Skip to main content
Search
Main content
Multiple sclerosis and related disorders
Published

The value of CSF diagnostic and prognostic biomarkers in NMOSD and MOGAD in real-life use

Authors

M Johnsson, K Eriksson, I Rosenstein, L Novakova, C Malmeström, J Lycke, S Sandgren, H Zetterberg, K Blennow, K Johansson, M Axelsson

Abstract

Mult Scler Relat Disord. 2025 Jan 30;94:106302. doi: 10.1016/j.msard.2025.106302. Online ahead of print.

ABSTRACT

BACKGROUND: Diagnosing neuromyelitis optica spectrum disorder (NMOSD) may be challenging owing to overlapping clinical features with multiple sclerosis (MS) and myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD), particularly in AQP4-IgG seronegative NMOSD patients. We evaluated cerebrospinal fluid (CSF) biomarkers, specifically glial fibrillary acidic protein (GFAP) and albumin quotient (QAlb), to improve diagnostic accuracy in NMOSD.

METHODS: In this retrospective study, we analyzed CSF samples for biomarkers GFAP, QAlb, neurofilament light, and total-Tau, from patients with AQP4-NMOSD, DNNMOSD, MOGAD, and MS in the Region Västra Götaland, Sweden. Receiver operating characteristic (ROC) analysis with calculation of the area under the curve (AUC) was used to identify optimal cut-off levels for discriminating AQP4-NMOSD from the other groups. Logistic regression models assessed the diagnostic power of GFAP and QAlb combined.

RESULTS: Patients with AQP4-NMOSD (N = 19) had significantly higher CSF GFAP levels than the others: median 2470 ng/L vs 330 ng/L (p < 0.001). The GFAP cutoff >715 ng/L gave a sensitivity of 81 % and specificity of 92 %. Combining GFAP and QAlb further increased the diagnostic accuracy (AUC = 0.96). MOGAD patients (N = 29) had the highest CSF lymphocyte counts, with elevated lymphocyte counts correlating with polyphasic MOGAD (R = 0.63; p = 0.016).

CONCLUSION: CSF GFAP is a valuable biomarker for distinguishing AQP4-NMOSD from other demyelinating diseases: Combining GFAP with QAlb enhances diagnostic precision.

PMID:39893752 | DOI:10.1016/j.msard.2025.106302