Macrostructural Brain Abnormalities in Spinal Muscular Atrophy: A Case-Control Study

Abstract

Background and objectives: Most individuals with spinal muscular atrophy (SMA) on disease-modifying therapies continue to have chronic motor impairment. Insights into brain involvement in SMA may open new pathways for adjunctive therapies to optimize outcomes. We aimed to characterize macrostructural brain abnormalities detected by MRI in individuals with SMA compared with peer controls.

Methods: We conducted a cross-sectional case-control study of children and adults with a confirmed genetic diagnosis of 5q SMA, and peer controls matched by age and sex. Brain MRIs acquired on a 3T MRI scanner through a standardized research protocol were reviewed to qualitatively assess the presence of macrostructural changes. The primary outcome was the presence of any structural brain anomaly on MRI. In addition, the total volume of each participant's lateral ventricles was quantified by volumetry using MRIcron. Genetic and clinical variables, including SMN2 copy number and motor function (Hammersmith Functional Motor Scale Expanded and Revised Upper Limb Module scores), were then correlated with neuroimaging findings.

Results: A total of 42 participants completed the study (mean age 17.4, range 7-40; 67% male). Of the 21 individuals with 5q SMA, 9 (43%) had macrostructural brain abnormalities identified on MRI compared with 2 of 21 (10%) peer controls (odds ratio 7.1, 95% confidence interval 1.4-34.0). In patients with SMA, the most common structural changes were widening of the arachnoid spaces (n = 4) and ventriculomegaly (n = 4). Individuals with SMA had larger median lateral ventricular volume than their normally developing peers (9.3 mL, interquartile range [IQR] 5.5-13.1 vs 5.3 mL, IQR 3.8-9.8; p = 0.034). Structural brain abnormalities were more frequent in those with 2 SMN2 copies (3/5, 60%) compared with 3 or 4 SMN2 copies (4/10, 40% and 2/6, 33% respectively), not reaching significance. We found no association between structural changes and motor function scores.

Discussion: Individuals with SMA have higher rates of macrostructural brain abnormalities than their neurotypical peers, suggesting CNS involvement in SMA. Understanding changes in the brain architecture of the SMA population can inform the development of adjunct therapies targeting the CNS and potentially guide rehabilitation strategies.

Figure 1. Ventricle Volumetry Mapping

Illustration of volume mapping of the left lateral ventricle using MRIcron.

Figure 2. Participant Flowchart

Flow diagram of participants and MRI findings.

Figure 3. Macrostructural Abnormalities

Examples of structural brain abnormalities observed in patients with SMA. (A) Widening of the anterior arachnoid spaces and widening of the frontal horns of the bilateral lateral ventricles. (B) Asymmetric widening of the arachnoid spaces and widening of the supratentorial ventricular system. (C) Left frontal paraventricular focal lesion of hyperintense T2 signal.

Figure 4. Macrostructural Abnormalities by SMN2 Copy Number and SMA Type

Rate of macrostructural brain abnormalities among patients with SMA, classified by (left) number of SMN2 copies and (right) SMA type. Error bars represent the 95% confidence interval.

Figure 5. Total Lateral Ventricular Volume

Total volume of the lateral ventricles for (left) peer controls and patients with SMA and (right) patients with SMA stratified by SMA type. The ventricular volumes were measured by MRI volumetry. The outlier type III SMA patient with a total lateral ventricular volume of 82.5 mL had no signs of hydrocephalus in the brain MRI study. They were born prematurely (29 + 5 weeks) and had evidence of intraventricular hemorrhage in the neonatal period with no known clinical sequelae.