The chemokine CXCL13 is a prognostic marker in clinically isolated syndrome (CIS)

Abstract

Background: There is increasing recognition of the importance of B lymphocytes in the immunopathogenesis of multiple sclerosis (MS), encouraging the evaluation of B cell-associated biomarkers in the cerebrospinal fluid (CSF). We aimed to evaluate the relevance of the B cell chemoattractant CXCL13 as a prognostic marker in patients with clinically isolated syndrome (CIS) regarding conversion to MS, and to compare it to Barkhof criteria in magnetic resonance imaging (MRI), oligoclonal bands (OCB) and the polyspecific intrathecal B cell response against measles, rubella and varicella zoster virus (MRZR).

Methodology/principal findings: CXCL13 was determined in a prospective study over 2 years including 46 patients that remained CIS over follow-up (CIS-CIS), 45 patients that developed MS (CIS-RRMS), and 30 controls using ELISA. CSF CXCL13 was significantly elevated in CIS-RRMS as compared to CIS-CIS and controls (p<0.001). It was significantly elevated in CIS with OCB (p<0.001), positive MRZR (p=0.04), and gadolinium enhancement in MRI (p=0.02) and showed a significant correlation with CSF leukocyte count (p<0.001) and QIgG (p<0.001). CXCL13 showed the best positive predictive value (PPV) of all parameters investigated (70%, 95%-CI: 53-84%), which could be further increased by combination with Barkhof criteria in MRI (80%).

Conclusions/significance: Our data indicate the relevance of CXCL13 in CIS to predict conversion to MS. It furthermore shows CXCL13 to be an important mediator in the inflammatory cascade associated with the polyspecific intrathecal B cell response that manifests itself in OCB and MRZR.

Figure 1. Reiber diagram and antibody index.

Left side: Reiber diagram visualizing intrathecal IgG synthesis using a double-logarithmic scaling to present the IgG CSF to serum quotient (QIgG) in relation to the albumin CSF to serum quotient (Qalb), and showing the hyperbolic function Qlim that indicates the upper reference range of QIgG. In the upper right corner of the diagram, the relative extent of intrathecal IgG synthesis is indicated. Right side: Formula for antibody index (AI) and Qlim.

Figure 2. CXCL13 in patients with CIS.

A) Boxplot shows CSF concentrations of CXCL13 in patients with CIS-CIS (patients with CIS that remained CIS over follow-up of two years), CIS-RRMS (CIS patients with conversion to MS over follow-up) and controls (CTRL). P-value refers to comparison of CIS-CIS, CIS-RRMS and controls. B) Boxplot shows CSF concentrations of CXCL13 in CIS patients with and without oligoclonal bands (OCB) in CSF. P-value refers to comparison of CIS patients with and without OCB. C) Boxplot shows CSF concentrations of CXCL13 in CIS patients with and without positive MRZR (“MRZ reaction”, intrathecal antibody production against measles, rubella and varicella zoster, with two antibody indices ≥1.5). P-value refers to comparison of CIS patients with and without MRZR. D) Boxplot shows CSF concentrations of CXCL13 in CIS patients with and without gadolinium-enhancing lesions in T1-weighted MRI. P-value refers to comparison of CIS patients with and without gadolinium-enhancing lesions. A)–D) The box represents the 25^th to 75^th quartile, the whiskers represent the range, and the horizontal line in the box represents the median. Black dots beyond the whiskers indicate outliers, d = number of patients with detectable CXCL13 concentrations in CSF.

Figure 3. Relation of CXCL13 to cell count.

Dot plot shows CSF CXCL13 in patients with CIS plotted against CSF leucocyte count. Straight line represents regression line; correlation was significant (p<0.001, R = 0.59).