Fluid Biomarkers in Multiple Sclerosis

Largely due to the absence of highly specific biomarkers for multiple sclerosis (MS), accurate diagnosis remains a challenge. In a recent systematic review and meta-analysis, the frequency of misdiagnosis ranged from 5% to 41%, with delay in rectifying misdiagnosis extending up to 16 years.¹ Misdiagnosis from misapplication of the McDonald criteria² is often the result of misinterpretation of clinical symptoms and overreliance on imaging abnormalities. The use of highly specific fluid biomarkers in combination with clinical history and radiologic findings may allow for more accurate MS diagnosis, early prognostication, and ongoing disease monitoring.

The clinical utility of a biomarker lies in its ability to provide measurable information about a disease or biologic process. Biomarkers can be physiologic, radiographic, histologic, or molecular, and can be found in body fluids, such as blood, urine, and cerebrospinal fluid (CSF). Because of the emphasis on lesion location and radiologic change over time in the McDonald criteria,² MRI has become a key complement to clinical history and examination in the diagnosis, prognostication, and disease monitoring of MS. However, MRI studies present limitations: they are costly, time-consuming, and stress-inducing, especially for individuals with claustrophobia. 

Fluid biomarkers may curtail reliance on imaging in MS. They are especially valuable in assessing an individual’s clinical state as they are collected efficiently through mostly minimally invasive methods. The development of highly sensitive immunoassay technology has allowed for more accurate measurement of blood molecules that may aid in the diagnosis and treatment of MS.³ Examples of assays include the single molecule array (Simoa) for neurofilament light chain (NfL) and glial fibrillary protein (GFAP) detection and the Octave Multiple Sclerosis Disease Activity (MSDA) panel (Octabe Bioscience, Menlo Park, CA), which measures several biomarkers putatively involved in MS pathophysiology. Care must be taken in the interpretation of these tests, however, as age, weight, the presence of other neurodegenerative conditions, and cross-reactivity may affect results.

A recent example of a fluid biomarker that transformed clinical diagnosis is the aquaporin-4 antibody diagnostic of neuromyelitis optica spectrum disorder (NMOSD). NMOSD was previously often misdiagnosed as MS,⁴ and this revolutionary discovery allowed for individuals with NMOSD to receive the appropriate therapy and avoid disease-modifying therapy (DMT) used to treat MS, some of which can lead to relapse in NMOSD. MS, by contrast, is a pathologically heterogenous disease⁵ with different patterns of progression and varying responses to treatment. It is unlikely that a single biomarker will capture all of MS, and a suite of tests may prove to have higher accuracy. 

The study of emerging fluid biomarkers, including those outlined in this article, will allow for improvement in diagnosis, prognostication, and monitoring of disease activity and treatment response in MS (Table).

Fluid Biomarkers for Diagnosis

CSF Oligoclonal Bands and Immunoglobulin G Index

Increased intrathecal immunoglobulin G (IgG) production is seen in >90% of individuals with MS.⁶ IgG production can be measured either quantitatively by the IgG index or qualitatively by the presence of IgG oligoclonal bands (OCBs), a long-estabAlished diagnostic biomarker. According to the 2017 revision of the McDonald criteria,⁷ the presence of ≥2 CSF-specific OCBs is evidence for dissemination in time, thus allowing for the diagnosis of MS to be made in the absence of a second attack, baseline MRI findings indicating dissemination in time, or serial MRI findings indicating new or active lesions. An elevated IgG index, a ratio calculated using CSF and serum IgG and albumin, indicates increased IgG production in the central nervous system. An IgG index >0.7 has been shown to have a positive predictive value of >99% for the presence of OCBs.⁸ OCBs, however, remain the gold standard for determining intrathecal IgG synthesis and for MS diagnosis. In addition, the number of OCBs and IgG index have been shown to correlate with the number of cervical and T2 lesions, and IgG index correlates with Expanded Disability Status Scale (EDSS) score, demonstrating that these markers may be able to predict MS disease activity and progression.⁹ 

Kappa Free Light Chains

Other fluid biomarkers of intrathecal IgG synthesis, kappa free light chains (κFLCs) and lambda free light chains (λFLCs), are produced by B cells during immunoglobulin production. Measuring λFLCs has not been found to be a valid test for diagnosing clinically isolated syndrome (CIS) or MS, but κFLCs have shown promise. The κFLC index, a ratio calculated using CSF and serum κFLC and albumin, and CSF κFLC/IgG ratio show high sensitivity and moderate specificity in diagnosing MS.¹⁰ The κFLC index was found to have a similar accuracy to OCBs in diagnosing MS, and it has the added advantage of being less costly, more objective, and more quickly processed. The proposed 2024 McDonald criteria include CSF κFLCs as an alternative to OCBs in MS diagnosis.¹¹

Fluid Biomarkers for Prognosis

Kappa Free Light Chains

In terms of prognostication, people with CIS or MS with a high κFLC index (>100) have a 4-fold risk of developing progression independent of relapse.¹² Another study demonstrated that individuals with CIS who converted to MS have shown higher CSF κFLC/IgG ratios, suggesting that measurement of κFLC may stratify the risk of conversion.¹⁰ 

Neurofilament Light Chains

NfLs are cytoskeleton proteins that are released during axonal injury. Measurement of CSF and serum NfL, especially early in the disease course, can reflect the extent of neuronal damage. Recent technologic advances now allow for sensitive measurement of the trace amounts of NfL that transfer from CSF to blood,¹³ allowing for more efficient detection. Although baseline serum NfL level has limited diagnostic value due to low specificity,¹⁴ it may reliably predict disease progression and disability, including future relapses and progression independent of relapse activity.¹⁵

Glial Fibrillary Acidic Protein

GFAP, a cytoskeletal filament of astrocytes, is a potential biomarker for disease activity and progression in people with MS. CSF GFAP has been found to be associated with future clinical progression in a study including participants with relapsing or primary progressive MS.¹⁶ Other studies have shown the utility of serum GFAP; in a study of people with progressive MS, elevated baseline serum GFAP levels were associated with higher risk of current disability as well as future progression.¹⁷

Chitinase‐3‐like Protein 1

Chitinase-3-like protein 1 (CHI3L1), also known as YKL40, is an extracellular glycoprotein produced mainly by macrophages and astrocytes. A meta-analysis found CSF levels of CHI3L1 to have a strong correlation with disease progression: higher levels were seen in converting CIS than in nonconverting CIS.18 This suggests that CHI3L1 may predict later conversion to MS. Elevated CSF CHI3L1 levels had a strong significant correlation with EDSS score. This meta-analysis also found that CSF CHI3L1 levels were higher in the remission stage than during relapse activity, indicating that CHI3L1 may have a role beyond the acute phase. Further study is needed to clarify the role of CHI3L1 in serum.

C-X-C Motif Chemokine Ligand 13

C-X-C motif chemokine ligand 13 (CXCL13) is a small chemokine that is involved in attracting B cells and a subset of T cells to lymphoid follicles. Elevated CSF CXCL13 levels in people with MS have been shown in multiple studies to predict disease course.¹⁹ The CXCL13 index, a measure of intrathecal production of CXCL13, has also demonstrated utility in predicting the success of certain DMTs. A recent study examining moderate-efficacy DMTs (ME-DMTs) versus higher-risk high-efficacy DMTs found that a low CXCL13 index predicted success with ME-DMTs or no DMT whereas a high CXCL13 index predicted failure with ME-DMTs in achieving no evidence of disease activity in individuals with CIS.²⁰

Fluid Biomarkers for Monitoring Disease Activity and Treatment Response

Neurofilament Light Chains

Treatment with DMTs has been shown to correlate with reduction in serum NfL levels. A study of plasma NfL levels following initiation of DMTs in people with MS found that the choice of DMT correlated with degree of reduction in plasma NfL levels, suggesting that NfLs may serve as a biomarker for DMT efficacy.²¹ NfL is more highly concentrated in the CSF than serum, but the invasiveness and logistic limitations of lumbar puncture make monitoring with CSF testing more challenging.²² With regard to clinical decision-making, baseline levels of serum NfL may be drawn prior to DMT initiation, then redrawn months later to inform treatment decisions, as was done in a study of people with MS treated with peginterferon β-1a.²³ This study found that decreased levels of serum NfL at 6 and 12 months posttreatment correlated with reduced brain volume loss at 4 years, and decreased serum NfL levels at 9 and 12 months correlated with reduced disability progression as measured by a 4-year change in EDSS score. Thus, elevated serum NfL levels may prompt escalation to a higher-efficacy DMT, whereas low or normal serum NfL levels may allow for continuation of the current DMT. It is important to remember when interpreting these results that NfL levels increase with age. Other potential confounders include trauma and neurodegenerative conditions.

C-X-C Motif Chemokine Ligand 13

In addition to being a prognostic biomarker for disease course and DMT success prior to initiation, CXCL13 may also correlate with disease activity following DMT initiation.¹⁹ Serum CXCL13 levels were found to drop in individuals showing optimal treatment response, defined as no evidence of disease activity after 1 year of treatment with teriflunomide.²⁴ In another study, which found that elevated CSF CXCL13 levels and CXCL13 indices at baseline predicted optimal rituximab response, these markers were also found to decrease 6 months posttreatment with rituximab.²⁵ Further study is needed examining CXCL13 in responders vs nonresponders after treatment with rituximab and other DMTs.

MicroRNA

MicroRNAs (miRNAs) are short noncoding single-stranded molecules involved in regulating gene expression and in the pathogenesis of MS. They regulate T-cell differentiation and are associated with MS disease activity. Several studies show that CSF, serum, or plasma miRNA levels can distinguish active from inactive disease after use of different DMTs, suggesting they may be used as a biomarker to assess DMT efficacy.²⁶ For example, in one study of participants with relapsing-remitting MS treated with dimethyl fumarate, plasma miR-125a-5p and miR-146a-5p levels were lower after 4 months of treatment in individuals showing EDSS progression at 12 months, whereas levels in individuals without progression were comparable to those in healthy controls.²⁷ In another study, serum miR-548a-3p levels after 6 months of treatment with fingolimod were significantly higher in individuals who had no evidence of disease activity at 2 years.²⁸ Interpretation of miRNA levels is dependent on the role of the specific miRNA; some miRNAs appear to play a protective role by reducing inflammation, whereas others are associated with disease activity and progression. For example, the serum levels of certain exome-associated miRNAs were found to be downregulated, but levels of other miRNAs were upregulated, in individuals with relapsing-remitting MS who were responsive to IFN-β therapy.²⁹ MicroRNAs have been studied as diagnostic biomarkers; however, because their levels may change with treatment, their value in clarifying cases of misdiagnosis may be limited.

Conclusion

Fluid biomarkers complement clinical history and radiographic findings, may allow for more accurate and efficient MS diagnosis, and may support earlier intervention by predicting prognosis and enabling ongoing monitoring of disease activity and personalized DMT selection. Though many biomarkers for prognosis and monitoring are still being studied for reliability, NfL has the most evidence for use in the clinical context and is currently commercially available. Research on GFAP, CHI3L1, and CXCL13 is ongoing and that on the many microRNAs is still in its infancy.