An Update on Laboratory-Based Diagnostic Biomarkers for Multiple Sclerosis and Beyond 

Abstract

Background

Multiple sclerosis (MS) is an immune-mediated central nervous system (CNS) inflammatory demyelinating disease in which analysis of clinical presentation, imaging studies, and laboratory tests aid in diagnosis.

Content

This review discusses laboratory tests ordered to rule out and rule in MS, such as the traditional measurement of cerebrospinal fluid (CSF) IgG index and oligoclonal bands. Biomarkers discovered in the past 2 decades, such as aquaporin-4 (AQP4) antibodies and myelin oligodendrocyte glycoprotein (MOG) antibodies, have been incorporated into clinical practice in the diagnosis of disorders referred to as MS mimics. The importance of test selection, assay methodology, optimal sample for testing, and diagnostic utility of these biomarkers is reviewed. Other laboratory testing that can aid in the differentiation between MS and these biomarker-defined CNS demyelinating diseases is described. There is a focus on emerging biomarkers such as the use of kappa immunoglobulin free light chain concentration in CSF and kappa CSF index measurement as an alternative to oligoclonal bands which has a potential for an improvement in laboratory workflows. Finally, the role of biomarkers of disease activity and prognosis are discussed, including neurofilament light chain, glial fibrillary acidic protein, and myelin basic protein. Future perspectives with improved laboratory testing tools and discovery of additional biomarkers are provided.

Summary

Laboratory testing for demyelinating disorders using CSF and serum are routine practices that can benefit from an update, as novel biomarker-defined entities have reduced the potential for MS misdiagnosis, and CSF/serum biomarkers reinstated in the diagnostic criteria of MS.

multiple sclerosis, oligoclonal bands, IgG index, CSF total protein, Kappa free light chains, neuromyelitis optica spectrum disorders

Introduction

Acquired inflammatory demyelinating diseases of the central nervous system (CNS) are an umbrella term used to categorize a group of diseases in which there is immune-mediated damage that leads to loss of myelin. Common manifestations of these disorders include attacks of inflammation of the optic nerve (optic neuritis), spinal cord (transverse myelitis), or brain (encephalitis). These regions can be involved concurrently (e.g., acute disseminated encephalomyelitis involving the brain and spinal cord) or sequentially (e.g. neuromyelitis optica, a syndrome with inflammation of the spinal cord and optic nerve that usually occurs separately). Multiple sclerosis (MS) is the commonest inflammatory demyelating disease in the CNS. The first disease presentation was described in 1868 by a French neurologist, Dr. Jean-Martin Charcot (1). The presentation of MS is heterogeneous, and the disease cause remains unknown. A single target antigen or causative autoantibody responsible for the disease has not been identified to date despite substantial efforts.

MS can be divided into different categories based on the clinical presentation, ranging from asymptomatic to progressive disease. There are at least 5 categories considered among the spectrum of MS: radiologically isolated syndrome (MRI findings without clinical features), clinically isolated syndrome (CIS, representing the first attack of MS), relapsing-remitting MS, primary progressive MS, and secondary progressive MS. MS has an onset between 15 and 50 years of age. Suggestive features of the disease include periods of attacks and periods of clinical remission or a progressive course from onset with a gradual decline in walking distance over time. Prevalence of MS in women has increased since 1950 from 1.4:1 to 2.3:1 in 2000 (2). The incidence and prevalence of MS vary geographically. High prevalence of MS is defined as 60 per 100 000 or more individuals. Europe, Southern Canada, Northern USA, New Zealand, and southeast Australia meet this criterion (3). It is estimated that 2.8 million people live with an MS diagnosis around the world (4).

The differential diagnosis for MS requires concomitant review of clinical presentation, imaging studies, and laboratory tests. The McDonald criteria are the most widely used tools to aid in the diagnosis of MS (5), and state that diagnosis of MS requires the identification of lesions in at least 2 separate distinctive regions (dissemination in space) in the CNS at 2 different points in time (dissemination in time). In their 2017 revision, the McDonald criteria reintroduced the presence of oligoclonal bands as an alternative to dissemination in time by a new clinical attack or MRI. This has increased the value of laboratory testing for MS. In addition, biomarker discovery in the past 2 decades helped identify diseases whose clinical presentation is an inflammatory demyelinating disorder of the CNS but atypical from MS, the “MS mimics.” This has reduced the risk of misdiagnosis. Conditions previously categorized as MS now may fall under less common, albeit inflammatory, CNS demyelinating diseases defined by autoantibody biomarkers: aquaporin-4 antibody positive neuromyelitis optica spectrum disorders (AQP4 + NMOSD) and myelin oligodendrocyte glycoprotein-IgG antibody-associated disease (MOGAD).

Most of this review will focus on laboratory tests as diagnostic tools in MS, how to confirm a diagnosis of AQP4 + NMOSD and MOGAD, and how they can be discriminated from MS. This distinction is important as there are crucial differences in treatments and prognosis, despite some overlap.

Differential Diagnosis for Demyelinating Diseases and CNS Inflammation

A patient with a clinical presentation suggestive of CNS demyelinating disease typically develops symptoms as episodes (attacks) that develop over days to weeks, plateau, and then symptoms tend to improve over subsequent months, particularly when given acute treatment. History, and physical and neurological examination are key (Table 1). After the history and neurologic examination have been completed, imaging of the CNS plays a major role in determining the localization to which the symptoms and signs are referable. This involves MRI of the brain, spine, or orbits with and without contrast to detect inflammation within the brain, spinal cord, or optic nerve. The location and pattern of MRI abnormalities, including their evolution over time, are useful to help determine the specific CNS demyelinating disease or alternative process (other inflammation, neoplasm, vascular, toxic/metabolic, infections, etc.). While a full review of the MRI features of inflammatory demyelinating diseases of the CNS is beyond the scope of this review, MRI examples of MS, AQP4 + NMOSD, and MOGAD are depicted in Fig. 1. A lumbar puncture is also often performed to evaluate the cerebrospinal fluid (CSF) composition and characteristics that can help diagnostically in those with inflammatory CNS demyelinating diseases.

Fig. 1.

Comparison of brain and spine MRI in multiple sclerosis (MS), aquaporin-4–IgG seropositive neuromyelitis optica spectrum disorder (AQP4 + NMOSD), and myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD). A = Multiple sclerosis; B = AQP4 + NMOSD; C = MOGAD. MRI head axial fluid-attenuated inversion recovery (FLAIR) images in a patient with MS reveals a periventricular hyperintense lesion in the left hemisphere and some additional hyperintense lesions in the right hemisphere (A1, arrows). MRI sagittal T2-weighted image of the cervical spine in an MS patient reveals a short T2-hyperintense lesions (A2, arrow) that on axial images involves the periphery of the cord in the dorsal column (A3, arrow). MRI head axial FLAIR images in a patient with AQP4 + NMOSD reveal a characteristic hyperintense lesion that is located adjacent to the third ventricle (B1, arrow). MRI sagittal T2-weighted image of the thoracic spine in a patient with acute myelitis from AQP4 + NMOSD reveals a longitudinally extensive T2-hyperintense lesion extending >3 vertebral segments (B2, arrows), with a brighter spotty T2-hyperintensity (B2, arrowhead) typical of this disease. MRI head axial FLAIR images in a patient with MOGAD during an attack of acute disseminated encephalomyelitis (ADEM) reveals the characteristic multifocal T2-hyperintense lesions in the white matter (C1, arrows), along with a characteristic T2-lesion within the deep gray matter involving the right caudate (C1, arrowhead). MRI sagittal T2-weighted image of the thoracic spine in a patient with acute myelitis from MOGAD reveals a longitudinally extensive T2-hyperintense lesion extending >3 vertebral segments (C2, arrows) that on axial images is central and highly restricted to the gray matter forming a characteristic H-sign (C3, arrow). Abbreviations: AQP4 + NMOSD, aquaporin-4–IgG seropositive neuromyelitis optica spectrum disorder; MOGAD, myelin oligodendrocyte glycoprotein antibody-associated disease.

Table 1.

Differentiation between multiple sclerosis and MS mimic conditions.

	Multiple sclerosis	Aquaporin 4 IgG seropositive neuromyelitis optica spectrum disorder (NMOSD)	Myelin Oligodendrocyte Glycoprotein antibody-associated disease (MOGAD)
Clinical presentation	In 85%, MS presents in the form of an attack with acute/subacute symptoms that later improve (termed relapsing-remitting MS) and the most common types include: myelitis (numbness with or without mild weakness, imbalance, or bowel/bladder dysfunction), ON (monocular visual loss with pain with eye movements), brainstem or cerebellar syndrome (diplopia that is painless and resolves with unilateral eye closure or imbalance from ataxia). In a timeframe usually between 10 and 30 years from MS onset, most will develop a slow decline in the ability to walk, usually from spinal cord dysfunction (myelopathy) termed secondary progressive MS. In 15%, this slow gradual worsening occurs from onset without preceding attacks and is termed primary progressive MS. The majority of disability in MS develops occurs during this progressive phase.	Clinical attack types are most commonly transverse myelitis, ON, and area postrema syndrome, the latter manifesting with intractable nausea, vomiting, and hiccups from involvement of the vomiting center of the brain, in which AQP4 is particularly enriched. Attacks tend to be more severe than MS with less recovery, and thus a single attack can lead to permanent blindness or paraplegia. Disability accumulates with each attack, but it differs from MS in that it lacks a progressive course.	Manifestations including optic neuritis (ON), acute disseminated encephalomyelitis, myelitis, or cerebral cortical encephalitis. The attacks are more severe than MS and similar to AQP4 + NMOSD but recover better. Approximately half of cases are monophasic and do not develop a recurrence, while half will have relapsing disease. MOGAD lacks the progressive course that is common in MS.
Lab work	CSF and serum samples drawn at differential diagnosis timeframe. Work-up frequently includes - Total protein in CSF - Glucose in CSF and serum - Oligoclonal banding - CSF IgG index - IgG synthesis rate - Albumin quotient	CSF and serum samples are only tested for AQP4 if the MS presentation is atypical. Oligoclonal banding is positive in 20–30% of cases. Aquaporin-4 antibody is the biomarker and best tested in blood using a cell-based assay.	CSF and serum samples are only tested for MOG antibodies if the MS presentation is atypical. Oligoclonal banding is positive in 20–30% of cases. MOG antibody is the biomarker that is best tested in blood using a cell-based assay.
McDonald criteria	Oligoclonal banding positive result of 2 or more unique CSF bands may replace dissemination in time or in space if only one clinical attack is observed.	Does not apply.	Does not apply.
Treatment options	Early treatment/attack prevention decreases CNS disease burden. Acute treatment includes high dose steroids. Plasma exchange can be used as an alternative in patients unresponsive to steroids. Long-term treatment with disease modifying therapies (DMT1) decreases relapse rates and new CNS lesions on MRI. Symptomatic treatment to manage complications of MS is also important.	Prevention of attacks is crucial, in addition to aggressive treatment of attacks with high dose steroids and plasmapheresis (6). Studies in vitro have shown the importance of B cells (CD20+, CD19+), plasmablasts (CD19+, CD20−), interleukin-6 (IL-6) and complement in the pathogenesis of AQP4 + NMOSD (7). Four prospective placebo-controlled randomized controlled trials evaluating attack-prevention treatments in AQP4 + NMOSD demonstrated efficacy of 4 monoclonal antibodies targeting CD20 (rituximab), CD19 (inebilizumab), IL-6 (satralizumab) and complement (eculizumab) in preventing attacks in this disease.	There are no proven treatments for MOGAD, although clinical trials are likely to get underway in the near future. Acute attacks tend to be treated with high dose steroids, and in patients with relapsing disease, empiric immunosuppressants are often used.

CSF, cerebrospinal fluid; NMOSD, neuromyelitis optica spectrum disorder. ¹DMT, disease modifying therapies include: Glatiramer acetate (Copaxone), beta-interferons (Avonex, Rebif, Plegridy, Extavia), fingolimod (Gilenya)/siponimod (Mayzent), dimethyl fumarate (Tecfidera), Teriflunomide (Aubagio), natalizumab (second line of therapy or severe disease), alemtuzumab, ocrelizumab (possible role in primary progressive multiple sclerosis), and autologous hematopoietic stem cell transplant.

The Role of the Laboratory in Diagnosis of MS

The evaluation of patients with immune-mediated CNS disorders includes analysis of CSF and serum. A macroevaluation of the composition of serum and CSF is traditionally performed to assess the integrity of the blood–brain barrier (BBB), which is a regulated network that separates the CNS from the circulation. The first line of testing for ruling in or ruling out MS includes oligoclonal banding and/or kappa free light chain testing, CSF IgG index, CSF total protein, CSF glucose, cell count, and cell differential.

Ruling Out MS: Total Protein, Glucose, Cell Counts, and Cytology

Total protein concentration in CSF may indicate the integrity of the BBB. CSF is an ultrafiltrate of plasma, and comparatively, the total protein concentration in CSF is 100-fold lower. Newborns and healthy elderly adults display higher concentrations than young adults (8, 9) and CSF total protein reference intervals could benefit from age adjustments (10, 11). Disruption of the BBB and infiltration of immune cells are key steps in inflammatory CNS conditions. Markedly increased total protein (>100 mg/dL) or increased white blood cell count beyond 50 cells/µL should be considered a red flag and is more suggestive of an infectious or other inflammatory disease rather than MS. In fact, the CSF total white blood cell count is normal in about 66% of patients (12). A decreased CSF glucose in relation to serum glucose (ratio <0.5) is atypical for MS and can help further suggest an infectious or neoplastic etiology (8, 13). BBB integrity may be assessed by calculation of the albumin quotient (Eq. 1), and despite the fact that the test is not sensitive or specific for the diagnosis of MS (14), the finding of high albumin quotient is associated with systemic inflammation (15).

$$Albumin-quotient={\displaystyle \frac{CSFAlbumin\left({\displaystyle \frac{mg}{dL}}\right)}{SerumAlbumin(g/L)}}$$

(1)

Ruling in MS: Oligoclonal Bands and CSF IgG Index

It is the oligoclonal bands and CSF IgG index assays that are held as the most sensitive and specific assays for MS diagnosis. For both assays, the interpretation requires a paired collection of serum and CSF, which are often run side-by-side in the laboratory. Fig. 2 shows an example of a patient who had testing done by several different methods.

Fig. 2.

Different clinical laboratory tests used to rule in multiple sclerosis, performed on a single sample set of paired CSF and serum. Results from CSF IgG index, Kappa Free Light Chain concentration, and Kappa index are automated, and results could be available on the same day of the sample collection, providing initial information to clinical teams and patients. Specialized testing such as oligoclonal banding run by protein electrophoresis or isoelectric focusing offer a minimal turnaround time of a few days. The oligoclonal banding test recommended method is isoelectric focusing electrophoresis. Other tests such as CSF total protein, CSF and serum glucose, CSF cell count, and cell differential would likely have been performed in additional to the assays represented here, for a complete CSF work-up in the investigation of a potential CNS inflammatory disorder. Figure created with BioRender.com. Abbreviations: CNS, central nervous system; CSF, cerebrospinal fluid.

Oligoclonal bands.

Testing for oligoclonal bands (OCB) can be performed using 2 methods: protein electrophoresis (PEP) or isoelectric focusing (IEF). The 2017 McDonald criteria recommend IEF for OCB detection given its superior sensitivity and specificity (5). The situations where OCB testing can be helpful are equivocal MS cases requiring a tie-breaker exam; however, this test, like every other, has limitations. Despite the recommendation, according to 2020 CAP proficiency testing surveys, 1/3 of participant clinical laboratories use PEP and 2/3 employ IEF for analysis. IEF is followed by blotting and staining specifically for IgG. It allows sharper separation of IgGs according to their isoelectric points as contrasted with PEP.

Reagents on the market for OCB IEF are Food and Drug Administration-approved but there is no guidance in the package insert on reference interval for positives. Overall, the threshold used in clinical practice varies between a cutoff of 2, 3, or 4 unique bands in CSF as positive, and it is recommended that clinical laboratories perform their own studies. The McDonald criteria called 2 bands positive, which raised potential for confusion if a clinician was basing their test interpretation on use of the McDonald criteria but the laboratory report stated this cutoff was a negative result for OCB. In a 2019 College of American Pathologists educational dry challenge, 25% of participants interpreted the finding of 2 unique CSF bands as negative.

Textbooks often state OCBs are positive in 95% of patients with MS and up to 8% of healthy subjects (16). However, the definition of clinical specificity and sensitivity for OCB can vary substantially, according to the control population used as true negatives, falling to 60% when inflammatory disorders are in the control groups (14). We reviewed OCB positivity in 1339 paired CSF/serum samples: 2 or more unique CSF OCBs were found in 77% of patients with MS (n = 118/155), 18% of patients with other CNS immune-mediated inflammatory conditions (n = 66/366), and 8.7% of other conditions such as headache, migraines (n = 71/819) (17) (Table 2). This cohort may not reflect positivity generally seen elsewhere for 2 main reasons. The first is that we are a tertiary referral center and could be seeing more atypical MS case presentations. Second, classic cases of MS may not need CSF testing when imaging and clinical presentation are clear, and it is estimated that more atypical cases are the ones in which testing is performed. OCB are still useful as a discriminator of MS from MOGAD and AQP4 + NMOSD, since OCB are encountered much less frequently in these conditions (Table 1).

Table 2.

Comparison of analytical performance of tests commonly used for diagnosis of multiple sclerosis.

Variable	MS, n = 155	CNS inflammation, n = 365		Non-CNS inflammation, n = 818
OCB ≥ 2 bands	118	64		71
Kappa FLC ≥0.06 mg/dL	123	113		121
Kappa FLC ≥ 0.1 mg/dL	112	81		76
Receiver Operating Characteristic Curves, MS = true positives, CNS inflammation, and non-CNS inflammation combined as control groups.
Variable (positive test)	AUC (95% CI)	Sensitivity (95% CI)	Specificity (95% CI)	PPV	NPV	Diagnostic OR
OCB (≥4)	0.801 (0.763, 0.839)	66.9 (59.1, 73.8)	93.3 (91.7, 94.6)	56.6	95.5	28.15
OCB (≥2)	0.826 (0.791, 0.860)	76.6 (69.3, 82.6)	88.5 (86.5, 90.2)	46.6	96.6	25.19
Kappa FLC (≥0.06 mg/dL)	0.798 (0.764, 0.832)	79.4 (72.4, 85.0)	80.2 (77.8, 82.4)	34.5	96.7	15.61
Kappa FLC (≥0.10 mg/dL)	0.798 (0.762, 0.834)	72.9 (65.4, 79.3)	86.7 (84.6, 88.5)	41.9	96.1	17.54
CSF IgG Index ≥0.85	0.668 (0.629, 0.708)	38.2 (30.9, 46.1)	95.6 (94.2, 96.7)	53.7	92.0	13.43
CSF IgG index ≥0.7	0.775 (0.729, 0.816)	49.7 (41.9, 57.5)	92.1 (90.4, 93.6)	45.4	92.9	11.03
Albumin quotient >9	0.609 (0.563, 0.654)	26.7 (24.0, 29.3)	86.5 (80.2, 91.0)	6.6	4.3	0.42
IgG synthesis rate (0–12 mg/24 h)	0.727 (0.685, 0.769)	21.9 (15.9, 29.4)	92.11 (90.3, 93.6)	28.1	89.3	3.28
Kappa concentration ≥0.06 mg/dL reflexes to OCB	0.823 (0.787, 0.860)	72.7 (65.2, 79.1)	92.0 (90.3, 93.4)	54.1	96.3	30.62

MS, multiple sclerosis; CNS, central nervous system; OCB, oligoclonal banding; PPV, positive predictive value; NPV, negative predictive value; OR, odds ratio; CSF, cerebrospinal fluid.

CSF IgG index and IgG synthesis rate.

Under normal conditions, most of the IgG present in CSF comes from peripheral sources by crossing the BBB; because IgG is not normally produced intrathecally, its presence in CSF does not necessarily indicate inflammation (18). Since albumin is synthesized in the liver, the presence of albumin in CSF suggests disruption of the BBB, and being a smaller protein, it crosses the BBB more easily than IgG. The IgG Index analysis is a quantitative measurement of intrathecal IgG synthesis in which a positive result for the CSF IgG index (Eq. 2) indicates an inflammatory CNS condition (8). Several laboratories also report the IgG synthesis rate (Eq. 3). Given the accessibility and low cost of the IgG reagents, IgG index is still considered a potential candidate for MS diagnosis. IgG index, however, displays lower sensitivity (50–75%) compared to OCB, although it is reported to have higher specificity (85%) (19). In an internal study at Mayo Clinic, we reported 92.1% specificity and 49.7% sensitivity by using a cutoff of >0.7. We also evaluated the synthesis rate formula and reported a specificity of 92.4% and sensitivity of 21.9% using a reference interval of 0–12 mg/24 h (Table 2). Presence of IgM in CSF has been associated with more aggressive progression to MS in patients with CIS (20, 21), and correlated with MRI lesion loads and severity of disease course. Presence of IgM is much less frequent than OCB or intrathecal IgG. OCB for IgM (22), formulas for IgM intrathecal synthesis (20) and CSF IgM index (Eq. 4) have been reported but are not performed frequently nor widely available in routine clinical laboratories. Such testing is more common in specialized CSF laboratories, mainly in Europe.

$$\begin{array}{r}\frac{IgG-index={\displaystyle \left(CSFIgG\left({\displaystyle \frac{mg}{dL}}\right)/CSFAlbumin\left({\displaystyle \frac{mg}{dL}}\right)\right)}}{\left(SerumIgG\left({\displaystyle \frac{g}{dL}}\right)/SerumAlbumin\left({\displaystyle \frac{g}{dL}}\right)\right)}\end{array}$$

(2)

$$\begin{array}{rl}& IgGSynthesisRate\\ & =\{\left(\mathrm{CSF}\mathrm{IgG}-{\displaystyle \frac{\mathrm{Serum}\mathrm{IgG}}{341}}\right)\\ & -\left[\left(\mathrm{CSF}\mathrm{Alb}-{\displaystyle \frac{\mathrm{Serum}\mathrm{Alb}}{169}}\right)\left({\displaystyle \frac{SerumIgG}{SerumAlb}}\right)(0.43)\right]\}x5\end{array}$$

(3)

$$\begin{array}{r}\frac{IgM-index={\displaystyle \left(CSFIgM\left({\displaystyle \frac{mg}{dL}}\right)/CSFAlbumin\left({\displaystyle \frac{mg}{dL}}\right)\right)}}{\left(SerumIgM\left({\displaystyle \frac{g}{dL}}\right)/SerumAlbumin\left({\displaystyle \frac{g}{dL}}\right)\right)}\end{array}$$

(4)

CSF immunoglobulin kappa free light chain indexes.

An alternative marker to the CSF IgG index calculation with higher sensitivity and comparable performance to OCB is the measurement of CSF immunoglobulin kappa free light chains (KFLC), either in place of the IgG in serum and CSF in different index calculation formats for linear synthesis (Eq. 5) or nonlinear intrathecal synthesis (Eq. 6) (23), or simply the measurement of CSF KFLC concentrations on their own, without a paired serum sample (17, 24). Among the calculations format, the kappa index (Eq. 5) is the most popular, and its equivalence to OCB as an alternative test has been conveyed in multiple publications (25). The main caveat for the acceptance of the new marker is that it is not included in the most up-to-date version of the McDonald criteria (5).

$$\begin{array}{r}\frac{K-index={\displaystyle \left(CSFKFLC\left({\displaystyle \frac{mg}{dL}}\right)/SerumKFLC\left({\displaystyle \frac{mg}{dL}}\right)\right)}}{\left(CSFAlbumin\left({\displaystyle \frac{mg}{dL}}\right)/SerumAlbumin\left({\displaystyle \frac{mg}{dL}}\right)\right)}\end{array}$$

(5)

$$Q_{KFLC}={\displaystyle \frac{CSFKFLC}{SerumKFLC}}$$

$$Q_{Alb}={\displaystyle \frac{CSFAlbumin}{SerumAlbumin}}$$

$$KFL{C}_{Lim}=0.9358xQAl{b}^{0.6687}$$

$$KFL{C}_{Loc}=(Q_{KFLC}-KFL{C}_{Lim})\mathrm{x}\mathrm{Serum}KFLC$$

$$KFL{C}_{IF}=(KFL{C}_{LOC}/CSFKFLC)x100$$

(6)

The kappa index is considered a more sensitive but slightly less specific test than OCB. This provides useful information to exclude MS in kappa index negative patients. Consensus on the most appropriate cutoff value for the kappa index is building: a value ≥5.9 had a sensitivity of 96% and specificity of 86% for the diagnosis of MS as opposed to sensitivities of 91% and specificity of 92% when compared to OCB (26), and a multicenter study determined the kappa index ≥6.6 to have a sensitivity and specificity of 88% and 83% as opposed to 82% and 92%, respectively, for OCB (25). The availability of multiple assays for free light chains will require each laboratory to evaluate performance of the cutoffs proposed to be associated with demyelinating disease.

Measurement of the absolute CSF KFLC concentration without a paired serum would simplify the workflow and calculations across laboratory information systems. A CSF KFLC concentration >0.06 mg/dL was shown to have similar value to the kappa index in 325 participants (24), and later a KFLC ≥0.06 mg/dL was shown to have sensitivity and specificity of 77% and 80% as opposed to 74% and 89% for OCB in a larger cohort, without the need of a paired serum. A reflex strategy was then adopted by our institution for the diagnosis of MS: testing starts with KFLC in CSF, and when the value is <0.06 mg/dL, it rules out MS. Any KFLC ≥0.06 mg/dL reflexes to OCB testing for confirmation (17). The reflex method reduces the laboratory workload on OCB, with shorter TAT and lower overall cost, and still fits in with the McDonald criteria for MS diagnosis. CSF KFLC has not been fully investigated in MOGAD and AQP4 + NMOSD (24).

Autoantibody Biomarker-Defined CNS Inflammatory Demyelinating Diseases

AQP4 + NMOSD and MOGAD represent recently defined CNS demyelinating diseases distinct from MS that have blood biomarkers available for diagnosis (Table 1). AQP4 is located on the end feet of astrocytes and is the predominant water channel in the brain. The AQP4 antibody is directly pathogenic and targets the astrocyte leading to astrocytic damage, AQP4 loss, and secondary immune-mediated demyelination (27–29).

MOG is a glycoprotein located on the myelin surface found exclusively in the CNS and represents 1% of the CNS myelin proteins. While the exact function is not fully understood, it is thought to be an adhesive protein important to the complement cascade and oligodendrocyte microtubule stability (30, 31). The human IgG antibodies against MOG are associated with a subset of inflammatory demyelinating disorders of the CNS that are distinct from MS and AQP4 + NMOSD.

AQP4 and MOG antibody assays in clinical practice: test selection, assay methodology, and interpretation.

The antibody biomarker pattern of AQP4-IgG was initially discovered by running serum on mouse brain tissue composite using immunohistochemistry, and subsequently the target was identified as the AQP4 water channel (28, 29). ELISAs for detection of AQP4-IgG later superseded the initial tissue immunohistochemistry but have a risk of false positivity, particularly at low titer (32, 33). The pioneering use of cell-based assays using human embryonic kidney (HEK) 293 cells transfected with the aquaporin-4 receptor on the surface with a secondary antihuman IgG along with an immunofluorescent tag has markedly improved both sensitivity and specificity of AQP4-IgG assay for NMOSD (29, 34). This cell-based assay technique can either involve fixed inactivated cells or live cells with detection of immunofluorescence either visually or with flow cytometry and a fluorescent activated cell sorting (FACS) technique (Fig. 3). It is recognized that serum yields higher sensitivity and similar specificity to CSF for detecting AQP4-IgG, and thus serum assessment is generally recommended over CSF (35). Analysis of 1371 consecutive serum samples sent for AQP4-IgG testing with a live cell-based FACS assay at Mayo Clinic over 2 years found 41 positives (3%) with a specificity of 100% (36). The absence of any false positives in this study highlighted its excellent utility as a diagnostic biomarker of NMOSD.

Fig. 3.

Serum aquaporin-4 IgG and myelin oligodendrocyte glycoprotein-IgG screening assessment using a live cell-based FACS assay. Heat inactivated patient serum (56 °C for 35 minutes) is added to live human embryonic kidney (HEK)-293 substrate cells transiently transfected with human aquaporin-4 (AQP4) M1-isoform or full length recombinant human myelin oligodendrocyte glycoprotein (MOG), both cloned into pIRES2-AcGFP vector that coexpresses nonlinked green fluorescent protein (GFP). The median fluorescence intensity (MFI) of bound AlexaFluor 647 antihuman IgG (G heavy and light chain–specific) for aquaporin-4-IgG or antihuman IgG1 (Fc region specific) for MOG-IgG is determined for both nontransfected and transfected cells. A ratio of the MFI values for GFP+ and GFP- cells is the immunoglobulin G (IgG) Binding Index (IBI). An IBI ≥2 is considered positive for AQP4-IgG and IBI ≥2.5 is positive for MOG-IgG. Panel (A) shows a negative AQP4-IgG sample by dividing the median GFP positive by median GFP negative (537/572) for an IBI of 0.94 (normal, <2). Panel (B) shows a positive AQP4-IgG result by dividing the median GFP positive by median GFP negative (11431/1036) for an IBI of 11.0 (normal, <2). Panel (C) shows a positive MOG-IgG result by dividing the median GFP positive by median GFP negative (3227/370) for an IBI of 8.85 (normal, <2.5). Positive samples are then diluted to determine the end-titer at which the IBI remains above the cutoff (not shown here). Abbreviations: AQP4, aquaporin-4; FACS, fluorescence activated cell sorting; GFP, green florescent protein; HEK, human embryonic kidney; IBI, immunoglobulin G binding index; MFI, median fluorescence intensity; MOG, myelin oligodendrocyte glycoprotein.

MOG antibodies (MOG-IgG) were first detected by enzyme-linked immunosorbent assay or western blot without using MOG in its conformational form, and were initially studied as a biomarker of MS but results were inconsistent and not reproducible. Subsequently, when utilizing MOG-IgG assays with MOG in its conformational state, it was found that MOG-IgG defined a disease distinct from MS (37). This disease was initially noted to have a predilection for children, and acute disseminated encephalomyelitis was a common manifestation (37). With improved cell-based assays with HEK293 cells transfected with full length MOG in its native conformational state, the MOGAD disease spectrum of manifestations became clearer (Table 1). While fixed and live cell-based assays for MOG-IgG are both commercially available, it is now recognized that the live-cell-based assays appear to yield higher sensitivity and specificity than inactivated cell-based assays (Fig. 3) (38, 39). An international multicenter assay comparison study showed that the serum MOG-IgG detection using cell-based assays shows excellent agreement for high positives and negatives, but, unsurprisingly, borderline or low positive results are often discordant among laboratories (40). It is recognized that low positives with cell-based assays can still occur in 1–2% of neurologic disease controls (including MS), particularly at low titer (41, 42). Serum is generally optimal for MOG-IgG assessment, although cases of isolated CSF MOG-IgG positives with a compatible MOGAD phenotype have been described, and further studies are required to determine its exact role (43). In 1260 consecutive serum samples sent for MOG-IgG testing at Mayo Clinic over 2 years for MOG-IgG testing, 92 positives (7.3%) were identified and MOG-IgG had a specificity of approximately 98% for MOGAD diagnosis (42). However, the positive predictive value (true positives/all positives) was titer dependent, with high titer positives of 1:1000 having 0% false positives, moderate titers of 1:100 having 14% false positives, and low titers 1:20 or 1:40 having an approximately 50% risk of false positivity (42). Testing in low probability situations increased the risk of false positivity in this study, highlighting the importance of reserving MOG-IgG testing for those with suspicious features for MOGAD and avoiding uniform testing in all MS patients (42). Caution is advised in the interpretation of low positive MOG-IgG results with atypical presentations for MOGAD. Overall, the risk of false positivity using cell-based assays for detection appears to be greater with MOG-IgG than AQP4-IgG (36, 42).

While persistence of AQP4-IgG titers does not appear to have prognostic value, with MOG-IgG persistence of the antibody and higher titers increase the risk of developing relapsing disease over following a monophasic course (44). Nonetheless, treatment decisions are generally made on clinical grounds and attack-prevention treatments are generally not initiated until patients develop a second attack to avoid the risk of exposing patients that are destined to never relapse and have a monophasic course (6).

Laboratory Tests with Potential Use for Disease Activity and Disease Prognosis

Neurofilament Light (NfL) in MS

Neurofilaments (NFs) are exclusively located in the neuronal cytoskeleton and are released to the interstitial fluid on axonal injury and/or neurodegeneration (45). NF concentrations in CSF and blood have been shown to correlate with the extent of axonal damage or neurodegeneration in various diseases (45). Of the family of NF proteins, the neurofilament light chain (NfL) has gained the most interest as a candidate marker of neurodegeneration. During axonal damage, NfL is released in into the CSF, and eventually into the blood, where concentrations are 40-fold lower than in the CSF (46). NfL concentrations in blood reflect the extent of axonal damage, making them a generic marker of disease activity.

In MS, NfL is increased in the CSF and blood of newly diagnosed patients, and concentrations correlate with disease severity and prognosis. Blood NfL measurements in newly diagnosed MS patients can predict brain atrophy (47) and lesion load on MRI (48). When evaluated in the context of clinical relapse, blood NfL concentrations showed increases beginning approximately 5 months prior to relapse, with a peak at clinical onset, and recovery within 4–5 months of remission (49). The use of blood NfL in serial disease monitoring and treatment response has been evaluated in various prospective clinical trials. Reductions in NfL concentrations after different treatments tend to follow the hierarchy of treatment efficacy, with greatest reductions observed with the most intensive treatments. A study that included over 1000 MS patients receiving various treatments reported the largest reductions in plasma NfL concentrations following alemtuzumab treatment (54% reduction), and the smallest reduction with teriflunomide treatment (7%) (50).

As a diagnostic marker, NfL might not provide additional value to the already available diagnostics tools, since its lack of specificity does not allow differentiation from MS mimics such as AQP4 + NMOSD, MOGAD, or other neuroinflammatory disorders (51, 52). Furthermore, as a generic marker of neurodegeneration, conditions such as Alzheimer disease, amyotrophic lateral sclerosis, Parkinson disease, Huntington disease, traumatic brain injury, and stroke, have also been shown to result in increased concentrations of NfL in CSF and blood. NfL has been suggested as a biomarker to monitor disease activity and treatment response; however, at this point, there is no consensus on how to implement its use in clinical practice. The use of NfL in clinical trials might be a valuable outcome measure and could help further define its best clinical application. The adoption of NfL in the clinical setting requires a few considerations. Currently, assays are not standardized, therefore, variability between clinical laboratories will be expected. This is particularly important if using NfL in serial disease monitoring and treatment response. Optimal frequency for testing, and the amount of change that would be considered significant, are not defined. Finally, robust reference intervals that reflect the change in NfL related to normal aging and take into consideration factors affecting NfL concentrations in blood need to be established for accurate result interpretation.

Serum Glial Fibrillary Acidic Protein

Serum Glial Fibrillary Acidic Protein (GFAP) is detected as a result of astroglial injury in both the CSF and serum. Interest in GFAP as a marker of CNS demyelinating disease stems from the fact that the protein is mainly derived from the CNS. A multicenter study observation was that CSF GFAP was found to be increased in AQP4-Ab+ patients but not in MOG-Ab+ and MS cases (53). This finding was replicated in another study that assessed the correlation between serum GFAP in patients with NMOSD vs controls (healthy and relapsing-remitting MS). The study concluded that serum GFAP increases correlate with the risk of subsequent attacks, attack severity, and response to treatment in NMOSD patients (54). The assay is not widely available to clinical laboratories and its broad adoption will require further studies validating initial findings.

Myelin Basic Protein

MBP is the second most abundant protein in CNS myelin. Together with others, proteolipid protein and myelin-associated glycoprotein and MOG, MBP is an essential component of the oligodendrocyte surface membranes, all of which have fundamental roles in the formation, maintenance, and disintegration of myelin sheaths. MPB concentration tends to be increased in both AQP4 + NMOSD and MOGAD (53). During demyelination, MBP and its fragments are released into the CSF and can be used as index for damage (55). Myelin basic protein concentrations in CSF were previously tested as predictor of clinical progression. CSF MBP increases in acute demyelination, however, MBP levels did not rise as the disease progressed (55, 56). Although a potential marker of acute demyelination, MBP does not seem to correlate with MS clinical progression. A difference between AQP4 + NMOSD and MOGAD is that in the former, GFAP concentrations are increased, whereas in the latter, MBP is increased; however, the increased MBP does not differentiate MOG-AD from MS (53, 57). Like GFAP, the test is not widely available and further studies are necessary to recommend routine testing of this marker.

Oligoclonal Banding and KFLC

A common question in the field is whether the number of OCB correlates with disease activity or outcomes. Conventionally, the test is only diagnostic and there is no correlation of the number of bands with disease severity or prognosis, so there is no recommendation for repeat testing over time. The 2017 revised McDonald criteria considered the identification of 2 or more CSF unique OCB a positive result, yet a few patients might have suggestive symptoms with no or one CSF unique band only. If the patient has a clinical or radiological picture highly suggestive of MS, yet not fulfilling the criteria, repeat testing might be warranted. One small study followed 27 patients with a single CSF specific OCB with serial OCB testing. Nine patients had 2 or more CSF specific OCB in 6 months with a definitive MS clinical phenotype, whereas the rest of the patients either had a clearance of the band or persistence with no clinical symptoms suggestive of MS over a 6-year follow-up period (58).

The prognostic value of KFLC in the CSF needs further study. There seems to be value in a high kappa index measured at diagnosis to correlate with CIS conversion to MS, whereas correlation between KFLC and the severity of MS is not as clear (59, 60).

Laboratory Tests to Identify Risk Factors for MS

Environmental risk factors have been linked to increased disease prevalence: viral infections by Epstein–Barr Virus (EBV) (61), limited sunlight exposure, and geographical latitude further from the Equator, as well as low vitamin D (62). Low vitamin D is suspected to be a risk factor, but as there can be a prodromal phase prior to clinical symptoms of MS, it is possible that vitamin D deficiency could occur secondary to that phase rather than precede the disease onset. Over 200 genetic variants have been associated with MS, but none has a strong effect. The strongest association is with class I and class II alleles of the MHC, HLA-DRB1 locus (63), and other susceptibility genes linked to cytokine expression (IL7R, CD6, IL2RA) have shown only a modest effect (64, 65). Overview information on EBV serology, vitamin D, biotin supplementation, and cytokines, is summarized in Table 3.

Table 3.

Highlights on MS risk factors and popular supplementation that may impact test results.

	Mechanism of action or association with MS	Recommendation
EBV	EBV is the most common virus with MS, previously associated with different types of lymphoma. EBV is a double stranded DNA virus transmissible through saliva, which usually presents with mononucleosis. It is not clear how EBV infection increases the prevalence of MS, although it is thought that the infection mediates molecular activation of antigen-presenting cells and autoreactive T cells (4).	The risk of MS in EBV-negative individuals is 15 times lower than that of the EBV positive individuals. It is of interest that the frequency of EBV positivity in children is much less in those with MOGAD than MS (66). EBV serology testing can be considered during MS work-up.
Vitamin D as a risk factor and supplementation	The bioactive form of vitamin D-1, 25 dihydroxyvitamin D can aid the T- and B-cell differentiation and maintains dendritic cells' tolerogenic status, which in turn suppresses T-cell activation. Interest in Vitamin D in relation to MS started in the early 1970s as an observation to the increased prevalence of MS as one moved further from the equator with less UV light exposure. Studies suggest there is a causal relationship between low vitamin D and the risk of MS. Evidence available to date suggests that serum vitamin D not only affects the risk of developing MS, but also modifies disease activity in patients with MS. Across multiple trials, associations between vitamin D and MS and its activity are generally stronger for MRI than for clinical outcomes, which may be due to the higher sensitivity of MRI compared to typical clinical parameters. Vitamin D is also noted to be low in AQP4 + NMOSD although there is limited data in MOGAD.	Testing for vitamin D is common practice. Though there is increasing evidence indicating that lower vitamin D is associated with increased risk of MS and with greater clinical and brain MRI activity in established MS, clinical trials of vitamin D supplementation are yet to show a benefit.
Biotin supplementation	Biotin has been discussed as a potential treatment in MS by promoting remyelination. Biotin is a cofactor involved in fatty acid synthesis and it is essential for carboxylases that protect against hypoxia-driven axonal degeneration by augmenting energy production in neurons. An initial high dose biotin human study showed promising results in terms of reduced disability, however in a phase 3 trial, all predefined endpoints were negative and 39 (12%) biotin-treated patients versus 29 (9%) placebo-treated patients reached the primary endpoint, hence this study concluded that biotin did not significantly improve disability or walking speed in patients with secondary progressive MS, and recommended against its use in secondary progressive MS (67). The use of high doses of biotin in MS prompted FDA warnings in several laboratory tests employing biotin as a reagent, so that assays should be validated to tolerate up to 1200 ng/mL of biotin as interferences or have a caution listed in the package insert or lab test catalog warning providers and patients about a potential interference. In non-MS biotin use, the expected biotin concentrations in serum range from 5 ng/mL at trough to 300 ng/mL at peak, significantly lower than the concentrations observed in MS potential use.	Biotin supplementation is no longer supported by clinical evidence for MS treatment (67).
Cytokines	Proinflammatory cytokines such as IL-17, IL-22, TNF-α, IL-1, IL-12, and IFN-γ may play a role in the pathogenesis of MS through several signaling pathways. Antiinflammatory circulating cytokines such as IL-4 and IL-10 are observed to be reduced in MS patients (64, 65). IL-1, IL-6, IL-8, and IL-10 have been noted to be elevated in both AQP4 + NMOSD and MOGAD (68).	Further studies are needed to establish the cause/effect relationship of cytokines in the pathogenesis of MS, AQP4 + NMOSD, and MOGAD. Monoclonal antibodies targeting elevated cytokines or their receptors have been used to treat these disorders and represent avenues to explore for developing novel therapeutics.

Sample Selection and Preanalytical Considerations

Sample selection and tests requested is an integral part of the diagnostic process. The clinical picture and clinical presentation guide the physician to appropriate testing. When testing for rare biomarkers, in general, even if the assay has a high specificity, when the prevalence of a marker is low, the true positive results can be outnumbered by the false positive results. Both AQP4 + NMOSD and MOGAD are rare diseases held as MS mimics. MS appears to be approximately 50 times more common than AQP4 + NMOSD in the USA and similarly more common than MOGAD (69). If all MS patients were to be tested for MOG-IgG, for example, there would be a risk of false positivity from testing in low probability situations. Even though the test specificity is approximately 98%, for every 100 patients tested that would yield approximately 2 false positive MOG-IgG results for every 2 true positive MOG-IgG results detected (70). Thus, testing for AQP4 and MOG antibodies should include only patients with clinical and radiologic features suggestive of those AQP4 + NMOSD and MOGAD. Uniform screening of all MS patients for AQP4-IgG and MOG-IgG is not recommended. Testing should be reserved for patients with an atypical MS clinical picture correlating with these MS-like disease entities, as outlined in Table 1.

CSF and serum samples are needed for different tests. In an initial work-up, BBB integrity problems and infectious processes can be ruled out using a comparison of results between both specimen types. Once the differential begins to center on the demyelinating conditions, the serum and CSF samples are likely already available in the laboratories. OCB is part of the McDonald criteria and requires CSF and serum run side-by-side on a gel (Fig. 3). The CSF IgG index and variations of the kappa index calculations all require a paired serum sample to CSF. For AQP4 + NMOSD or MOGAD testing, serum is the optimal and most cost-effective specimen for both antibodies (34, 71). Even if a reflex strategy is employed starting with KFLC concentration measurement in CSF, there is a chance OCB will be performed and hence a serum sample needed. In summary, at this time, if there is high suspicion of MS, both CSF and serum samples should be collected for the patient.

Future Directions

AQP4 and MOG antibodies are an integral part of the disease entities AQP4 + NMOSD and MOGAD, forming a backbone of their diagnostic criteria (6, 44, 71). That is not the case for MS, where a single antigenic target is still lacking. Even animal models of MS with different knock-out genes are imperfect. In the search for a unique antigenic target for MS, mass spectrometry has shown potential for future application, as the study of CSF proteins in healthy controls and disease states evolves. Our group has shown oligoclonal peaks on a time-of-flight mass spectrometry platform, similar to oligoclonal bands by IEF (72). The study was promising as pattern recognition tool, but challenges arose for data interpretation and definition of a positive peak in terms of signal-to-noise. Limiting implementation, there was no improvement in laboratory workflow compared to OCB for a high-volume laboratory and no elucidation of a single antigenic target, as all IgGs were being analyzed. Other groups have taken advantage of the specificity of mass spectrometry to survey metabolomics patterns (73) and lipidomics unique signatures (74) in MS. These studies can identify pathways involved in disease, and potentially future targets for diagnosis, prognosis, and even therapies. Other technologies, such as single-cell RNA sequencing and flow cytometry, identify CSF cell types (75), and differential expression of miRNAs (76) to be employed as potential diagnostic markers. Machine learning may become essential to connect the networks of findings. These proof-of-concept studies are exciting and add another toolset for laboratorians to continue discovery as we gain more insight into how MS will be diagnosed and followed-up in the future. Automated and objective tests are one step forward in this journey. A partnership between clinical laboratories and MS clinicians should allow new discoveries and biomarkers to be incorporated in disease guidelines, diagnostic criteria, and continue to improve clinical practice.

Nonstandard Abbreviations

MS, multiple sclerosis; CNS, central nervous system; CSF, cerebrospinal fluid; AQP4, aquaporin-4; MOG, myelin oligodendrocyte glycoprotein; CIS, clinically isolated syndrome; NMOSD, neuromyelitis optica spectrum disorder; MOGAD, myelin oligodendrocyte glycoprotein-IgG antibody-associated disease; BBB, blood–brain barrier; OCB, oligoclonal bands; PEP, protein electrophoresis; IEF, isoelectric focusing; KFLC, kappa free light chains; HEK, human embryonic kidney; FACS, a fluorescent activated cell sorting; NfL, neurofilament light; GFAP, Serum Glial Fibrillary Acidic Protein; EBV, Epstein–Barr Virus.

Author Contributions

All authors confirmed they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved.

Employment or Leadership

E.P. Flanagan, editorial board member of the Journal of the Neurological Sciences and Neuroimmunology Reports; M.A. Willrich, CAP diagnostic immunology and flow cytometry committee (DIFCC), AACC clinical diagnostic immunology division, Association of Medical Laboratory Immunologists.

Consultant or Advisory Role

A. Algeciras-Schimnich, advisory boards for Roche Diagnostics and Fujirebio Diagnostics; E.P. Flanagan, advisory boards for Alexion, Genentech, and Horizon Therapeutics, medical advisory board of the MOG project; M.A. Willrich, advisory board for Myeloma360; S.J. Pittock, Alexion, MedImmune/Viela Bio, Euroimmun, Roche/Genentech, Sage Therapeutics, Prime Therapeutics, UCB, Astellas, Roche SAkuraBONSAI Steering Committee.

Stock Ownership

None declared.

Honoraria

E.P. Flanagan, Pharmacy Times; M.A. Willrich, Myeloma360; S.J. Pittock, UCB, Inc., Hoffman/LaRoche AG, Alexion.

Research Funding

E.P. Flanagan, NIH (R01NS113828); S.J. Pittock, grants from Alexion Pharmaceuticals, Inc., MedImmune/Viela Bio/Horizon, Grifols, Autoimmune Encephalitis Alliance, NovelMed, Genentech/Roche; M.A. Willrich, research grant support from Sebia, Siemens Healthineers, and The Binding Site.

Expert Testimony

None declared.

Patents

S.J. Pittock has a Patent no. 8,889,102 (Application no. 12-678350)—Neuromyelitis Optica Autoantibodies as a Marker for Neoplasia-issued; a patent no. 9,891,219B2 (Application no. 12-573942) —Methods for Treating Neuromyelitis Optica (NMO) by Administration of Eculizumab to an individual that is Aquaporin-4 (AQP4)-IgG Autoantibody positive-issued; a patent GFAP-IgG pending, a patent Septin-5-IgG pending, a patent MAP1B-IgG pending, and a patent KLHL11 pending, PDE10A.

Other Remuneration

E.P. Flanagan, royalties from UpToDate. E.P. Flanagan was a site primary investigator in a randomized clinical trial on Inebilizumab in neuromyelitis optica spectrum disorder run by Medimmune/Viela-Bio/Horizon Therapeutics. E.P. Flanagan, personal fees and nonfinancial support from Alexion Pharmaceuticals, Inc. and Viela Bio/Horizon, personal fees from Astellas, UCB, and Roche/Genentech; S.J. Pittock, personal compensation from UCB, Inc. to attend the UCB Advisory Board Meeting in Stockholm, Sweden on 9/10/19, personal compensation from Hoffman/LaRoche AG to attend the NMOSD Biomarker Advisory Board Meeting in Munich, Germany on 11/8/19; personal compensation from Roche SAkuraBONSAI, travel expenses paid by Alexion to attend the NMOSD Advisory Board Meeting March 8–10, 2019, and the NMOSD Advisory Board Meeting June 1, 2019, personal compensation from Genentech to attend the virtual Genentech NMOSD Advisory Board Meeting 10/1/2020.

Acknowledgments

The authors would like to thank Ms. Melissa S. Derksen, instructor in laboratory medicine and pathology, and Ms. Jamie M. Shulman, clinical laboratory technologist, both at Mayo Clinic, Rochester, MN, for their assistance with the images of testing shown in this review.

Author notes

Authors’ Disclosures or Potential Conflicts of Interest:Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest: