CSF CXCL13 and Chitinase 3-like-1 Levels Predict Disease Course in Relapsing Multiple Sclerosis

Abstract

Several biomarkers from multiple sclerosis (MS) patients' biological fluids have been considered to support diagnosis, predict disease course, and evaluate treatment response. In this study, we assessed the CSF concentration of selected molecules implicated in the MS pathological process. To investigate the diagnostic and prognostic significance of CSF concentration of target candidate biomarkers in both relapsing (RMS, n = 107) and progressive (PMS, n = 18) MS patients and in other inflammatory (OIND, n = 10) and non-inflammatory (ONIND, n = 15) neurological disorders. We measured the CSF concentration of APRIL, BAFF, CHI3L1, CCL-2, CXCL-8, CXCL-10, CXCL-12, CXCL-13 through a Luminex Assay. MS patients were prospectively evaluated, and clinical and radiological activity were recorded. CHI3L1 and CXCL13 CSF levels were significantly higher in both MS groups compared to control groups, while CCL2, BAFF, and APRIL concentrations were lower in RMS patients compared to PMS and OIND. Considering RMS patients with a single demyelinating event, higher concentrations of CHI3L1, CXCL10, CXCL12, and CXCL13 were recorded in patients who converted to clinically defined MS(CDMS). RMS patients in the CXCL13 and CHI3L1 high concentration group had a significantly higher risk of relapse (HR 12.61 and 4.57), MRI activity (HR 7.04 and 2.46), and of any evidence of disease activity (HR 12.13 and 2.90) during follow-up. CSF CXCL13 and CHI3L1 levels represent very good prognostic biomarkers in RMS patients, and therefore can be helpful in the treatment choice. Higher CSF concentrations of neuro-inflammatory biomarkers were associated with a higher risk of conversion to CDMS in patients with a first clinical demyelinating event. Differential CSF BAFF and APRIL levels between RMS and PMS suggest a different modulation of B-cells pathways in the different phases of the disease.

Keywords: APRIL; BAFF; Biomarker; CCL2; CNS; CSF; CXC13; CXCL10; CXCL12; CXCL8; Chemokine; Chitinase 3-like1; Multiple sclerosis.

Fig. 1

CSF concentration of APRIL (A), BAFF (B), CHI3L1 (C), and CCL2 (D) for each study group in pg/ml except for CHI3L1 measured as ng/ml. The boxes represent median and interquartile range. Statistical differences between groups were highlighted: a singler asteris (*) for p-value < 0.05 and double asterisks (**) for p-value < 0.01. Comparisons of CSF biomarkers concentration between multiple groups were explored with ANOVA for APRIL, BAFF, and CCL-2 and with Kruskal–Wallis for CHI3L1. Comparisons between two independent groups were assessed through Student T-test for APRIL, BAFF, and CCL-2 and with Mann–Whitney test for CHI3L1. RMS, relapsing multiple sclerosis; PMS, progressive multiple sclerosis; ONIND, other non-inflammatory neurological disorders; OIND, other inflammatory neurological disorders

Fig. 2

CSF concentration of CXCL8 (A), CXCL10 (B), CXCL12 (C), and CXCL13 (D) for each study group in pg/ml except for CHI3L1 measured as ng/ml. The boxes represent median and interquartile range. Statistical differences between groups were highlighted: a singler asteris (*) for p-value < 0.05 and double asterisks (**) for p-value < 0.01. Comparisons of CSF biomarkers concentration between multiple groups were explored with ANOVA for CXCL8, CXCL10, and CXCL12 and with Kruskal–Wallis for CXCL13. Comparisons between two independent groups were assessed through Student T-test for CXCL8, CXCL10, and CXCL12 with Mann–Whitney test for CXCL13. RMS, relapsing multiple sclerosis; PMS, progressive multiple sclerosis; ONIND, other non-inflammatory neurological disorders; OIND, other inflammatory neurological disorders

Fig. 3

CSF concentration of CHI3L1 (A), CXCL10 (B), CXCL12 (C), CXCL13 (D) in patients with a single demyelinating event at baseline. The two groups were divided following the occurrence (relapsing) or the absence (non-relapsing) of new relapse. The boxes represent median and interquartile range. Statistical differences (Student t-test for CXCL12 and Mann–Whitney test for CHI3L1, CXCL12, and CXCL13) between groups were highlighted: a singler asteris (*) for p-value < 0.05 and double asterisks (**) for p-value < 0.01

Fig. 4

ROC curve to predict conversion to CDMS in patients with a first demyelinating event at baseline. A Each line represents the ROC curve for a single CSF biomarker (CHI3L1, CXCL13, CXCL2, and CXCL10). B ROC curve obtained through a multivariate regression analysis including biomarkers associated with the risk of conversion to CDMS. p < 0.01 for all analysis. ROC, receiver operating characteristic; CDMS, clinically defined multiple sclerosis

Fig. 5

Cox regression analysis to represent the proportion of patients with disease activity (relapse in A; MRI activity in B; disability progression in C; any disease activity in D) based on the CSF concentration of CHI3L1. MRI, magnetic resonance imaging; CSF, cerebrospinal fluid

Fig. 6

Cox regression analysis to represent the proportion of patients with disease activity (relapse in A; MRI activity in B; disability progression in C; any disease activity in D) based on the CSF concentration of CXCL13. MRI, magnetic resonance imaging; CSF, cerebrospinal fluid