M. Abdoli

University of Ottawa

Author Of 1 Presentation

Biomarkers and Bioinformatics Poster Presentation

P0051 - Comparison of serum and CSF fluid biomarkers for predicting long term disease progression in MS
  (ID 1448)

Speakers
Presentation Number
P0051
Presentation Topic
Biomarkers and Bioinformatics

Abstract

Background

With the availability of more powerful treatments for multiple sclerosis (MS), prognostic biomarkers are badly needed.

Objectives

Our objective was to evaluate the long-term prognostic value of 4 protiens in paired serum and CSF samples obtained early-on following MS diagnosis.

Methods

In this prospective cohort study, we identified patients with serum collected within 5 years of first MS symptom onset (baseline) and more than 15 years of routine clinical follow-up. Neurofilament light Chain (NfL), Glial Fibrillary Acidic Protein (GFAP), Tau and UCHL-1 were quantified in paired serum (s) and CSF (c) samples from patients and matched controls using digital immunoassay (SiMoA HD-1 Analyzer, Quanterix). Outcomes of biomarker performance included conversion to progressive MS phenotype and reaching an EDSS ≥4.

Results

67 patients had a median follow-up of 17.4 years (range:15.1-26.1), by which time 10/67 had been classified as PPMS, 16 SPMS and 41 RRMS. 29 had developed EDSS ≥4. Baseline CSF levels 3 of the candidate markers were higher than MS patients compared to controls: cNfL (Mann Whitney p=0.0001, median 624 vs. 277pg/mL), cGFAP (p<0.0001, 6900 vs. 694pg/mL) and cTau (p=0.0001, 15.4 vs. 8.12pg/mL) but not UCH-L1. Patient-control differences were less marked in serum: sNfL (p= 0.0037, 10.1 vs. 7.3pg/mL), sGFAP (p=0.0011, 68 vs 51pg/mL), no difference in sTau and sUCH-L1. Positive correlations existed between paired serum and CSF samples only for NfL (Spearman r=0.71, p<0.0001) and GFAP (r=0,4, p=0.003). ROC curve analysis showed cUCH-L1 was most predictive of developing EDSS ≥4 after 15 years of follow-up (AUC 0.72, p=0.003) followed by sNfL (AUC 0.70, p=0.012) and cGFAP (AUC 0.66, p=0.03). Similarly, cUCH-L1 was most predictive of developing a progressive phenotype (PP/SPMS, AUC 0.69, p=0.0097), followed by cGFAP (AUC 0.66, p=0.024) and barely by sNfL (AUC 0.64,p=0.057). cNfL (AUC 0.60,p=0.17), sGFAP, sTau, cTau and sUCH-L1 were not predictive of either reaching EDSS ≥4 or converting to a progressive phenotype (PP/SPMS).

Conclusions

This is the first study to report and association of baseline CSF UCH-L1 levels with long term clinical outcomes in MS. This marker was more predictive of EDSS worsening and conversion to a progressive phenotype than well-established markers NfL and GFAP. More generally, CSF biomarker levels better segregating MS patients from controls at baseline compaired to levels in paired serum samples.

Collapse