The complex relationship between oligoclonal bands, lymphocytes in the cerebrospinal fluid, and immunoglobulin G antibodies in multiple sclerosis: Indication of serum contribution

Abstract

Introduction: Intrathecal immunoglobulin G (IgG) and oligoclonal bands (OCBs) are the most consistent and characteristic features of Multiple Sclerosis (MS). OCBs in MS are considered products of clonally expanded B cells in the cerebrospinal fluid (CSF), representing the sum of contributions from B cells in the brain. However, large amounts of IgG can be eluted from MS plaques in which lymphocytes are absent, and there is no correlation between levels of plaque-associated IgG and the presence of lymphocytes. It is calculated that it would take 3.2 billion lymphocytes to generate such large amounts of intrathecal IgG (30 mg in 500 ml CSF) in MS patients. Therefore, circulating lymphocytes in CSF could only account for <0.1% of the extra IgG in MS.

Methods: We analyzed clinical laboratory parameters from sera and CSF of 115 patients including 91 patients with MS and 24 patients with other inflammatory central nervous system (CNS) disorders (IC). We investigated the relationship between oligoclonal bands, IgG antibodies, CSF cells, IgG Index, albumin, and total protein.

Results: MS patients have significantly elevated serum concentrations of IgG antibodies, albumin, and total protein, lower levels of lymphocytes, albumin, and total protein in the cerebrospinal fluid, but no difference in CSF IgG concentration compared to those with other inflammatory neurological disorders. Furthermore, in MS there was no linear relationship between the numbers of OCBs, CSF lymphocytes, CSF IgG, and IgG Index, and between serum IgG and serum albumin, but significant correlation between IgG in CSF and serum, and between CSF IgG and CSF albumin.

Conclusion: There are unique differences between MS and patients with other inflammatory neurological disorders. Our data suggest that in MS patient (a) B cells and their products in the CSF may not be the sole source of intrathecal IgG; (b) oligoclonal bands may not be the products of single B cell clones in the CSF; and (c) there is a strong connection between serum components in the peripheral circulation and the central nervous system.

Fig 1. Summary of mean difference in IgG, albumin, and total protein concentrations between MS and IC patients in CSF and serum.

MS patients have significantly lower levels of IgG, albumin, and protein concentrations in the CSF (Fig 1A), but significantly higher levels in serum (Fig 1B) compared to IC patients. *p values <0.05.

Fig 2. Similar non-linear relationships exist between number of OCBs and CSF lymphocytes, CSF IgG, and IgG Index in MS.

A. No linear relationship between number of OCBs and CSF lymphocytes (n = 61) were observed. However, when CSF number of cells = 7, the association changed from an increasing linear association (r = 0.22, p = 0.09) to a decreasing linear association (r = -0.27, p = 0.93). B. No linear relationship between number of OCBs and CSF IgG concentration (n = 63). C. No linear relationship between number of OCBs and IgG Index. However, there is a highly significant positive correlation when IgG Index ≤ 1.14 (r = 0.394, slope = 11.45, p = 0.002), and the correlation becomes decreasing linear association when IgG Index >1.14, although not significant (r = -0.1381, slope = -6.09, p = 0.68). D. No association between CSF IgG concentration and number of OCBs when number of OCBs > 2. Spearman’s correlation coefficient r = 0.12, p = 0.35.

Fig 3. CSF IgG and serum IgG are highly correlated in MS.

A. Significant correlation between CSF IgG and serum IgG concentrations is present in MS patients (r = 0.33, p = 0.0015, n = 90); B. In IC patients, CSF IgG is correlated with serum IgG concentrations (r = 0.37, p = 0.1, n = 20).

Fig 4. Immunoblot data demonstrate a near perfect correlation between CSF IgG and serum IgG in MS.

Slot blot immune-assays with paired CSF and serum were performed with 15 MS and 8 IC. Equal amounts IgG in CSF and sera loaded onto nitrocellulose membranes were probed with anti-IgG (H+L) antibody-HRP followed by chemiluminescent detection. A. A near perfect correlation between CSF IgG and serum IgG concentration is present in MS (r = 0.99, p < 0.0001, n = 15), and in IC (r = 0.88, p = 0.045, n = 8) (B). C. There is no significant difference in the ratio of CSF IgG to serum IgG in MS compared in IC (p = 0.313).

Fig 5. Strong linear correlation between CSF IgG and CSF albumin is found only in MS CSF but not in MS serum.

A. Highly significantly correlation exists between concentration of CSF IgG and CSF albumin in MS (r = 0.55, p < 0.0001, n = 70). B. No correlation between serum IgG and serum albumin in MS (r = -0.08, p = 0.53, n = 68). In IC patients, highly correlated relationship was found in both CSF (C) and serum (D).

Fig 6. Summary of relationships between the laboratory parameters in CSF and serum of MS.

In MS, highly correlated relationships are present between CSF IgG, CSF albumin, and CSF total protein. In contrast, such correlations are not observed in MS serum. Further, there are no direct linear relationship between the number of OCBs, CSF IgG concentration, and CSF cells. Black arrows indicate correlation, and the red ‘not equal’ sign indicates that no correlations were found.