GFAP and VEGF Trajectories in Adult Type 3 Spinal Muscular Atrophy Under Nusinersen

For adults with type 3 spinal muscular atrophy (SMA), nusinersen has shifted expectations from stabilization to measurable motor improvement. What remains less clear is how to biologically track treatment response in this population.

In a single-center longitudinal study published in the Journal of the Neurological Sciences, Pronto-Laborinho and colleagues examined whether glial fibrillary acidic protein (GFAP) and vascular endothelial growth factor (VEGF) in cerebrospinal fluid (CSF) and plasma could serve as biomarkers during 30 months of nusinersen therapy in adults with type 3 SMA.

The clinical question is timely: while neurofilament and other neuronal markers have been explored, reliable biomarkers in treated adult SMA remain elusive. Given emerging evidence implicating astrocytic dysfunction and vascular alterations in SMA pathogenesis, GFAP and VEGF were selected as hypothesis-generating candidates.

A Longitudinal Look at Treated Adult SMA

The investigators followed 11 adults, aged 19 to 60 years at treatment initiation, all with genetically confirmed type 3 SMA. Most had four SMN2 copies, and average disease duration at treatment start exceeded three decades. Participants received nusinersen per standard protocol and were evaluated at baseline, 6, 14, 22, and 30 months. Primary biomarker analyses focused on 14- and 30-month changes.

Functional outcomes included the Hammersmith Functional Motor Scale Expanded (HFMSE), Revised Upper Limb Module (RULM), and respiratory measures. GFAP and VEGF levels in CSF and plasma were measured using ELISA, with all samples analyzed in batch to reduce interassay variability.

Motor Gains were Sustained

Over 30 months, motor outcomes improved significantly. Mixed-effects modeling estimated a 6.96-point increase in HFMSE, with the 95% CI ranging from 5.44 to 8.48 points. For RULM, the model suggested a non-linear 9.02% increase over the same period.

These changes exceed previously reported minimal clinically important differences of 3 to 4 HFMSE points in adult type 3 SMA, supporting not only statistical significance but likely clinical relevance. Respiratory measures remained largely stable, consistent with expectations in a relatively preserved adult cohort.

Biomarkers Did Not Track with Clinical Change

In contrast, neither GFAP nor VEGF demonstrated statistically significant longitudinal change in CSF or plasma.

CSF GFAP showed numerical decreases at 14 and 30 months, but without statistical significance. Plasma GFAP fluctuated modestly over time. Similarly, CSF and plasma VEGF levels varied but did not show consistent directional change or significant time effects. In mixed-effects models, month coefficients for all four biomarker measures were nonsignificant, with p values well above 0.05.

Importantly, baseline CSF GFAP levels appeared elevated relative to values reported in other neurodegenerative contexts, suggesting ongoing astroglial activation in this adult SMA cohort. However, that activation did not meaningfully shift with treatment over the study period.

Interpreting the Disconnect

Several factors temper interpretation. The cohort was small, and biomarker variability was substantial, with standard deviations in some cases approaching or exceeding mean concentrations. Such dispersion can obscure subtle longitudinal effects. The single-center design and heterogeneous disease duration further limit generalizability.

It is also possible that GFAP and VEGF reflect aspects of disease biology not directly modified by nusinersen in longstanding adult SMA. Nusinersen increases full-length SMN protein primarily within motor neurons, but chronic astroglial or vascular alterations may be less reversible after decades of disease. Alternatively, any treatment effect on these pathways may be too modest to detect with current assay sensitivity in small samples.

Clinical Implications

For now, functional scales remain the most reliable indicators of treatment response in adult type 3 SMA. This study reinforces the durability of motor improvement over 30 months while underscoring the ongoing challenge of identifying molecular correlates in treated adults.

GFAP and VEGF remain biologically plausible markers, particularly as interest grows in non-neuronal contributors to SMA pathogenesis. Larger, multicenter cohorts with stratification by baseline severity and SMN2 copy number may clarify whether subgroups demonstrate distinct biomarker trajectories.

Until then, the message is pragmatic: in adult type 3 SMA, nusinersen’s clinical benefit is measurable at the bedside. The search for a parallel molecular signal continues.

Reference:

Pronto-Laborinho AC, Freitas T, Domingues S, Rosa MG, de Carvalho M, Oliveira Santos M. CSF and plasma GFAP and VEGF in adult type 3 spinal muscular atrophy patients treated with nusinersen. J Neurol Sci. Published online February 5, 2026. doi:10.1016/j.jns.2026.125795