Cytokines cannot distinguish MS

Mouzaki A, Rodi M, Dimisianos N, Emmanuil A, Kalavrizioti D, Lagoudaki R, Grigoriadis NC, Papathanasopoulos P.

Immune Parameters That Distinguish Multiple Sclerosis Patients from Patients with Other Neurological Disorders at Presentation. 


PLoS One. 2015;10(8):e0135434.

BACKGROUND/AIM:Multiple sclerosis (MS) is an inflammatory, demyelinating disease of the central nervous system. Effector T helper cells, mainly Th1 and Th17, cytotoxic T-cells, B-cells, macrophages, microglia, and the cytokines they secrete, are implicated in the initiation and maintenance of a deregulated immune response to myelin antigens and the ensuing immune-mediated demyelination. In this study, we investigated whether signature cytokines exist in MS patients at presentation to gain an insight into the underlying immunopathogenic processes at the early stage of the disease.

METHODS: We collected serum and cerebrospinal fluid (CSF) samples from 123 patients at presentation, eventually diagnosed with MS or non-inflammatory (NIND) or inflammatory neurological diseases (IND) or symptomatic controls (SC). The levels of cytokines IFN-γ, TNF-α, TGF-β1, IL-2, IL-4, IL-6, IL-10 and IL-17 were measured, and cytokine ratios, such as Th1/Th2, Th1/Th17, and Type-1/Type-2, were calculated. All parameters were tested for their correlations with the intrathecal IgG synthesis.

RESULTS: Cytokine levels in CSF were lower than in serum in all the patients, with the exception of IL-6. Serum or CSF cytokine levels of MS patients did not differ significantly from NIND or SC, with the exception of serum IFN-γ and TNF-α that were significantly higher in NIND. IND patients presented with the highest levels of all cytokines in serum and CSF, with the exception of serum IL-10 and CSF IL-17. MS patients had a significantly lower serum Th1/Th2 ratio compared to the NIND and IND groups, and significantly lower serum Type-1/Type-2, IFN-γ/IL-10 and CSF Th1/Th17 ratios compared to IND patients. MS patients had a significantly higher CSF IL-17/IL-10 ratio compared to IND patients. The IgG index was higher in MS patients compared to the control groups; the differences reached statistical significance between the MS and the NIND and SC groups. Reiber-Felgenhauer analysis of the QIgG and QAlb indices revealed higher intrathecal IgG synthesis in MS patients, and higher blood-CSF barrier dysfunction in IND patients. The IgG index correlated with CSF IL-4 in MS patients only.

CONCLUSIONS: We found no signature cytokines or profiles thereof in MS patients at presentation. Only IND patients presented with a clear Th1 cytokine polarization in serum and CSF. The parameters that distinguished MS patients from patients with other neurological disorders were IgG intrathecal synthesis, the IgG index and its correlation with CSF IL-4 level





This is what I call a negative publication that normally would not have seen the light of day. Kudos to the authors for paying for this work to be published in an open-access journal - it doesn't come cheap at $1,350/article.

There are several reasons for the negative findings aside from the obvious (i.e. that there is no true relationship between immune parameters and MS), and includes sensitivity of cytokine assays which become quite tricky to measure at the lower ends (if you look at the measurement range, the y-axis on average is around 100 pg/ml; which is equivalent to 0.0000001 mg/ml) and the age of the samples (cytokines are quite labile and can disappear over time and through repeated freeze/thaw of the samples).

Having said this, MS is defined by the presence of inflammatory activity in CSF samples with IgG and oligoclonal bands being the commonest thing found in MS patient samples (and also noted in this publication), and in my books the most useful test to perform.

Inflammation also occurs outside of the typical inflammatory disorders, such as stroke and neuropathy, and therefore there will be people in the NIND group with very high cytokine levels (as noted in the plots for IL-2, IL-10 and IL-4 which have long whiskers to their box-plots). It is reassuring to note that the controls in most cases have low values, and these are not healthy controls but those perceived to have nothing serious going on with them. What is normal anyway?

Labels: , ,