assessing dystrophinopathies. The qMRI technique combines fat fraction (FF), water T2 relaxation time (wT2), and diffusion-tensor imaging (DTI). A total of 42 participants were enrolled: 30 genetically confirmed male patients (all males, age 4–22 years; Group 1: 20 with DMD, mean age 10.40 ± 3.28 years, Group 2: 8 with BMD plus 2 with IMD, mean age 11.80 ± 4.66 years) and 12 healthy controls (all males, mean age 14.92 ± 3.55 years). All underwent 3 T MRI using Dixon-based FF, wT2, and DTI (yielding fractional anisotropy (FA), apparent diffusion coefficient (ADC), and eigenvalues λ1, λ2, λ3). AI-assisted semiautomatic muscle segmentation and regional qMRI measurements were performed using DAFNE. Then, qMRI parameters were compared across groups and correlated with clinical parameters. Compared to controls, DMD demonstrated markedly increased FF and prolonged wT2 across all examined muscle groups, accompanied by significantly reduced ADC values (all adjusted p < 0.001). Exploratory DTI analysis revealed a modest increase in FA limited to the anterior muscle group. The BMD+IMD group exhibited a milder, more heterogeneous imaging phenotype, with elevated FF and wT2 values in specific muscle groups and increased diffusion eigenvalues (λ1 and λ2; adjusted p < 0.05). Correlation analyses demonstrated that FF values showed robust positive correlations with Gower’s sign across all muscle groups, whereas ADC values were strongly negatively correlated with Gower’s sign (all |r|> 0.7). Exploratory β-CTX correlations were observed with all qMRI parameters, displaying inverse associations with FF and T2 and positive associations with ADC and diffusion eigenvalues (p < 0.05). Conventional serum muscle enzymes showed inconsistent associations with qMRI measures. Multiparametric qMRI, particularly FF and wT2, characterizes muscle involvement and its associations with clinical phenotype in dystrophinopathies; diffusion and biochemical associations remain exploratory. AI-assisted semiautomatic segmentation supports integrated multiparametric qMRI assessment of muscle involvement in dystrophinopathies. qMRI metrics show strong associations with clinical indicators but only weak correlations with conventional serum enzymatic biomarkers, highlighting the limitations of relying solely on traditional biochemical markers. Multiparametric qMRI provides non-invasive, radiation-free characterization of muscle involvement and its associations with clinical phenotype.