Longitudinal spinal cord MRI in adult 5q-SMA: biomarker and pathophysiological insights

Spinal muscular atrophy (SMA) is a progressive neuromuscular disease caused by biallelic SMN1 mutations and widespread lower motor neuron degeneration. Objective biomarkers to monitor progression and treatment response are critically needed. Spinal cord (SC) MRI, including diffusion tensor imaging (DTI), is a promising non-invasive approach, but data in adults remain scarce.

We conducted an exploratory longitudinal study in 26 adults with 5q-SMA (type 2: n = 4; type 3: n = 22; median age 29.5 years) and 27 controls undergoing 3 T cervical SC MRI. Gray matter (GM) and total cross-sectional area (CSA) and DTI metrics (FA, MD, AD, RD) were extracted from three white matter tracts. Patients underwent motor (MFM-32, HFMSE, RULM), neurophysiological (CMAP, MUNIX) and fatigue assessments. 22 patients completed 12-14-month follow-up.

Total SC CSA was reduced across C2-C6 (Cohen d up to 1.02) but showed limited functional correlations. GM CSA was reduced in SMA versus controls (11.90 ± 2.56 vs. 13.92 ± 1.72 mm²; p = 0.001; Cohen d = 0.93) and correlated robustly with MFM-32 (R = 0.60), HFMSE (R = 0.56), RULM (R = 0.54) and CMAP (R = 0.46). DTI revealed FA reduction and markedly elevated MD in fasciculi gracilis and cuneatus (Cohen d up to 2.44), without significant RD changes, suggesting axonal pathology. No DTI metric correlated with clinical variables. No longitudinal changes were detected at 12-14 months.

Selective GM CSA yields robust functional correlations, supporting its role as a candidate imaging biomarker and outperforming total CSA. DTI abnormalities in sensory tracts extend the known pathological scope of adult SMA beyond motor neurons. Longer follow-up is warranted to detect disease progression.