301PPatterns of volume and fat infiltration in skeletal muscle of adults with spinal muscular atrophy

Abstract

Spinal muscular atrophy (SMA) is a genetic disorder characterized by protein deficiency affecting lower motor neurons and the neuromuscular junction, leading to muscle weakness and fatigue. Pathologically, muscles exhibit progressive fat infiltration and atrophy. Whole-body MRI is a valuable tool for assessing these changes in vivo. Structural T1 MRI sequences can quantitatively measure fat replacement in skeletal muscle, while fluid-sensitive T2 sequences, such as short tau inversion recovery (STIR), can detect inflammatory changes that precede clinical symptoms and fat infiltration. These imaging techniques can serve as diagnostic biomarkers and monitor disease progression. Recent advancements in artificial intelligence (AI) have enabled reliable automated methods for masking and identifying individual muscles in whole-body MRI, facilitating group-level analysis of muscle volume and fat fraction. Previous studies have highlighted distinct patterns of muscle involvement across different SMA types. This study aims to characterize the neuromuscular pathway structures affected in adults with SMA type 3. We conducted whole-body 3T MRI on six subjects with SMA type 3 (five ambulatory, one non-ambulatory; ages 26-67). Imaging sequences included T1 (2-point) DIXON, T2 STIR, and structural T1/T2 of the brain. An AI-based masking approach (Springbok) was employed to isolate and assess 71 individual muscles across 36 functional groups for muscle volume, fat fraction, and left-right asymmetry. Group-level analysis revealed reduced volumes in hip flexors, hip adductors, knee extensors, and proximal upper extremity muscles, with significant fat infiltration in these areas. Asymmetry levels varied among subjects, with individual muscles exhibiting localized fat infiltration patterns. Notably, atrophy and fat fraction were most pronounced in the hip and knee extensors in the lower extremities, while axial muscles of the trunk and proximal limbs were predominantly affected in the upper extremities. These findings align with previous studies and clinical presentations, suggesting a targeted pathological process affecting specific motor units. This study supports the hypothesis that SMN protein deficiency in SMA selectively impacts certain neuromuscular motor groups, correlating with segmental lower motor neuron weakness observed clinically and in post-mortem analyses. Limitations include the small cohort size and absence of additional imaging modalities. Future research will aim to enhance statistical power and explore central effects on higher structures within the neuromuscular pathway, potentially identifying specific biomarkers or therapeutic targets.