Characteristic Cerebrospinal Fluid Cytokine/Chemokine Profiles in Neuromyelitis Optica, Relapsing Remitting or Primary Progressive Multiple Sclerosis

Takuya Matsushita,Hiroyuki Murai,Takahisa Tateishi,Ryo Yamasaki,Noriko Isobe,Jun-Ichi Kira,Dai Matsuse,Tomomi Yonekawa,Akio Suzumura

doi:10.1371/journal.pone.0061835

Takuya Matsushita, Hiroyuki Murai + Show 7 more

Open Access

https://doi.org/10.1371/journal.pone.0061835

Copy DOI

Journal: PLoS ONE	Publication Date: Apr 18, 2013
Citations: 184	License type: CC BY 4.0

Affiliation: Kyushu University

Abstract

BackgroundDifferences in cytokine/chemokine profiles among patients with neuromyelitis optica (NMO), relapsing remitting multiple sclerosis (RRMS), and primary progressive MS (PPMS), and the relationships of these profiles with clinical and neuroimaging features are unclear. A greater understanding of these profiles may help in differential diagnosis.Methods/Principal FindingsWe measured 27 cytokines/chemokines and growth factors in CSF collected from 20 patients with NMO, 26 with RRMS, nine with PPMS, and 18 with other non-inflammatory neurological diseases (OND) by multiplexed fluorescent bead-based immunoassay. Interleukin (IL)-17A, IL-6, CXCL8 and CXCL10 levels were significantly higher in NMO patients than in OND and RRMS patients at relapse, while granulocyte-colony stimulating factor (G-CSF) and CCL4 levels were significantly higher in NMO patients than in OND patients. In NMO patients, IL-6 and CXCL8 levels were positively correlated with disability and CSF protein concentration while IL-6, CXCL8, G-CSF, granulocyte-macrophage colony-stimulating factor (GM-CSF) and IFN-γ were positively correlated with CSF neutrophil counts at the time of sample collection. In RRMS patients, IL-6 levels were significantly higher than in OND patients at the relapse phase while CSF cell counts were negatively correlated with the levels of CCL2. Correlation coefficients of cytokines/chemokines in the relapse phase were significantly different in three combinations, IL-6 and GM-CSF, G-CSF and GM-CSF, and GM-CSF and IFN-γ, between RRMS and NMO/NMOSD patients. In PPMS patients, CCL4 and CXCL10 levels were significantly higher than in OND patients.ConclusionsOur findings suggest distinct cytokine/chemokine alterations in CSF exist among NMO, RRMS and PPMS. In NMO, over-expression of a cluster of Th17- and Th1-related proinflammatory cytokines/chemokines is characteristic, while in PPMS, increased CCL4 and CXCL10 levels may reflect on-going low grade T cell and macrophage/microglia inflammation in the central nervous system. In RRMS, only a mild elevation of proinflammatory cytokines/chemokines was detectable at relapse.

Highlights

Neuromyelitis optica (NMO) is characterized by severe and selective involvement of the optic nerves and spinal cord, which frequently presents as longitudinally extensive spinal cord lesions (LESCLs) extending over three or more vertebral segments, and neutrophil-predominant pleocytosis of the cerebrospinal fluid (CSF) [1]
Our findings suggest distinct cytokine/chemokine alterations in CSF exist among neuromyelitis optica (NMO), relapsing remitting multiple sclerosis (RRMS) and primary progressive MS (PPMS)
Statistically significant differences by Kruskal– Wallis test were noted in the levels of IL-6, CXCL8 (0.0017), IL-17A (0.010), granulocyte-colony stimulating factor (G-CSF) (0.010), CCL4 (0.010), and CXCL10 (0.00031) among patients with NMO, RRMS, PPMS and other non-inflammatory neurological diseases (OND) after correction for multiple tests by the Benjamini-Hochberg method (Fig. 1)

Summary

Introduction

Neuromyelitis optica (NMO) is characterized by severe and selective involvement of the optic nerves and spinal cord, which frequently presents as longitudinally extensive spinal cord lesions (LESCLs) extending over three or more vertebral segments, and neutrophil-predominant pleocytosis of the CSF [1]. In an ex vivo study, mouse optic nerves exposed to anti-AQP4 antibody-positive serum showed reduced levels of myelin basic protein [7] These findings suggested anti-AQP4 antibody induces pathogenic effects. Anti-AQP-4 antibody injection into young rats with a leaky blood–brain barrier did not induce disease or neuropathological changes in the central nervous system (CNS), despite apparent vascular leakage of human immunoglobulin to the perivascular tissue [8]. It is still unknown whether NMO is caused by anti-AQP4 antibody and complement alone or whether T cell involvement is required to trigger inflammation. A greater understanding of these profiles may help in differential diagnosis

Results

Discussion

Conclusion