Related Experiment Video
Updated: Aug 29, 2026

Quantitative Magnetic Resonance Imaging of Skeletal Muscle Disease
Published on: December 18, 2016
Whole-Body Muscle MRI in Non-5q Spinal Muscular Atrophy: Patterns, Genotype Prediction, and Diagnostic Implications
Edouard Berling1,2,3, Harmen Reyngoudt3, Adrien Le Guillou4
1AP-HP, Service de Neurologie, Hôpital Raymond Poincaré, Centre de référence Nord-Est-Ile-de-France, Garches, France.
Background And Objectives:
Spinal muscular atrophy (SMA) is a group of genetically heterogeneous motor neuron disorders characterized by progressive, predominantly proximal weakness. Non-5q SMA refers to forms occurring without SMN1 pathogenic variants or deletions. More than half of non-5q SMA cases remain genetically unresolved, and muscle MRI may reveal characteristic patterns supporting diagnosis.
Methods:
We retrospectively analyzed multicenter data from adults presenting with chronic, progressive, motor-predominant weakness, beginning proximally in the upper or lower limbs, with exclusive or predominant motor involvement on nerve conduction studies and neurogenic changes on EMG. Whole-body MRI was qualitatively graded (Mercuri scale, 42 paired muscles) and quantitatively analyzed in the lower limbs.
Results:
Among 28 patients (median age 46 [26-50] years), 23 (82.1%) carried pathogenic variants in 13 genes, most commonly VWA1 (n = 5), TRPV4 (n = 4), and VRK1 and DYNC1H1 (n = 3 each). Distinct and reproducible MRI patterns were identified in VRK1-related, DYNC1H1-related, and VWA1-related cases. In VRK1-related SMA, the pattern combined upper limb and iliopsoas sparing with severe gluteal involvement, diffuse thigh involvement, and relative preservation of the posterior tibialis and medial gastrocnemius despite early calf involvement. DYNC1H1-related cases were characterized by sparing of the gluteus maximus, isolated adductor magnus involvement, predominant anterior thigh involvement with marked long head of biceps femoris involvement, and severe triceps surae involvement. In VWA1-related SMA, the pattern included selective gluteus maximus and quadriceps involvement with iliopsoas, gracilis, and sartorius preservation, and predominant peroneal involvement with relative extensor digitorum longus sparing. In TRPV4-related SMA, patterns of muscle involvement were heterogeneous, but the sternocleidomastoid was severely affected in 2 of 4 patients, whereas this muscle remained unaffected in all other cohort members.
Discussion:
Whole-body muscle MRI appears to be a valuable diagnostic tool for non-5q SMA. Gene-specific imaging signatures, particularly for VRK1, DYNC1H1, and VWA1, can guide targeted genetic testing and support interpretation of variants of uncertain significance. Sternocleidomastoid involvement may represent a specific marker of TRPV4-related SMA. However, confirmation in larger cohorts is needed, although this is constrained by the rarity of these disorders.

