Proteomic pattern analysis discriminates among multiple sclerosis-related disorders

Abstract

Objective: To use a new, unbiased biomarker discovery strategy to obtain and assess proteomic data from cerebrospinal fluid (CSF) of patients with multiple sclerosis (MS)-related disorders.

Methods: CSF protein profiles were analyzed from 107 patients with either MS-related disorders (including relapsing remitting MS [RRMS], primary progressive MS [PPMS], anti-aquaporin4 antibody seropositive-neuromyelitis optica spectrum disorder [SP-NMOSD], and seronegative-NMOSD with long cord lesions on spinal magnetic resonance imaging [SN-NMOSD]), amyotrophic lateral sclerosis (ALS), or other inflammatory neurological diseases (used as controls). CSF peptides/proteins were purified with magnetic beads, and directly measured by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. The obtained spectra were analyzed with multivariate statistics and pattern matching algorithms. These analyses were replicated in an independent sample set of 84 patients composed of those with MS-related disorders or with other neurological diseases (the second cohort).

Results: MS-related disorders differed considerably in terms of CSF protein profiles. SP-NMOSD and SN-NMOSD, both of which fit within the NMO spectrum, were distinguishable from RRMS with high cross-validation accuracy on a support vector machine classifier, especially in relapse phases. Some peaks derived from samples of relapsed SP-NMOSD can discriminate RRMS with high area under curve scores (>0.95) and this was reproduced on the second cohort. The similarity of proteomic patterns between selected neurological diseases were demonstrated by pattern matching analysis. To our surprise, the spectral differences between RRMS and PPMS were much larger than those of PPMS and ALS.

Interpretation: Our findings suggest that CSF proteomic pattern analysis can increase the accuracy of disease diagnosis of MS-related disorders and will aid physicians in appropriate therapeutic decision-making.