fbpx Skip to main content

Serum neurofilament light chain level shows promise as a biomarker for predicting spinal cord atrophy in neuromyelitis optica spectrum disorder.

Serum neurofilament light chain (sNfL) and serum glial fibrillary acidic protein (sGFAP) levels are linked to disease activity and disability in patients with neuromyelitis optica spectrum disorder (NMOSD). High sNfL levels are associated with brain and spinal cord atrophy in multiple sclerosis; however, the relationship is unclear in NMOSD. 

A prospective study in the Journal of Neuroimmunology investigated the association of sNfL and sGFAP levels with clinical disability and magnetic resonance imaging measurements in NMOSD.

Patient Characteristics

Fifteen NMOSD patients were enrolled in the study. The median age was 43 years, and 14 (93.3%) patients were female. The median disease duration was 2.5 years. The baseline median Expanded Disability Status Scale (EDSS) score was 4.7. The participants completed a median follow-up of 12.2 months.

Baseline Log-sNfL Positively Predicted Spinal Cord Atrophy in NMOSD

Univariate analysis demonstrated a positive association of baseline log-sNfL levels with age at baseline (p=0.035), baseline EDSS score (p=0.043), follow-up EDSS score (p=0.040), and the percentage change in cervical spinal cord volume (CSCV) (p=0.007). Multivariable analysis revealed that baseline log-sNfL levels were independently associated with the percentage change in CSCV over the follow-up period (p=0.025). Baseline log-sNfL strongly and positively predicted spinal cord atrophy in NMOSD patients in both univariate and multivariate models.

Univariate analysis showed that baseline log-sGFAP levels were significantly associated with age (p=0.011) and brain gray matter volume at baseline (p=0.021). However, the association between baseline log-sGFAP and the percentage change in CSCV was not statistically significant. Multivariable analysis revealed that baseline log-sGFAP was not independently associated with age or percentage change in CSCV.

Longitudinal Change in sNfL Predicts Cervical Spinal Cord Atrophy in NMOSD

You May Also Like::  Brain Volume Loss in Ocrelizumab-Treated Relapsing Multiple Sclerosis

The values of the changes in log-sNfL (∆log-sNfL) and log-sGFAP (∆log-sGFAP) were calculated between the baseline and follow-up values. A linear regression model revealed a significant association between ∆log-sNfL and the percentage change in CSCV (average CSCV reduction of 7.48% per 10 pg/ml increase in sNfL, p=0.001). A positive but non-significant trend was observed between ∆log-sGFAP and the percentage change in CSCV. Thus, the longitudinal change in log-sNfL can be a potent biomarker for predicting cervical spinal cord atrophy in NMOSD.

Log-sGFAP Reduction and EDSS Score Reduction in NMOSD

Of the 15 participants, 10 demonstrated EDSS score reduction in the follow-up visit. Univariate analysis showed an association between EDSS score reduction and EDSS baseline score (p=0.172), baseline log-sNfL (p=0.960), baseline log-sGFAP (p=0.181), log-sNfL changes (p=0.106), and log-sGFAP changes (p=0.050) over the follow-up period. The reduction in log-sGFAP was the only relevant factor influencing EDSS score reduction; however, the multivariate model found this association to be not significant.


Li, Y., Liu, Y., Wang, Z., An, X., Zhang, F., Zhang, T., Liu, Y., Du, C., Zeng, P., Yuan, M., Zhang, N., & Zhang, C. (2023). Serum neurofilament light chain predicts spinal cord atrophy in neuromyelitis optica spectrum disorder. Journal of Neuroimmunology, 384, 578218. https://doi.org/10.1016/j.jneuroim.2023.578218