Markers for different glial cell responses in multiple sclerosis: clinical and pathological correlations

Disease progression in multiple sclerosis occurs within the interface of glial activation and gliosis. This study aimed to investigate the relationship between biomarkers of different glial cell responses: (i) to disease dynamics and the clinical subtypes of multiple sclerosis; (ii) to disability; and (iii) to cross-validate these findings in a post-mortem study. To address the first goal, 51 patients with multiple sclerosis [20 relapsing remitting (RR), 21 secondary progressive (SP) and 10 primary progressive (PP)] and 51 neurological control patients were included. Disability was assessed using the ambulation index (AI), the Expanded Disability Status Scale score (EDSS) and the 9-hole PEG test (9HPT). Patients underwent lumbar puncture within 7 days of clinical assessment. Post-mortem brain tissue (12 multiple sclerosis and eight control patients) was classified histologically and adjacent sites were homogenized for protein analysis. S100B, ferritin and glial-fibrillary acidic protein (GFAP) were quantified in CSF and brain-tissue homogenate by ELISA (enzyme-linked immunosorbent assay) techniques developed in-house. There was a significant trend for increasing S100B levels from PP to SP to RR multiple sclerosis (P < 0.05). S100B was significantly higher in RR multiple sclerosis than in control patients (P < 0.01), whilst ferritin levels were significantly higher in SP multiple sclerosis than in control patients (P < 0.01). The S100B : ferritin ratio discriminated patients with RR multiple sclerosis from SP, PP or control patients (P < 0.05, P < 0.01 and P < 0.01, respectively). Multiple sclerosis patients with poor ambulation (AI greater than or equal to7) or severe disability (EDSS > 6.5) had significantly higher CSF GFAP levels than less disabled multiple sclerosis or control patients (P < 0.01 and P < 0.001, respectively). There was a correlation between GFAP levels and ambulation in SP multiple sclerosis (r = 0.57, P < 0.01), and between S100B level and the 9HPT in PP multiple sclerosis patients (r = -0.85, P < 0.01). The post-mortem study showed significantly higher S100B levels in the acute than in the subacute plaques (P < 0.01), whilst ferritin levels were elevated in all multiple sclerosis lesion stages. Both GFAP and S100B levels were significantly higher in the cortex of multiple sclerosis than in control brain homogenate (P < 0.001 and P < 0.05, respectively). We found that S100B is a good marker for the relapsing phase of the disease (confirmed by post-mortem observation) as opposed to ferritin, which is elevated throughout the entire course. GFAP correlated with disability scales and may therefore be a marker for irreversible damage. The results of this study have broad implications for finding new and sensitive outcome measures for treatment trials that aim to delay the development of disability. They may also be considered in future classifications of multiple sclerosis patients.