During the last decades, the effort of establishing satisfactory biomarkers for multiple sclerosis has been proven to be very difficult, due to the clinical and pathophysiological complexities of the disease. Recent knowledge acquired in the domains of genomics-immunogenetics and neuroimmunology, as well as the evolution in neuroimaging, has provided a whole new list of biomarkers. This variety, though, leads inevitably to confusion in the effort of decision making concerning strategic and individualized therapeutics. In this paper, our primary goal is to provide the reader with a list of the most important characteristics that a biomarker must possess in order to be considered as reliable. Additionally, up-to-date biomarkers are further divided into three subgroups, genetic-immunogenetic, laboratorial, and imaging. The most important representatives of each category are presented in the text and for the first time in a summarizing workable table, in a critical way, estimating their diagnostic potential and their efficacy to correlate with phenotypical expression, neuroinflammation, neurodegeneration, disability, and therapeutical response. Special attention is given to the “gold standards” of each category, like HLA-DRB1* polymorphisms, oligoclonal bands, vitamin D, and conventional and nonconventional imaging techniques. Moreover, not adequately established but quite promising, recently characterized biomarkers, like TOB-1 polymorphisms, are further discussed.
Multiple Sclerosis (MS) is the most common reason of neurological disability among young adults. Its clinical course varies greatly, reflecting complexity in pathophysiology. Different mechanisms of inflammation-demyelination, axonal damage-neurodegeneration, gliosis, and remyelination-repair combine together in various degrees (influenced by idiosyncratic factors) to create a unique clinical result for each patient. Identifying those idiosyncratic factors, as well as understanding which mechanism is prominent in each case, is the first step towards a rational therapeutical choice. Thus, guiding research towards distinguishing reliable biomarkers for every independent MS pathogenic factor is of primary importance.
An adequate definition of the term “biomarker” would be as follows: “Biomarker is a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes or pharmacological responses to a therapeutical intervention” [
Biomarkers in MS.
(A) Diagnostic biomarkers (criteria i, iv, v, and vi) | ||
| ||
(1) Genetic-immunogenetic | ||
HLA-DRB1*1501 | +++ Risk for MS | See also B, E |
DR3 and DR4 haplotypes | ++ Risk for MS | |
HLA-DRB1*04 | ++ Risk for MS | |
HLA-DRB1*0401 | + Risk for high familial autoimmunity in MS patients | See also F |
HLA-DQ1*0102 | + Risk for MS, in coexistence with HLA-DRB1*1501 | |
HLA-DPB1*0501 | + Risk for opticospinal MS | |
HLA-DPB1*0301 | + Risk for opticospinal MS | |
IL2RA and IL7RA polymorphisms | + Risk for MS | |
EVI5, CD58, KIAA0350, and RPL5 polymorphisms | +/− Risk for MS | |
(2) Laboratorial | ||
OCB IgG | +++ But with low specificity | See also E |
KFLC | +++ But with low specificity | See also E |
MRZ reaction | +++ Higher specificity than OCB IgG | See also E, F |
Anti BRRF2, anti EBNA-1 | ++ | See also B, C |
Anti MBP 48–70 and 85–170 | + | See also B, E |
Anti MBP 43–68 and 146–170 | + Differential diagnosis with OND’s | See also B, E |
MBP/MOG conformational epitopes antibodies | + But low specificity | See also B, E, F |
VEGF-A | + Lower CSF levels in all disease forms, but low specificity | See also D, E |
Vitamin D | +++ Lower levels, higher risk for MS | See also C, F |
TRECs | + Lower serum levels in all disease forms, but low specificity | See also B |
CSF levels of lipocalin 2 | + Higher CSF levels in MS, but low specificity | See also F |
AR | +++ Differential diagnosis of MS and NMO | See also C, E |
NO and NO metabolites | + Higher CSF and serum levels in MS, but low specificity | See also C, E |
NF-L | ++ Higher CSF levels in MS patients | See also C, F |
NAA | +++ Differential diagnosis of RRMS and NMO | See also D, E |
GFAP | +++ Differential diagnosis of MS and NMO | See also C, E |
|
+ Differential diagnosis of MS and NMO | See also C, E |
Nogo-A | ++ For MS forms with prominent neurodegenerative element | See also D |
(3) Imaging | ||
Contrast-enhanced T1 lesions | +++ | See also C |
Hyperintense T2-weighted lesions | +++ | See also C, D, E |
Corpus callosum DTI abnormalities | ++ Early diagnostic biomarker | See also E |
MRS findings (glutamate/choline) | +++ | See also C, D, E |
PET | ++ But still experimental | |
EPs |
+++ |
See also C, |
SSR | ++ Autonomic dysfunction assessment in MS patients | See also E |
| ||
(B) Biomarkers of phenotypical expression (criteria ii, iv, v, and vi) | ||
| ||
(1) Genetic-immunogenetic | ||
HLA-DRB1*1501 | +++ Early disease onset | See also A, E |
HLA-DRB1*1501 | + Risk for cognitive decline | |
HLA-DRB1*01 | ++ Protection against malignant disease form | |
ApoE |
++ Greater risk for mental disorders | |
(2) Laboratorial | ||
OCB IgM against myelin lipids | +/− Aggressive disease course | See also E |
EBV antibodies | + Early disease onset | See also A, C |
Anti-MBP | +++ ADEM-like onset in childhood MS | See also A, E |
Anti-MOG | +++ Childhood MS, ADEM, isolated optic neuritis, anti-AQP4 (−) NMO | See also A, E, F |
rMOG index | +++ Progressive disease forms | |
IL-6 serum levels | +++ Age at onset | See also C |
TRECs | ++ Lower levels PPMS | See also A |
Amyloid- |
++ Lower levels, higher risk for mental disorders | |
(3) Imaging | ||
UCCA atrophy | +++ Progressive disease forms | See also E |
NAGM DTI abnormalities | +++ Progressive disease forms | |
| ||
(C) Biomarkers of demyelination-neuroinflammation-relapse (criteria i, ii, iii, iv, v, and vi) | ||
| ||
(1) Genetic-immunogenetic | ||
TOB1 | +++ Underexpression, higher Th1 and Th17 percentage | See also E |
(2) Laboratorial | ||
EBV antibodies | + Higher inflammatory activity | See also A, B |
CXCL13 | ++ Mobilizes B-cells, T-helper cells | |
CXCL12 | +/− Neuroprotection against inflammation in EAE/ experimental | |
IFN- |
+++ Th1 immune response | |
IL-1 levels imbalance | + Triggering factor for neuroinflammation | |
IL-6 | +++ B-cell and T-cell immunity link, Th17 immune response triggering factor |
See also B |
IL-10 −592 position polymorphisms | ++ Regulation of CNS autoimmunity | |
IL-15 | ++ BBB disruption, enhanced CD8(+) T cytotoxicity | |
IL-33 | + Increase in IFN- |
|
sICAM-1 | ++ Higher levels, higher inflammatory activity |
See also F |
sVCAM-1 | +++ Higher levels in NMO than MS—marker of BBB disruption | See also F |
Laminin 411 | ++ TH-17 enhancement | |
|
++ Correlation with gadolinium-enhanced lesions during CIS | See also E, F |
Osteopontin | ++ Serum and CSF elevation during relapse | |
Fetuin-A | +++ Overexpression in active demyelinating lesions | See also F |
Vitamin D | +++ High levels, anti-inflammatory role—lower radiological disease activity | See also A, F |
CSF mature B-cells/plasma-blasts | ++ Bigger accumulation, higher inflammatory activity | |
CXCR3 | ++ Helps T-cells to enter the brain | |
CX(3)CR1 | ++ CD4(+)CD28(−) cytotoxic cells biomarker | |
CSF CCR2(+)CCR5(+) T cells | +++ Increase during MS relapse—osteopontin enhancement | |
CD56 Bright NK | ++ Remission phase | |
AR | +++ Biomarker of BBB disruption | See also A, E |
MMP-9 | ++ Higher CSF levels during relapse | |
Ninjurin-1 | ++ Upregulation in active demyelinating lesions | |
MBP and fragments | +++ Higher CSF levels during relapse | See also F |
|
+++ Over-expression in active demyelinating lesions | |
NO and metabolites | ++ | See also A, E |
7-Ketocholesterol | ++ | |
Glutamate | +++ Higher levels in active demyelinating lesions | |
Cystine/glutamate antiporter | + Over-expression in active demyelinating lesions | |
NF-L | +++ Higher CSF levels, especially the 3rd week after relapse onset | See also A, F |
GFAP | ++ Higher levels during relapse | See also A, E |
S100B | +/− Higher CSF levels during MS/NMO relapse | See also A, E |
N-CAM | + CSF elevation at remission onset | |
BDNF | ++ Lower levels inhibit demyelination and axonal loss | See also D, E, F |
(3) Imaging | ||
Contrast-enhanced T1 lesions | +++ Active lesions | See also A |
Hyperintense T2-weighted lesions | ++ Combination of different mechanisms | See also A, D, E |
MTR decrease | + Demyelination and axonal loss combined | See also D |
DTI abnormalities | ++ Combination of different mechanisms | See also D, E |
MRS findings (especially changes in glutamate and choline) | +++ Active lesions | See also A, D, E |
DTS | ++ Promising but still experimental | See also D |
EP’s delayed conduction | ++ Demyelination biomarker | See also A, D, E |
| ||
(D) Biomarkers of axonal loss-neurodegeneration (criteria i, iv, v, and vi) | ||
| ||
(1) Laboratorial | ||
VEGF-A | ++ Lower levels, higher risk for neurodegeneration | See also A, E |
14-3-3 | +/− Axonal loss | |
NAA | +++ Axonal loss | See also A, E |
BDNF | ++ Lower levels inhibit demyelination and axonal loss | See also C, E, F |
Nogo-A | +++ Higher CSF levels, failure in axonal repair | See also A |
(2) Imaging | ||
RNFL thinning | +++ Axonal loss in the optic nerve | See also E, F |
Hyperintense T2-weighted lesions | ++ Combination of different mechanisms | See also A, C, E |
Black holes | +++ Axonal loss | See also E |
MTR decrease | ++ Demyelination and axonal loss combined | See also C |
DTI abnormalities | ++ Combination of different mechanisms | See also C, E |
MRS findings (especially NAA) | ++ | See also A, C, E |
DTS | +++ Promising but still not widely accessible | See also C |
Visual and motor EPs | ++ | See also A, C, D |
| ||
(E) Prognostic biomarkers—biomarkers of disability progression (criteria ii, iv, v, vi, and viii) | ||
| ||
(1) Genetic-immunogenetic | ||
HLA-DRB1*1501 | +/− Early progression from RRMS to SPMS | See also A, B |
HLA-DRB1*1501 | + Worst brain atrophy measures | |
HLA-DQB1*0301 | + Worst brain atrophy measures | |
HLA-DQB1*0602 | + Worst whole and gray matter atrophy measures | |
TOB1 | +++ Early conversion from CIS to CDMS | See also C |
(2) Laboratorial | ||
OCB IgG | +++ Conversion from CIS to CDMS | See also A |
KFLC | +++ Conversion from CIS to CDMS | See also A |
OCB IgM | +/− Bad prognostic biomarker | See also B |
MRZ reaction | +++ Conversion from CIS to CDMS | See also A, F |
Anti-MBP | +/− Conversion from CIS to CDMS | See also A, B |
Anti-MOG | +/− Conversion from CIS to CDMS | See also A, B, F |
AR | ++ Marker of clinical severity in NMO | See also A, C |
VEGF-A | ++ Lower levels, progression from RRMS to SPMS | See also A, D |
NO and NO metabolites | ++ Higher CSF levels, longer relapses/higher disability progression rates | See also A, C |
NF-H | +++ Higher CSF levels, progressive forms/bad prognostic biomarker | |
NF-H and tau | +++ Combined high CSF levels, conversion from CIS to CDMS | |
Tubulin/actin | ++ Higher CSF levels, progressive forms/worst disability scores | |
NAA | +++ Lower CSF levels, progressive forms/worst disability scores | See also A, D |
GFAP | ++ Higher CSF levels, progressive MS forms/worst disability scores |
See also A,C |
S100B | + Disability progression in NMO | See also A,C |
BDNF | ++ Lower CSF levels in SPMS patients | See also C, D, F |
Unblocked |
+ Prognostic factor of risk for PML | See also C, F |
(3) Imaging | ||
RNFL thinning | + Correlation with brain atrophy measures and disease progression | See also D, F |
Hyperintense T2-weighted lesions | +/− | See also A, C, D |
Black holes | +/− | See also D |
Whole brain atrophy measures | ++ Worsening rates at MS onset, prognostic biomarker of disability after 8 years | |
Gray matter atrophy measures | +++ Higher worsening rates, progressive forms/early CIS conversion to RRMS | |
UCCA atrophy | ++ Progressive forms, good correlation with EDSS, bad prognostic in RRMS | See also B |
DTI abnormalities | +++ Early prognostic biomarker of relapse | See also C, D |
Corpus callosum DTI abnormalities | +++ Bad prognostic biomarker | See also A |
Spinal cord DTI abnormalities | +++ Good correlation with EDSS scores | |
Early MRS abnormalities | ++ Bad prognostic biomarker | See also A, C, D |
Combined EPs | +++ Good prognostic biomarker, especially for benign disease forms | See also A, C, D |
SSR | ++ Correlation with higher EDSS scores | See also A |
| ||
(F) Biomarkers of therapeutical response (criteria i, iv, v, vi, and vii) | ||
| ||
(1) Genetic-immunogenetic | ||
HLA-DRB1*0401, 0408, 1601 | +++ Higher risk for developing neutralizing antibodies against IFN-B | See also A |
(2) Laboratorial | ||
MRZ reaction | ++ B-cell immunity targeted therapy | See also A, E |
Anti-MOG | ++ B-cell immunity targeted therapy | See also A, B, E |
Fetuin-A | +++ Decreased CSF levels in Natalizumab responders | See also C |
MBP | +++ Decrease in CSF levels in methylprednizolone responders | See also C |
CSF lipocalin 2 | ++ Decreased CSF levels in Natalizumab responders | See also A |
Unblocked |
+++ Therapeutical response to Natalizumab | See also C, E |
NF-L | +++ Normalized CSF levels in Natalizumab responders | See also A, C |
BDNF | +++ CSF elevation in Glatiramer Acetate responders | See also C, D, E |
TRAIL | ++ Serum levels good predictors of response in IFN-B | |
MxA | ++ Serum levels good predictors of response in IFN-B | |
sVCAM | ++ CSF alterations in IFN-B responders | See also C |
Th17 immune profil | +/− Immune response exacerbation by IFN-B | |
Vitamin D | +++ Increased levels in IFN-B responders | See also A, C |
sICAM-1 | + Lower levels in Cladribine responders | See also C |
sE-Selectin | + Lower levels in Cladribine responders | |
(3) Imaging | ||
RNFL | +++ Biomarker of therapeutical efficacy for several agents | See also D, E |
Classification of biomarkers. +++ very strong correlation, ++ strong correlation, + modest correlation, and +/− controversial correlation. Criteria used for classification., (i) Biological rationale; (ii) clinical rationale; (iii) predictability of disease initiation, reactivation or progression, or of disease differentiation; (iv) sensitivity and specificity; (v) reproducibility; (vi) practicality; (vii) correlation with therapeutical outcome; (viii) correlation with prognosis and disability. Biomarkers of more than one category are indicated in the third column.
Its value as a means of providing biomarkers is undebatable, due to its natural proximity to the Central Nervous System (CNS). The levels of a CSF biomarker cannot be influenced by liver or kidney function.
On the other hand, the invasiveness of the collecting method narrows the potential of multiple measurements. Circadian fluctuation in CSF’s production rate dictates the necessity of standardizing the time of performing a lumbar puncture [
It is easier to collect compared to CSF, with fewer limitations regarding safety. There is also a limitation concerning circadian fluctuations. Interleukin-(IL-)6 has maximum concentration at 08:00 a.m., and minimum at 22:00 p.m. [
It is the easiest material to collect, even in a 24-hour basis, overcoming the obstacles of fluctuations previously mentioned. Bacterial colonization of the urinary tract though can distort the measurements. MS patients with bladder dysfunction may regulate the amount of the fluids taken in a daily basis, affecting the quantity of produced urine.
There have been previous efforts in measuring OligoClonal Bands (OCBs) in tears, with results comparable to those of CSF [
It has served as a means of specifying soluble Human Leucocyte Antigens (HLAs) type II [
Magnetic Resonance Imaging (MRI) was considered, till recently, as the most accurate imaging method for MS. There are though considerable difficulties in correlating MRI findings with disability progression. Certain novel ambitious techniques promise to overcome all these problems. The most important of them are further analyzed in paragraph 7.
Although in depth analysis of all the different techniques providing biomarkers for MS is out of the scopes of this paper, a short comment in relevance to the most important of them is going to follow. An example biomarker will be provided in each case, extracted from the literature and references sited in this paper.
It is the most common method for specifying soluble proteins. The term refers to a solid-phase enzyme immunoassay which gives the ability to detect the presence of a substance, in a liquid or wet sample. Performing an ELISA technique requires the use of at least one specific antibody against the antigen under investigation. There are numerous different commercial kits available. By the use of standardization protocols, the assays are easy to reproduce by different laboratories. ELISA techniques give also the opportunity to analyse samples that are properly stored (i.e., IL-1 and IL-6 measurements; see Section
This is a highly sensitive technique which gives the opportunity to specify, in vivo, proteins of variable size, haptens, and antibodies. Immunofluorescence techniques use the affinity of a certain antibody against its epitope, which is made visible through microscope by the use of proper fluorescent dyes. Further categorization in direct and indirect methods refers to the use of one or two antibodies. Time-resolved immunofluorescence assays display high sensibility, tracking down proteins in very small concentrations. Like ELISA, this technique is highly reproducible and accessible by different laboratories. Major drawback is the danger of photobleaching, as a result of prolonged light exposure (i.e., sICAM-1 and sVCAM-1; see Section
It is a laser-based method applied in the study of cellular subpopulations and biomarkers by the form of surface antigens, DNA and RNA variations, protein expression, enzymes, and intracellular antigens. Several techniques, like fluorescein dyes and intracellular cytokine dying, allow the collection of functional data for specific cellular populations. The major drawback of these assays is that sample handling may influence the results. Because of that, comparisons between different biomarker measurements are at risk of being inaccurate (i.e.,
PCR provides the researcher with the ability to detect trivial amount of genetic information. Basic components of the method are primers (short DNA fragments, sequences complementary to the target area), as well as a DNA polymerase, which enhances the replication. Quantitative PCR methods include competitive, noncompetitive, and real-time PCR. Posttranslational modifications are not taken into account by PCR techniques. Subsequently, the results may not be fully compatible with the functional state, in vivo. Cost and sample handling limitations should also be considered (i.e., HLA-DRB1* polymorphisms; see Section
The term nephelometry refers to a technique of estimating protein concentrations in different bodily fluids. The liquid sample is beamed by light by a certain angle, and afterwards the degree of light scatter is estimated. Nephelometry techniques are widely performed, reliably reproducible by many different laboratories (i.e., CSF albumin; see Section
The term refers to a protein detection method, which uses gel electrophoresis of a sample and subsequent dying of the protein target by the use of a specific antibody, on a membranic surface. Several final detection techniques have been developed, namely, colorimetric and fluorescent (i.e.,
This is another electrophoresis technique, which takes advantage of the phenomenon of the different isoelectric point between different molecules, in order to separate them. For this purpose, acrylamide gels with pH gradient are used (i.e., detection of CSF oligoclonal bands-kappa and lambda free chains; see Sections
The general term “-omics” refers to a group of rapidly emerging novel technologies that give the opportunity of large-scale analysis and identification of candidate biomarkers in multiple levels of cell biology (DNA, RNA, proteins, lipids, metabolites, and epigenetic modifications). Subsequently, the “-omics” technologies are further categorized in: genomics: large-scale studies of the whole DNA sequence (i.e., vitamin D Receptor Element recognition; see Section transcriptomics: genome-wide studies of RNA sequences. Two main types of transcriptomics technologies are in common use, microarrays and next generation sequencing (i.e., TOB-1 gene downregulation; see Section proteomics: large-scale studies of protein distribution (i.e., Ninjurin-1; see Section lipidomics: recognition studies of important cellular lipid pathways. Recent knowledge implicates specific CNS lipid epitopes in the generation of anti-lipid antibodies in MS (see also IgM against myelin lipids Section metabolomics: studies of important metabolic pathways, as a result of MS specific pathogenic mechanisms (i.e., N-acetyloaspartate CSF measurements; see Section epigenomics: large-scale studies of epigenetic modifications. They explore the potential influence that alterations in chromatin architecture may have in MS susceptibility. One such study reported similar methylation profiles between twins discordant for MS [
In this paper biomarkers are further categorized in three subgroups for reasons of systematic approach, and according to their pathophysiological implication in MS pathogenesis: genetic/immunogenetic: biomarkers specified via genomics and immunogenetic techniques; laboratorial: all other biomarkers that can be measured in body fluids; imaging: biomarkers provided by imaging techniques.
The fact that genetic factors may influence MS was already known by epidemiological studies of previous decades. Recent research, using modern techniques previously mentioned, led to the implication of multiple genetic loci, with polymorphisms of Major Histocompatibility Complex (MHC) antigens having the primary role.
Polymorphisms in HLA class II antigens seem to be decisive in attributing genetic burden for MS. Initial studies found positive correlation between DRB1*1501-DRB5*0101-DQA1*0102-DQB1*0602 haplotype and disease frequency. Multiple recent researches, conducted in many MS cohorts, made clear that HLA-DRB1*1501 is the mainly responsible allele for attributing genetic risk in MS population [
Positive correlation of HLA-DRB1*1501 and negative correlation of HLA-DRB1*0405 alleles with OCB in the CSF of MS patients were established by observations in a Japanese cohort [
HLA-DRB1*15 was found to correlate positively with early onset MS [
Zivadinov et al. observed the following in a study of MS patients [ DRB1*1501 positive had worst brain atrophy scores and bigger T1 lesions’ burden in MRI; DQB1*0301 positive had worst brain atrophy scores and bigger T2 lesions’ burden; DQB1*0602 positive had worst grey matter atrophy scores.
In another study [ lower bigger white matter lesions; worst brain atrophy scores; impaired cognitive function.
HLA-DRB1*0401, 0408, and 1601 alleles correlate with greater risk of developing neutralizing antibodies against interferon beta (IFN-
Various genome-wide studies revealed many non-MHC single nucleotide polymorphisms as candidates for genetic burden augmentation in MS. Most of them though had only a modest effect on susceptibility. Polymorphisms of the IL2RA and IL7RA regions seem as the most promising at the moment [
TOB-1 gene has a role against T-cell multiplication, keeping autoreactive cells in a dormant state. Its degreased expression leads towards a more intense immune response (higher percentage of Th1 and Th17 cells and lower percentage of T-regulatory cells). TOB-1 polymorphisms represent an independent factor influencing the progression from clinically isolated syndrome (CIS) to clinically definite multiple sclerosis (CDMS) [
ApoE is a protein regulating lipid homeostasis, located mostly in astrocytes. Carrying
Positive OCB IgG in the CSF of patients with CIS was found to duplicate the risk of progression in CDMS in a 4-year observation period [
Some researchers consider them as a bad prognostic biomarker, correlating with disability progression both qualitatively and quantitatively (IgM index) [
KFLC high CSF levels have been repeatedly reported in MS. In comparison to OCB IgG, slightly higher sensitivity with slightly lower specificity has been found [
MRZ IgG reaction in CSF displays, compared to OCB IgG, a higher specificity for MS diagnosis and higher prognostic value of progression from CIS to CDMS [
Cepok et al. reported a high percentage of IgG antibodies against protein epitopes BRRF2 and EBNA-1 of the virus, in the serum and CSF samples from MS patients [
Their diagnostic and prognostic values in MS remain highly controversial. Initially, they were regarded as satisfactory predictors of conversion from CIS to CDMS [
Considerably higher levels of serum autoantibodies against the fragments 48–70 and 85–170 of MBP were found in MS patients compared to healthy controls [
Finally, children with CIS were found to have high titres of anti-MOG in a percentage of 30%–40% [
Chemokine CXCL13 mobilizes B-cells and T-helper cells towards active demyelinating lesions by interacting with CXCR5 receptor. Consistent correlation of CXCL13 CSF levels with CSF B-cells, plasmablasts, and intrathecal Ig synthesis has been reported [
Inflammatory activity in active demyelinating lesions leads to the liberation of many different cytokines that can be used as biomarkers of disease activity. Proinflammatory cytokines in the periphery primarily originate from T- and B-cells, whereas B-cells seem to be mainly responsible for their intrathecal production in RRMS [
Moreover, studying IL-1 levels in mice led to the conclusion that any imbalance in the IL-1 signalling (increased or decreased) may lead to CNS demyelination [
Proinflammatory cytokines cause a rise in CSF expression of sICAMs. High levels of ICAM-1 molecule correlate positively with higher disease activity [
Osteopontin is a macrophage derived phosphoprotein which enhances IFN-
Fetuin-A (alpha2 Hermans Schmid glycoprotein) is a calcium-regulating surface glycoprotein. Protein’s coding m-RNA is overexpressed in MS patients’ CNS, resulting in its high concentrations in active demyelinating lesions. Fetuin-A seems to antagonize anti-inflammatory TGF-
VEGF-A is a factor of angiogenesis with neuroprotective properties. Diminished m-RNA expression of VEGF-A in serum monocytes of patients with SPMS compared to RRMS patients has been reported. VEGF-A could serve as biomarker of progression from RRMS to SPMS [
The vitamin’s potential pathogenic role in MS can be deducted by multiple previous epidemiological studies that showed correlation of latitude and sun exposure with relative risk for developing the disease. Vitamin D suppresses Th1 immune response in multiple levels and enables the production of many neurotrophic factors. 25-Hydroxyvitamin D levels in untreated MS patients correlate inversely with radiologic disease activity [
Interestingly, Stewart et al. recently concluded that part of IFN-
Evidence from animal model research also implicates vitamins other than vitamin D in MS pathogenesis. Fat-soluble vitamins A and E are considered as modulators of disease activity [
Mature B-cells and plasma-blasts were found to accumulate in the CSF of RRMS patients, correlating positively with higher disease activity in the MRI [
Autoreactive memory T-cells enter CNS with the help of CXCR3 cytokine receptor. CXCR3 has a low diagnostic specificity for MS, due to its high levels in many other inflammatory disorders [
RRMS patients in a remission phase display high levels of CD-56 surface antigen in their NK cells (>36%). CD56 bright NK cells may regulate T-cell survival in MS [
TRECs are intracellular by-products of T-cell receptor remodelling, gradually rejected via homeostatic mechanisms. Existence of TRECs inside a T-cell is a good marker of naivety. Thymic gland’s functional state can be estimated by the percentage of naïve T-cells in the peripheral blood. A definite reduction of TRECs in MS patients has been reported, indicative of thymic dysfunction in the disease. Naïve T-cells are found further diminished in patients with primary progressive multiple sclerosis (PPMS), compared to RRMS [
Lipocalins are a family of proteins that transport small hydrophobic molecules, taking part in several processes of the immune system. The gene encoding for lipocalin 2 was found upregulated during relapses in the EAE model of MS, mainly originating from neutrophils infiltrating the Choroid Plexus (CP), as well as from astrocytes in affected regions. CSF levels of lipocalin 2 were found increased in two different MS cohorts. Decrease, subsequent to clinical response after Natalizumab treatment, was also reported [
BBB disruption is an early feature of lesion formation, leading to edema, excitotoxicity, and entry of serum proteins and inflammatory cells inside CNS. Intercellular endothelial tight junctions breakdown possesses a primary role between events leading to BBB and blood-cerebrospinal fluid barrier (BCB) disruption [
Apart from CSF production, CP is actually considered as important regulator of CNS autoimmunity, displaying properties of early BCB disruption site, allowing sentinel T-cells to enter noninflamed regions [
Various measures of BBB permeability have been proposed, like the CSF : serum albumin ratio (AR). AR levels are constantly higher in NMO in comparison to MS, and display correlation with clinical severity only in NMO [
Serum and CSF MMPs levels are constantly elevated during MS relapse. MMP-9 levels have been found elevated in patients with RRMS [
The degree of expression of the protein Ninjurin-1 by endothelial cells of the BBB and myeloid antigen-presenting cells (APCs) plays an important role in the transmigration and localization of the latter inside CNS, as it was made obvious by proteomic screen of human BBB cells. Ninjurin-1 was found up-regulated in active demyelinating lesions [
The CSF sICAM-1 levels from NMO patients were found to correlate adequately with other measures of BBB disruption, like the albumin quotient and the gadolinium-enhanced lesions in MRI [
The term refers to an endothelial proteinic system that plays role in the transmigration of monocytes through the BBB. Major components of this system are the proteins endothelin-1, endothelin type B receptor, and endothelin-converting enzyme-1 [
Further information about EVB infection is in Section
MBP and its fragments are found in large quantities in the CSF of most MS patients during a relapse (80%) [
Immunohistochemical analysis of demyelinating lesions revealed increased expression of this protein, comparatively to healthy myelin.
NO and its metabolites can cause mitochondrial damage and tissue hypoxia leading to further damage in MS lesions. High serum and CSF levels of NO in inflammatory neurological disorders were reported. Higher CSF concentrations were further correlated with higher disability progression rates in MS [
ROS damage oligodendrocytes and myelin through radical mediated oxidation. Myelin cholesterol breaks down to 7-ketocholesterol, whose levels in the CSF of MS patients have been reported to be elevated [
Extracellular levels of glutamate are normally regulated through its active reabsorption in oligodendrocytes. In active demyelinating lesions, homeostatic mechanisms are distorted resulting in extracellular glutamate accumulation that causes further axonal damage [
Neurofilaments are major axonal cytoskeleton proteins consisting of three subunits (light chain/NF-L, intermediate chain/NF-M, and heavy chain/NF-H). NF-L CSF levels in MS patients are considerably higher compared to healthy controls and ONDs patients, reaching their peak approximately three weeks after relapse onset [
On the other hand, NF-H chains seem to correlate better with disease progression, with significant elevation recorded only in progressive disease forms [
Tau is a cytoskeleton protein whose basic responsibility is microtubular stabilization. High CSF levels in MS patients have been reported. Simultaneous elevation in Tau and NF-H values in CSF, in patients with CIS, has a 70% predictive value of conversion to CDMS, which is superior to the predictive value of MRI [
Microtubules represent a major structural cytoskeleton component, consisting of two subunits, A- and B-tubulin. They are closely associated with tau protein and microfilaments, especially actin. Elevated CSF tubulin and actin values have been reported in progressive disease forms, correlating well with disability measured by EDSS [
In Alzheimer’s disease, amyloid-
Apart from Creutzfeldt-Jacobs disease, elevated CSF values have been reported in 10%–30% of patients with RRMS [
NAA is an aminoacid, highly expressed in neurons, which transfers actively water molecules extracellularly against concentration gradient. Spectroscopy techniques revealed decreased NAA values in MS lesions, but also in NAWM, in conventional MRI. CSF-NAA reduction correlates adequately with disability progression [
On the contrary, serum and CSF NAA levels were significantly higher in RRMS patients, in comparison to healthy donors and NMO patients. Subsequently, NAA could be helpful in differential diagnosis between MS and NMO [
S100B is a calcium-binding protein, primarily expressed in astrocytes, whose CSF elevated values have been previously correlated with cerebral injury. There are reports of CSF elevation in RRMS patients [
GFAP is a structural protein of the astrocytes whose CSF levels increase in association with gliosis-astrocytosis. High CSF values have been found in SPMS patients, but rarely in RRMS patients, and seem to correlate well with disability progression [
Constant CSF elevation of N-CAM has been repeatedly reported immediately after MS relapse, in adequate correlation with clinical improvement. N-CAM is assumed to have a key role in remyelination process. The exact pathway still remains unclear [
Lower CSF-BNDF levels in SPMS patients comparatively to RRMS patients have been reported. Low BDNF levels are considered to contribute in demyelination and axonal damage progress [
Nogo-A is a CNS myelin component that inhibits axonal repair. Its presence in MS patients CSF constitutes a bad prognostic marker of axonal repair. Nogo-A is adequately specific for MS, as it could not be isolated in other autoimmune or infectious neurological disorders [
IFN-
Progress in understanding of MS pathophysiology shed light over the important role of Th17 immune response. Recently, researchers reached the conclusion that MS patients with prominent Th17 response are probably more harmed than benefited by treatment with IFN-
Vitamin D levels during
Predictive role of BDNF is mentioned in Section
Expression levels of unblocked
CSF levels reduction of sICAM-1 and sE-Selectin may potentially serve as biomarkers of therapeutical efficacy after cladribine treatment [
OCT is a noninvasive technique using emission of infrared light through the pupil and detection of its reflection from the retina. Retinal nerve fiber layer (RNFL) thickness can then be estimated. RNFL thinning can be used as a reliable biomarker of axonal loss, correlating adequately with brain atrophy measures [
MRI provides the clinical doctor with a substantial variety of neuroinflammation biomarkers. On the other hand, classical MRI techniques lack in adequate correlation with neurodegeneration and disability progression.
The most important MRI biomarkers for MS are the following: T1 lesions with contrast enhancement: biomarkers of acute neuroinflammation. Although they are considered as the gold standard for BBB disruption imaging, recent research claims that the same diagnosis can be inferred in many cases by combination of T1, T2, and T2-weighted FLAIR images characteristics alone [ hyperintense T2-weighted lesions: reflecting a combination of mechanisms like inflammation, demyelination, axonal damage and edema. Their diagnostic value is high, but they correlate moderately with disability [ hypointense T1-weighted lesions (black holes): considered as satisfactory biomarkers of axonal damage [ whole brain atrophy biomarkers: the most widely used measure is the brain parenchymal fraction. Brain atrophy worsening rates are higher in untreated MS patients (0.5%–1% annualized decrease) in comparison with healthy controls (0.1%–0.3%) [ gray matter atrophy biomarkers: recently acquired knowledge suggests gray matter demyelination, axonal damage, and atrophy in MS. Double inversion recovery imaging techniques display gray matter atrophy in all MS stages and types, with higher worsening rates in SPMS patients [ spinal cord atrophy biomarkers: upper cervical cord area (UCCA) measuring techniques display atrophy most apparently in progressive MS forms, correlating well with disability progression. UCCA atrophy presence in early disease stages in RRMS patients is a bad prognostic biomarker of future disability [
It is a novel MRI technique based on proton interaction between free water and macromolecules. In the absence of axonal loss, acute MRI lesions that show recovery display also increase in MTR [
DWI is based on mobility and spatial distribution of water molecules, while DTI measures movement in several directions in space. DTI technique provides two different measures, mean diffusivity (MD) and fractional anisotropy (FA).
MD increases and FA decreases in hyperintense T2-weighted lesions. Similar alterations can be recorded in NAWM areas in conventional MRI, as well as in normal appearing gray matter (NAGM) areas, especially in progressive disease forms [
MRS is a novel imaging method for assessment of pathobiochemical disease processes. The following substances spectroscopic measurements are of particular value in MS: NAA: biomarker of neuronal and axonal integrity. NAA showed a progressive decline pattern in a two-year MRS followup of patients with RRMS [ choline: biomarker of myelin loss; myoinositol and creatine: biomarkers of gliosis; glutamate: biomarker of acute inflammation.
Early spectroscopic changes represent a bad prognostic factor of future disability [
Diffusion tensor spectroscopy (DTS), a technique combining properties of DTI and MRS, seems promising in better distinguishing axonopathy, demyelination, inflammation, edema, and gliosis [
Modern PET tracers have the ability to bind in proteins that show upregulation in activated microglia, making possible an early visualization of NAWM and NAGM disorders, even before contrast enhancement in conventional MRI [
EPs estimate action potential conduction along somatosensory, motor, visual, and auditory pathways, providing a reliable means of demyelination and axonal loss assessment. MRI techniques have diminished their spectrum of use, although visual and motor EPs may have some utility as biomarkers of neurodegeneration [
Combined EP data, like the EP score, for instance, could offer a reliable prognostic biomarker, especially for early recognition of benign MS forms [
Recent studies of SSR in MS patients report correlation of SSR abnormalities with higher EDSS scores and disease duration. SSR may be a useful tool of autonomic function assessment among MS patients [
Pathophysiological complexity of MS leads inevitably to a great variety of potential biomarkers, as it was made obvious by the previous analysis. Thus, for systematization reasons and after the completion of an elaborate quest for biomarkers in the international literature we subgrouped all biomarkers in six broad categories, that is, diagnostic biomarkers, biomarkers of phenotypical expression, biomarkers of demyelination—neuroinflammation—relapse, biomarkers of axonal loss—neurodegeneration, prognostic biomarkers—biomarkers of disability progression, and biomarkers of therapeutical response; gradually evaluated every biomarker with (+++) to (+/−) according to their implication in the category they referred to; commented on biomarker abilities to reflect the properties of the category in which they are presented.
The previous three steps demanded as prerequisite hard, critical, and thorough data processing, in order to achieve the best and accurate results, since it is the first time that an attempt of such a systematization is done in a workable table. In this table, biomarkers such as HLA-II alleles and OCBs are for many years the “gold standard” for MS, while other well-described biomarkers are being implicated more and more every day, like conventional and nonconventional MRI scans. Additionally, many other laboratorial and imaging parameters are at the beginning of their characterization as biomarkers in MS.
From every group of biomarkers, we collected those with characterization of (+++) and presented them as an easy summary to the reader, as it follows. Additionally, we present the biomarkers of differentiation between MS and NMO:
Even after decades of research, MS still remains at a significant proportion an unsolved mystery. This is mainly the reason why finding a biomarker with absolute surrogacy abilities remains elusive. Further research in the field of MS biomarkers must be directed towards an earlier and accurate diagnosis and a more prompt, targeted, and individualized therapeutical approach, with the minimum intervention and economical cost.
The authors declare no conflict of interests.