Unbiased bioinformatic and clinical analysis identifies glial fibrillary acidic protein as a potential biomarker in spinal muscular atrophy.

Spinal muscular atrophy (SMA) is an autosomal recessive disorder caused by bi-allelic pathogenic variants in SMN1, leading to motor neuron degeneration, muscle atrophy and progressive weakness. Although disease-modifying therapies have recently become available, reliable biomarkers for disease monitoring remain limited. This study aimed to identify potential biomarkers of disease severity and monitoring in SMA using an unbiased approach based on differential gene expression datasets. A meta-analysis of datasets from 15 studies in SMA mouse models identified genes differentially expressed in at least three studies. Of the 134 differentially expressed genes, 122 had human orthologs, with only two being CNS-specific and consistently altered across studies, including glial fibrillary acidic protein (GFAP), a marker of astrocyte reactivity. Candidate gene expression was subsequently evaluated in two human datasets. GFAP and neurofilament light chain (NfL) serum levels were then measured in a case-control study including 25 SMA patients (11 type 1, 14 types 2/3) and 24 matched controls, followed by a 12-month cohort study of the SMA patients. GFAP levels in cerebrospinal fluid (CSF) were assessed in 14 patients before and up to 14 months after initiation of disease-modifying therapies. Serum GFAP levels were lower in SMA patients compared with controls (P = 0.014), but higher in those with SMA type 1 (P = 0.028) and in patients with two SMN2 copies (P = 0.03). Patients under disease-modifying treatment exhibited higher baseline GFAP levels (P = 0.01). CSF GFAP levels decreased after 14 months of treatment in SMA type 1 and in pre-symptomatic patients predicted to develop type 1 (P = 0.045). NfL levels were elevated in SMA patients, correlating with disease severity and SMN2 genotype, and further increased over 12 months. Taken together, this study, employing an unbiased biomarker discovery strategy, identified GFAP as a candidate biomarker for SMA. Clinical validation demonstrated reduced serum GFAP levels, which correlated with disease type, SMN2 copy number and treatment status. In SMA type 1, CSF GFAP levels declined over time with treatment. These findings implicate astrocytes in SMA-related neurodegeneration and support GFAP as a potential biomarker for the disease.