Neuroglial markers of damage in autoimmune neurology

Abstract

The advancements in assays for the detection of neuroglial markers of damage, such as neurofilament light chain and glial fibrillary acid protein, have led to a growing interest in their investigation in neurological disorders. In this review, we aim to summarize evidence regarding the role of biomarkers of neuroglial damage in autoimmune neurological disorders of the central nervous system, including myelin oligodendrocyte glycoprotein antibody-associated disease, neuromyelitis optica spectrum disorder, autoimmune encephalitis, and paraneoplastic neurological syndromes. We will discuss how these biomarkers can help our understanding of disease mechanisms and how these results can be potentially translated into clinical practice, helping clinicians in the differential diagnosis, in prognostication, and prediction of the risk of relapse.

Keywords

autoimmune encephalitis biomarker glial fibrillary acid protein myelin oligodendrocyte glycoprotein antibody-associated disease neurofilament light chain neuromyelitis optica paraneoplastic neurological syndrome

Introduction

In recent years, the improvement of diagnostic assays for the detection of markers of neuroglial damage, such as neurofilament light chain (NfL) or glial fibrillary acid protein (GFAP), have allowed these proteins to be routinely detected in serum or plasma samples, and not only in the cerebrospinal fluid (CSF), as for older generation enzyme-linked immunoassays (ELISA).^1,2

These advancements have led biomarker investigation in easily accessible samples, leading to further studies of neurological disorders in which CSF analysis is not routinely performed, and also allowing longitudinal evaluation of samples.

Among inflammatory neurologic disorders, neuroglial biomarkers have been widely investigated in multiple sclerosis (MS), but recently, a growing body of evidence has supported their potential role also in less frequent autoimmune disorders of the central nervous system (CNS).³ Biomarkers of neuroglial damage may have a potential role in clinical practice, including helping clinicians in the differential diagnosis and predicting disease course or relapses.

The aim of this review is to provide an updated overview of neuroglial markers of damage in the field of rare autoimmune disorders of the CNS. The discussion of biomarkers in MS, the most common CNS neuroimmune disorder, goes beyond the scope of this review and thus will not be included, even though MS will be mentioned in the manuscript for comparative purposes.

Myelin oligodendrocyte glycoprotein antibody-associated disease and neuromyelitis optica spectrum disorder

Myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) and neuromyelitis optica spectrum disorder (NMOSD) emerged as distinct CNS inflammatory disorders, with unique clinico-radiological, immunological, and pathological characteristics. As a consequence, the differential diagnosis is of utmost importance to administer proper treatment strategies.⁴

MOGAD, as an incidence of ≈3–4 million-person-years, affects both sexes almost equally, and both children and adults, with approximately 30% of patients presenting below the age of 18 years and 25% over 50 years.⁵ The most common clinical manifestations are optic neuritis, myelitis, acute disseminated encephalomyelitis (ADEM), brainstem/cerebellar involvement, and cortical encephalitis.^6,7 Diagnosis is based on the presence of serum/CSF MOG antibodies (MOG-Abs) and a compatible clinical/radiological phenotype.⁸ A relapsing disease course is reported in 40%–80% of MOGAD, and many patients show a favorable recovery, with 10% experiencing a poor outcome.^9,10 Disease course and final outcome are highly unpredictable. Acute treatment strategies include intravenous steroids and intravenous immunoglobulins (IvIg)/plasma exchange. The use of long-term immunosuppressive treatment in MOGAD is still a matter of debate, as a relapsing course is observed in about half of cases; thus, the benefits and risks of immunosuppressants should be carefully evaluated in each case.¹¹ Immunosuppressive treatment can be proposed in patients with incomplete recovery from the onset of an attack, clinically severe/life-threatening first attack, or evidence of a relapsing disease course, and includes among the more common off-label options oral immunosuppressants (azathioprine, mycophenolate mofetil), tocilizumab, monthly IvIg, and rituximab. Trials with satralizumab, rozanolixizumab, azathioprine, and CAR-T treatment are currently ongoing.⁶

NMOSD is a rare CNS disorder with the highest estimated incidence/prevalence in the Afro-Caribbean region and Asia, and the lowest in Western countries and Australia/New Zealand.¹² NMOSD is more common in females (9:1) and peaks in middle-aged adults, with rare occurrence in children.¹³ The main clinical manifestations include optic nerve, spinal cord, and/or are postrema involvement with compatible radiological involvement. Although a subgroup of patients is antibody negative, the diagnosis of NMOSD is based on the detection of antibodies to the water channel aquaporin-4 (AQP4-Abs), in serum.¹⁴ NMOSD is a severe condition with a high risk of disability and a relapsing disease course.¹⁵ Acute treatment includes intravenous steroids and plasma exchange/IvIg, whose prompt administration is of utmost importance to reduce final disability.¹⁶ Due to the frequently relapsing nature of the disease (about 90% of cases),¹⁷ chronic immunosuppression is always recommended.¹⁸ Long-term oral corticosteroids and/or steroid-sparing agents such as azathioprine, mycophenolate mofetil, and methotrexate have been used in NMOSD and are still employed in many countries as first-line agents.¹⁹ B-cell depleting agents, in particular rituximab, are also widely used in clinical practice as first-line treatments, and their role has been further supported by a recent clinical trial.²⁰ In recent times, clinical trials have shown the efficacy of drugs targeting disease mechanisms of NMOSD, such as complement inhibition (eculizumab/ravulizumab), anti-IL6 receptor blockade (satralizumab), and depletion of anti-CD19-positive cells (inebilizumab), leading to their approval by regulatory agencies.^21–25 In contrast to MOGAD,^26,27 where neoplasms are rarely encountered, AQP4 + NMOSD may have a paraneoplastic origin, in particular in male patients presenting with longitudinally extensive myelitis in association with adenocarcinomas.^26,27

Biomarkers in MOGAD

The extent of axonal and glial damage in MOGAD has been investigated in several studies, but remains unclear.²⁸ Increased serum NfL has been reported in MOGAD, with higher levels during attacks and fluctuating over the 3 months following relapse, usually normalizing within 6 months.^29–34 Some contradictory findings exist, with similar NfL levels in relapses and remission phases despite a correlation with disability at the time of sampling.³⁵ NfL correlates with attack severity, but some studies report that it does not predict disease course or correlate with clinical/radiological phenotypes, though one study found higher levels in patients with brain MRI abnormalities and seizures.^32,36 Longitudinal studies suggest NfL levels remain stable or decline over time in MOGAD, indicating that axonal injury mainly occurs at disease onset.³⁷ A recent study showed that NfL decreases over time, is associated with disability at attack and radiological activity, independently predicts relapse risk (with higher concentration predicting shorter time to second relapse), and longitudinal changes forecast clinical recovery (with lowering concentrations during follow-up predicting an EDSS score <3 in monophasic patients). This highlights the potential value of monitoring NfL for predicting relapses and recovery. Treatment did not appear to affect NfL levels, and increases independent of clinical events were not observed.³⁸

Data on GFAP are inconsistent: GFAP is typically lower in MOGAD compared to NMOSD, with some reports showing stable levels during relapse and remission without correlation to disability,³⁹ while others observed increased GFAP during relapses correlating with disability.^30,33,35 The serum NfL/GFAP ratio has been proposed as a useful marker to distinguish NMOSD, MOGAD, and MS.³³ A recent study found that GFAP levels decrease over time, correlate with disease severity, and may predict clinical recovery. GFAP increase at the first attack seems to be minor compared to NfL peak, with less pronounced fluctuations over time and early normalization.³⁸ Neither NfL nor GFAP increases during inter-attack periods, suggesting no attack-independent axonal or glial injury.³⁰

Serum NfL is also elevated in pediatric MOGAD cases, particularly in those presenting with ADEM-like symptoms, encephalopathy, or longitudinally extensive transverse myelitis.^40,41 NfL rises during clinical or radiological disease activity in children (within 20 days of onset), supporting its role as a biomarker for disease activity.³⁸ While NfL alone does not predict relapsing disease, the ratio of NfL-to-MOG antibodies is significantly lower in relapsing than in non-relapsing pediatric cases, suggesting its potential utility for predicting disease course in this population.⁴²

Of note, a recently published systematic review and meta-analysis report that neither NfL nor GFAP correlates with disease activity at sampling, although the heterogeneity of the included study might have influenced the results.⁴³

Biomarkers in NMOSD with AQP4-Abs

Markers indicating damage to axons and glial cells have demonstrated notable associations with disability and disease activity in NMOSD.²⁹ Earlier research reported elevated CSF GFAP levels during relapses, which were higher than those detected in MS. CSF GFAP levels tend to decline following treatment and are linked with disability both at the time of attacks and over the follow-up.⁴⁴ CSF NfL concentrations are also increased in NMOSD and correspond with disease activity. Some studies reported comparable NfL levels between NMOSD and MS, while others detected higher levels in NMOSD.^45,46 These studies mostly employed first-generation ELISAs for CSF biomarker detection, as these early assays could not reliably identify the lower biomarker concentrations present in serum. This limited the translation into clinical practice of biomarkers and did not allow longitudinal studies.

More recently, advanced sensitive methods have facilitated the measurement of these biomarkers in serum, providing a useful tool for longitudinal monitoring. The strong correlation between CSF and serum levels further supports serum as a more accessible sample source.⁴⁷ Both serum NfL and GFAP generally rise in NMOSD patients, with GFAP showing particularly high elevations, likely reflecting predominant astrocyte involvement. A limited number of studies reported similar GFAP levels in NMOSD patients and controls, primarily involving clinically stable cases.³⁹ This suggests that GFAP may help predict AQP4 antibody positivity and assist in distinguishing NMOSD from MS. The serum GFAP/NfL ratio during relapse has been shown to be useful for differentiating MS from NMOSD, with higher values achieving diagnostic sensitivity of 73% and specificity of 75.8%.^{29,32,33,47–50} NfL levels also support differentiation between inflammatory and vascular conditions, with higher levels and more common in the latter.⁵¹

During NMOSD attacks, NfL and GFAP levels rise and correlate with disability measured by EDSS at the time of attacks, with GFAP showing the most consistent association. Recently, Z-score cutoffs of 3.0 for GFAP and 2.1 for NfL have been established to distinguish attack phases from remission in NMOSD.⁵² However, some evidence indicates that NfL levels do not predict recovery post-attack, and neither NfL nor GFAP reliably forecasts future relapses.^47,52–56 Importantly, data from the N-Momentum trial analyzing inebilizumab’s effect in NMOSD showed that NfL strongly correlates with disability during attacks and predicts worsening disability post-attack, whereas GFAP levels can predict future attacks.^57,58 Moreover, serum NfL, but not GFAP, may be useful for predicting spinal cord atrophy in NMOSD.⁵⁹

Treatment with rituximab appears to influence biomarker levels, preventing significant serum GFAP elevation and reducing occurrences of subclinical GFAP increases.^53,54 Tocilizumab similarly lowers NfL and GFAP concentrations, supporting the potential use of these biomarkers in monitoring therapeutic response.⁵⁶ According to findings from the NMomentum trial, inebilizumab also decreases GFAP and NfL levels compared to placebo during attacks and over time.^57,58

The dynamics of biomarker levels should be considered when interpreting NfL and GFAP in NMOSD patients, as GFAP peaks within 1 week of an attack and rapidly declines over the following 12 weeks, whereas NfL peaks approximately 5 weeks post-attack and decreases more gradually over 20 weeks.⁵² Increases in biomarker levels have also been reported in the week prior to relapse.^57,58 In addition to disease phase and timing, factors such as age, BMI, coexisting conditions (which may affect biomarker clearance), and sex (particularly for serum GFAP) need to be considered.^52,60,61 Age is also a relevant factor that influences serum NfL and GFAP levels.^50,62

Lastly, recent evidence suggests that increased CSF concentrations of 14-3-3 protein can distinguish AQP4 + NMOSD from MOGAD and MS and track with disease status, but further independent validation of these results is warranted.⁶³

Biomarkers in double seronegative NMOSD

Double seronegative NMOSD remains a poorly understood condition with overlapping clinical-radiological features with AQP4 + NMOSD, but with scant evidence supporting shared disease mechanisms. As an example, IL6 blockade is an effective treatment for AQP4 + NMOSD, but no clear benefit was observed in double seronegative cases, implying heterogeneity in disease mechanisms.²⁵ So far, limited studies have examined biomarker profiles in seronegative NMOSD. CSF levels tend to be lower in seronegative patients compared to those positive for AQP4 antibodies, with very high GFAP levels (>2050 ng/L) during relapse reported only in seropositive cases.^64,65 However, some seronegative cases show increased GFAP, implying a possible unknown antibody targeting astrocytes and causing astrocytic damage.⁶⁶ Recent serum studies confirmed significantly higher GFAP concentrations in seropositive (median 308.3 pg/mL) versus seronegative (median 103.4 pg/mL) NMOSD patients. Interestingly, two seronegative patients exhibited high GFAP, supporting astrocytic injury in a subset of seronegative NMOSD.⁶⁷ NfL levels did not differ between seropositive and seronegative patients and, notably, did not correlate with disability in seronegative patients.⁶⁷

Autoimmune encephalitis and paraneoplastic neurological syndromes

Autoimmune encephalitis (AE) is characterized by an abnormal immune response against neuronal and glial proteins, leading to autoimmunity. Although infrequent in terms of epidemiology,⁶⁸ AE is one of the most frequent causes of rapidly progressive cognitive impairment and represents a potentially treatable condition.^69,70 Different clinical syndromes fall under the umbrella term of AE, including limbic encephalitis, extra-limbic encephalitis, brainstem encephalitis, and immune-mediated cerebellar ataxia. These syndromes have heterogeneous clinical manifestations, depending on the site of inflammation and specific associated autoantibody signatures.⁷¹

On a biological standpoint, AE can be associated with antibodies targeting neuronal surface or intracellular proteins. Antibodies targeting neuronal surface antigens can bind their target in vivo and are the pathogenic cause of the disease through cross-linking and antigen internalization, functional blocking of the cognate receptor, and blocking of protein–protein interactions.⁷²

On the contrary, antibodies targeting intracellular proteins cannot access their targets in vivo and lack a pathogenic potential. Intracellular antigens are often shared between the nervous system and an underlying tumor, leading to an immune response against cancer cells and neurons. This immune response is mainly mediated by cytotoxic CD8+ T cells, which are responsible for early neuronal loss. The double-edge immune response against cancer and neuronal tissue defines paraneoplastic neurological syndromes (PNS).⁷³

A third group of antibodies, including glutamic acid decarboxylase isoform 65 (GAD-65) and amphiphysin, shares common features with both antibodies against neuronal and intracellular targets, as the antigens are usually located intracellularly, but can be transiently expressed on the cell surface, implying a pathogenic potential.⁷⁴

The different disease mechanisms of antibody-mediated AE and PNS mirror distinct clinical features, therapeutic approach and response, and long-term outcomes in these groups.

In antibody-mediated AE, the disease course is often stereotyped, as seen in anti-N-methyl-D-aspartate receptor (NMDAR) or anti-leucine-rich glioma inactivated 1 (LGI1) encephalitides, and specific clinical findings, such as facio-brachial dystonic seizures, may be pathognomonic.^75–77

Furthermore, some autoantibodies, such as anti-immunoglobulin-like cell adhesion molecule 5 (IgLON5), may be present with a more chronic disease course and without overt signs of inflammation, mimicking neurodegenerative conditions rather than other causes of encephalitis.⁷⁸

Treatment response is often favorable, and steroids, plasma exchange, IvIg, and rituximab are common therapeutic choices, given the B-cell-mediated pathogenesis. In about 10%–30% of patients, relapses can occur during the disease course and may resemble the first clinical event.^79–81

In PNS, CNS involvement includes “high-risk” phenotypes such as limbic encephalitis, encephalomyelitis, opsoclonus-myoclonus syndrome, and rapidly progressive cerebellar syndrome. Different autoantibodies can present with the same clinical phenotype, and, in this setting, antibodies are useful markers for cancer, helping clinicians to identify the underlying malignancy with focused investigations.⁸² Recently, the introduction of immune checkpoint inhibitors has revolutionized cancer therapy and prognosis, but immune-mediated side disorders involving the nervous system have emerged as treatment side effects.⁸³ Among neurological immune-related adverse events (n-irAEs), a subset of patients may have clinical features resembling classic PNS and have less frequent treatment response and a more unfavorable prognosis compared to other phenotypes.⁸⁴

The identification and treatment of cancer is vital in PNS, as treatment outcomes are frequently unsatisfactory due to early neuronal death. Steroids and cyclophosphamide are usually preferred due to their rapid action, blood–brain barrier permeability, and depletion of both B and T cells. Relapses are rare but can occur as manifestations of cancer recurrence.⁸⁰

Lastly, anti-GAD-65 neurologic autoimmunity encompasses autoimmune epilepsy, cerebellar ataxia, and stiff-person syndrome. The pathogenesis of GAD-65 autoimmunity is still not fully understood, and both humoral and cellular mechanisms may be involved. Response to treatment is variable across the spectrum, with stiff-person syndrome being the most responsive phenotype. Improvement can also be observed in patients with cerebellar ataxia, but seizure control in GAD-65 autoimmune epilepsy is often unsatisfactory.^85,86

Biomarkers in autoimmune encephalitides

Investigating markers of neuroglial damage can contribute to deciphering the pathogenesis, differential diagnosis, disease status, and prognosis of AE, with potential translation into clinical practice.

Many studies have demonstrated that neuroaxonal damage, reflected by increased concentrations of NfL in serum and CSF, occurs across different subtypes of AE, regardless of antibody specificity. These studies have mostly included patients with anti-NMDAR, anti-LGI1, anti-CASPR2, and anti-IgLON5 antibodies.^87–93 Of note, NfL concentration can vary across different phenotypes associated with the same autoantibody, such as in GAD65 neurologic autoimmunity. In this case, patients with cerebellar ataxia and limbic encephalitis present higher concentrations of NfL than patients with stiff-person syndrome, implying different degrees of axonal damage across the spectrum.^94,95 Similarly, patients with anti-NMDAR encephalitis usually have lower concentrations of NfL than those with other autoimmune encephalitides, mirroring the predominant synaptic dysfunction at disease onset in this condition.^88,89

GFAP concentrations in AE may mirror astrocytic damage as seen in patients with anti-GFAP antibodies, an autoimmune astrocytopathy, who have CSF GFAP concentrations comparable to NMOSD with AQP4-Abs, reflecting a T-cell-mediated astrocytic injury.⁹⁶ Alternatively, GFAP concentrations in AE may also indicate an astroglial reaction to inflammation, as hypothesized for anti-NMDAR encephalitis, in which serum GFAP concentration is increased only in the acute stage with a subsequent rapid decline and normalization after 6 months from onset.⁹⁷ In addition, in patients with anti-IgLON5 disease, GFAP concentration is lower in those patients with subacute and pauci-symptomatic disease course.⁹³

Neuronal damage also seems to occur in AE, as reflected by increased concentrations of total tau, phosphorylated tau, and 14-3-3 protein in some patients, usually to a lesser degree compared to neurodegenerative conditions.^98–101 Furthermore, reduced amyloid beta concentration can also be observed, implying altered CSF dynamics and protein clearance due to neuroinflammation.⁹⁸

Lastly, CSF markers of synaptic dysfunction, such as neurogranin, SNAP-25, beta-synuclein, and YKL-40, have been investigated with heterogeneous results. Some studies showed a reduction in synaptic marker concentration, implying antibody-mediated internalization, while others demonstrated increased concentration, suggesting their release after synaptic damage.^102,103

The dynamics of these biomarkers over time has also important implications for the understanding of AE pathogenesis. The increase in NfL is delayed compared to the nadir of clinical disability and the peak of antibody titers, suggesting that neuroaxonal damage is not the major pathogenic cause of the disease, at least in anti-NMDAR encephalitis.^90,104

The concentration of biomarkers is higher during the acute stages of the disease compared to remission, and, after the onset event, biomarkers tend to decrease over time, but may remain higher compared to reference values, suggesting residual neuroaxonal loss.^{88,92,97,101,105} Intriguingly, the dynamics of NfL concentration in patients with stable disease seems to be divergent in anti-NMDAR and anti-LGI1 encephalitides: indeed, NfL concentration tends to decline over time in anti-NMDAR encephalitis, while, in anti-LGI1 encephalitis, the decline is slower, and some patients may have fluctuating NfL concentrations during their follow-up.^92,97,105 This different trend over time could be explained by frequent subclinical and potentially under-recognized disease activity, such as seizures occurring during sleep, recently demonstrated in anti-LGI1 encephalitis.¹⁰⁶

Lastly, biomarkers tend to increase during relapses, although their concentration peaks tend to follow the clinical onset, mirroring what is observed at the onset event, but limiting the potential role of monitoring biomarkers over time to predict disease recurrence.^90,105

Markers of neuroglial damage have a potential role in the differential diagnosis between AE and its mimics, although further prospective validation in clinical practice is mandatory before real-life implementation.

NfL evaluation may be useful to discriminate anti-NMDAR encephalitis from the relevant mimics, herpes simplex virus (HSV) 1 encephalitis, and the first episode of psychosis. Indeed, NfL concentration is higher in HSV 1 encephalitis and lower in psychosis without neurological accompaniments (primary psychiatric disease) compared to anti-NMDAR encephalitis. Of note, a cut-off concentration of serum NfL of 15 pg/mL allows to correctly identify 96% of patients with first episode psychosis and 85% of patients with AE.⁸⁷ In addition, CSF NfL concentration is higher in neurodegenerative rapidly progressive dementia compared to anti-LGI1 encephalitis.¹⁰⁰

Serum NfL concentration can also be useful to discriminate anti-IgLON5 disease from other bulbar mimics such as amyotrophic lateral sclerosis and myasthenia gravis,¹⁰⁷ whereas CSF NfL concentration has moderate accuracy in differentiating autoimmune versus infectious inflammatory CNS disorders.¹⁰² In the same study, CSF markers of synaptic dysfunction, such as CSF beta-synuclein, SNAP25, and neurogranin, did not correctly discriminate infectious versus infective causes of encephalitis.¹⁰²

Markers of neuronal damage have been investigated in the differential diagnosis of rapidly progressive dementias. Total tau can be increased in patients with AE,^93,98,99 usually at a lower level compared to neurodegenerative conditions such as Alzheimer’s disease and Creutzfeldt-Jakob disease: for example, a cut-off of 100 pg/mL of total tau can be observed more frequently in patients with prion disease compared to autoimmune causes of dementia.⁹⁸ Rarely, phosphorylated tau can be increased, and amyloid beta 42 can be decreased in patients with AE, but most patients do not have a CSF profile suggestive of Alzheimer’s disease.^93,98–100 Lastly, 14-4-3 protein can be increased in patients with AE, especially in those harboring antibodies to GABA-B, and, in about 20% of cases, a CSF protein profile suggestive of CJD can be observed.⁹⁹

The association between clinical features, disease outcomes, and markers of neuroglial damage has been investigated in different studies with divergent results. These discrepancies could be explained by heterogeneity in the spectrum of AE phenotypes and antibody specificities, the use of different outcome measures, variable sample timing from disease onset, or by differences in the analyzed matrix (i.e., serum, plasma, or CSF). Resolving this heterogeneity is crucial, as it could allow the development and translation into clinical practice of multimodal biomarkers (i.e., biological, neuroimaging, optical computerized tomography) to improve diagnostic accuracy and predict disease course, similarly to what is observed in demyelinating diseases.¹⁰⁸

In anti-NMDAR encephalitis, NfL concentration is increased in patients with post-HSV-1 encephalitis or with coexistent demyelination or MRI abnormalities. CSF NfL concentration tracks with inflammatory CSF findings, movement disorders, seizures, and disease severity, while serum concentration seems to predict ICU admission. In the pediatric population, NfL predicts long-term outcomes and residual cognitive symptoms, while these findings are less consistent in adults, with divergent results regarding the correlation of long-term outcomes and NfL.^{87,88,90,97,109}

In anti-LGI1 encephalitis, serum NfL seems to be associated with seizures and to predict long-term cognitive outcomes, while CSF concentration predicts the development of epilepsy.^91,92

In anti-IgLON5 disease, serum NfL tracks with severity score at sampling, bulbar symptoms, and disability, while serum GFAP tracks with bulbar symptoms. In addition, serum NfL can predict treatment response and 1-year mortality.^93,107

In anti-GFAP encephalomyelitis, CSF NfL correlates with poor outcome and the presence of brain MRI abnormalities.⁹⁶

Lastly, studies investigating cohorts of patients with different antibody specificities and clinical syndromes have found heterogeneous correlations between concentration of serum or CSF total tau, GFAP, NfL, and markers of synaptic dysfunction with long-term clinical outcomes or with the development of hippocampal sclerosis.^{89,95,101,102,110} These divergent findings could be explained by the inclusion of patients with diseases that underlie heterogeneous pathogenic mechanisms and should be translated with caution in clinical practice.

Biomarkers in PNS

As PNS are less frequently encountered compared to “idiopathic” AE cases, less evidence about the role of markers of neuroglial damage is available for this spectrum of disorders.

Some studies have highlighted that neuroaxonal loss is a major disease mechanism in the PNS, as serum NfL concentration is higher in patients with paraneoplastic neuronal intermediate filament autoimmunity, and CSF NfL concentration is increased in pediatric opsoclonus myoclonus syndrome compared to controls.^111,112 In addition, patients harboring intracellular antigens or with an underlying malignancy tend to have higher CSF NfL concentration compared to their “non-paraneoplastic” counterparts.^113,114

In patients with neurologic adverse events from immune checkpoint blockade, in particular those harboring anti-ANNA1/Hu antibodies, serum NfL is increased at comparable levels to HSV-1 encephalitis.¹¹⁵ Furthermore serum NfL concentration, but not GFAP, can discriminate patients with neurologic irAEs from other diseases. Similarly, markers of neuroglial damage are increased in patients with nervous system irAEs compared to immune-related hypophysitis and controls.¹¹⁶ NfL increase tend to be slower compared to other biomarkers such as S100B, but can predict treatment response and correlates with worst outcomes.^115–117 NfL concentration is higher in patients with peripheral nerve involvement compared to neuromuscular toxicities, while GFAP concentration is more elevated in patients with CNS involvement.¹¹⁶

Lastly, longitudinal studies are currently lacking, but serum biomarker concentrations ((NfL, GFAP)) decrease over time and track with clinical status in some individual reports of patients with PNS or neurologic irAEs.^{115,117,118–123} Although promising in terms of potential clinical translations, the use of biomarkers to monitor these conditions is still in a preliminary phase, and longitudinal studies, also accounting for other factors such as chemotherapy-related neuroaxonal damage, are mandatory to draw definitive conclusions.

Conclusion

In this review, we summarized the current evidence regarding the role of markers of neuroglial damage in autoimmune neurology.

Soluble biomarkers can aid our understanding of disease mechanisms and pathophysiology. Furthermore, these biomarkers can help clinicians in obtaining diagnostic precision by discriminating autoimmune disorders from their mimics, or by identifying specific biomarker signatures (i.e., astrocytopathy in NMOSD with AQP4-Abs). Lastly, these biomarkers can be helpful in predicting disease course and relapses, thus representing potential aid for treatment decisions.

Limitations of available studies hinder the translation of these findings into clinical practice and should be acknowledged. The heterogeneity of sampling timing, of assays for biomarker detection, and variability within included cohorts, especially for AE, are major confounders.

Future perspectives for clinical implementation of biomarkers include the standardization of assays used for their detection, the development and validation either of cut-offs or adjusted z-scores that can be translated from a study population to an individual patient level and are clinically significant (i.e., cut-offs defining impeding relapses) and, lastly, longitudinal studies to clarify the optimal time frame to assess biomarkers concentration at diagnosis and during follow-up.

The inclusion of biomarkers in NMOSD clinical trials has greatly contributed to our understanding of their dynamics and potential role in clinical practice, highlighting the importance of conducting multicenter prospective studies for biomarker investigation and clinical practice translation, also in other autoimmune disorders of the nervous system (Figures 1 and 2).⁵⁷

Figure 1.

Immune pathogenesis and biomarkers of neuroglial damage in NMOSD and MOGAD.

Figure 2.

Summary of biomarkers of neuroglial damage investigated in patients with autoimmune encephalitis and paraneoplastic neurological syndromes.

Footnotes

Acknowledgements

None.

Declarations

ORCID iD

Alessandro Dinoto

References

Khalil

Teunissen

Otto

, et al. Neurofilaments as biomarkers in neurological disorders. Nat Rev Neurol 2018; 14: 577–589.

Petzold

Glial fibrillary acidic protein is a body fluid biomarker for glial pathology in human disease. Brain Res 2015; 1600: 17–31.

Di Filippo

Gaetani

Centonze

, et al. Fluid biomarkers in multiple sclerosis: from current to future applications. Lancet Reg Health Eur 2024; 44: 101009.

Uzawa

Oertel

Mori

, et al. NMOSD and MOGAD: an evolving disease spectrum. Nat Rev Neurol 2024; 20: 602–619.

Cacciaguerra

Sechi

Komla-Soukha

, et al. MOG antibody-associated disease epidemiology in Olmsted County, USA, and Martinique. J Neurol 2025; 272: 118.

Sechi

Gastaldi

Cortese

, et al. Myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD): practical recommendations for diagnosis and management. J Neuroimmunol 2026; 410: 578781.

Sechi

Cacciaguerra

Chen

, et al. Myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD): a review of clinical and MRI features, diagnosis, and management. Front Neurol 2022; 13: 885218.

Banwell

Bennett

Marignier

, et al. Diagnosis of myelin oligodendrocyte glycoprotein antibody-associated disease: international MOGAD Panel proposed criteria. Lancet Neurol 2023; 22(3): 268–282.

Trewin

Dale

Qiu

, et al. Oral corticosteroid dosage and taper duration at onset in myelin oligodendrocyte glycoprotein antibody-associated disease influences time to first relapse. J Neurol Neurosurg Psychiatry 2024; 95(11): 1054–1063.

10.

Lopez-Chiriboga

Sechi

Buciuc

, et al. Long-term outcomes in patients with myelin oligodendrocyte glycoprotein immunoglobulin G-associated disorder. JAMA Neurol 2020; 77: 1575.

11.

Whittam

Karthikeayan

Gibbons

, et al. Treatment of MOG antibody associated disorders: results of an international survey. J Neurol 2020; 267: 3565–3577.

12.

Papp

Magyari

Aktas

, et al. Worldwide incidence and prevalence of neuromyelitis optica: a systematic review. Neurology 2021; 96: 59–77.

13.

Arnett

Chew

Leitner

, et al. Sex ratio and age of onset in AQP4 antibody-associated NMOSD: a review and meta-analysis. J Neurol 2024; 271: 4794–4812.

14.

Wingerchuk

Banwell

Bennett

, et al. International consensus diagnostic criteria for neuromyelitis optica spectrum disorders. Neurology 2015; 85: 177–189.

15.

Flanagan

EP.

Neuromyelitis optica spectrum disorder and other non-multiple sclerosis central nervous system inflammatory diseases. Continuum (Minneap Minn) 2019; 25(3): 815–844.

16.

Abboud

Petrak

Mealy

, et al. Treatment of acute relapses in neuromyelitis optica: steroids alone versus steroids plus plasma exchange. Mult Scler 2016; 22: 185–192.

17.

Jarius

Ruprecht

Wildemann

, et al. Contrasting disease patterns in seropositive and seronegative neuromyelitis optica: a multicentre study of 175 patients. J Neuroinflammation 2012; 9: 14.

18.

Kümpfel

Giglhuber

Aktas

, et al. Update on the diagnosis and treatment of neuromyelitis optica spectrum disorders (NMOSD)-revised recommendations of the Neuromyelitis Optica Study Group (NEMOS). Part II: Attack therapy and long-term management. J Neurol 2024; 271: 141–176.

19.

Wingerchuk

Lucchinetti

CF.

Neuromyelitis optica spectrum disorder. N Engl J Med 2022; 387: 631–639.

20.

Tahara

Oeda

Okada

, et al. Safety and efficacy of rituximab in neuromyelitis optica spectrum disorders (RIN-1 study): a multicentre, randomised, double-blind, placebo-controlled trial. Lancet Neurol 2020; 19: 298–306.

21.

Cree

BAC

Bennett

Kim

, et al. Inebilizumab for the treatment of neuromyelitis optica spectrum disorder (N-MOmentum): a double-blind, randomised placebo-controlled phase 2/3 trial. Lancet 2019; 394: 1352–1363.

22.

Pittock

Berthele

Fujihara

, et al. Eculizumab in Aquaporin-4-positive neuromyelitis optica spectrum disorder. N Engl J Med 2019; 381: 614–625.

23.

Pittock

Barnett

Bennett

, et al. Ravulizumab in Aquaporin-4-positive neuromyelitis optica spectrum disorder. Ann Neurol 2023; 93: 1053–1068.

24.

Traboulsee

Greenberg

Bennett

, et al. Safety and efficacy of satralizumab monotherapy in neuromyelitis optica spectrum disorder: a randomised, double-blind, multicentre, placebo-controlled phase 3 trial. Lancet Neurol 2020; 19: 402–412.

25.

Yamamura

Kleiter

Fujihara

, et al. Trial of satralizumab in neuromyelitis optica spectrum disorder. N Engl J Med 2019; 381: 2114–2124.

26.

Dinoto

Borin

Campana

, et al. Investigating paraneoplastic aquaporin-4-IgG-seropositive neuromyelitis optica spectrum disorder through a data-driven approach. Eur J Neurol 2022; 29(11): 3466–3472.

27.

Sepúlveda

Sola-Valls

Escudero

, et al. Clinical profile of patients with paraneoplastic neuromyelitis optica spectrum disorder and aquaporin-4 antibodies. Mult Scler J 2018; 24: 1753–1759.

28.

Dinoto

Sechi

Flanagan

, et al. Serum and cerebrospinal fluid biomarkers in neuromyelitis optica spectrum disorder and myelin oligodendrocyte glycoprotein associated disease. Front Neurol 2022; 13: 866824.

29.

Mariotto

Farinazzo

Monaco

, et al. Serum neurofilament light chain in NMOSD and related disorders: comparison according to aquaporin-4 and myelin oligodendrocyte glycoprotein antibodies status. Mult Scler J Exp Transl Clin 2017; 3: 2055217317743098.

30.

Hyun

J-W

Kim

, et al. Absence of attack-independent neuroaxonal injury in MOG antibody-associated disease: longitudinal assessment of serum neurofilament light chain. Mult Scler J 2022; 28(6): 993–999.

31.

Schindler

Bellmann-Strobl

Kuhle

, et al. Longitudinal change of serum NfL as disease activity biomarker candidate in MOGAD: a descriptive cohort study. Mult Scler Relat Disord 2024; 88: 105729.

32.

Luo

Chen

Mao

, et al. Serum neurofilament light chain in adult and pediatric patients with myelin oligodendrocyte glycoprotein antibody-associated disease: correlation with relapses and seizures. J Neurochem 2022; 160: 568–577.

33.

Chang

Huang

Wang

, et al. Serum neurofilament light and GFAP are associated with disease severity in inflammatory disorders with aquaporin-4 or myelin oligodendrocyte glycoprotein antibodies. Front Immunol 2021; 12: 647618.

34.

Gomes

ABAGR

Kim

Pretzsch

, et al. Neurofilament light chain as a discriminator of disease activity status in MOG antibody-associated disease. Neurol Neuroimmunol Neuroinflamm 2025; 12: e200347.

35.

Kim

Lee

Kim

, et al. Serum biomarkers in myelin oligodendrocyte glycoprotein antibody-associated disease. Neurol Neuroimmunol Neuroinflamm 2020; 7: e708.

36.

Mariotto

Ferrari

Gastaldi

, et al. Neurofilament light chain serum levels reflect disease severity in MOG-Ab associated disorders. J Neurol Neurosurg Psychiatry 2019; 90: 1293–1296.

37.

Mariotto

Gastaldi

Grazian

, et al. NfL levels predominantly increase at disease onset in MOG-Abs-associated disorders. Mult Scler Relat Disord 2021; 50: 102833.

38.

Marignier

Villacieros-Álvarez

Espejo

, et al. Assessment of neuronal and glial serum biomarkers in myelin oligodendrocyte glycoprotein antibody-associated disease: the MULTIMOGAD study. J Neurol Neurosurg Psychiatry 2025; 96: 884–892.

39.

Schindler

Grittner

Oechtering

, et al. Serum GFAP and NfL as disease severity and prognostic biomarkers in patients with aquaporin-4 antibody-positive neuromyelitis optica spectrum disorder. J Neuroinflammation 2021; 18: 1–14.

40.

Wendel

Bertolini

Kousoulos

, et al. Serum neurofilament light-chain levels in children with monophasic myelin oligodendrocyte glycoprotein-associated disease, multiple sclerosis, and other acquired demyelinating syndrome. Mult Scler J 2022; 28: 1553–1561.

41.

Simone

Palazzo

Mastrapasqua

, et al. Serum neurofilament light chain levels and myelin oligodendrocyte glycoprotein antibodies in pediatric acquired demyelinating syndromes. Front Neurol 2021; 12: 1–7.

42.

Horellou

Flet-Berliac

Leroy

, et al. Early blood neurofilament light chain and myelin oligodendrocyte glycoprotein antibody levels associate with different disease courses of myelin oligodendrocyte glycoprotein-associated disease in children. Brain Commun 2023; 5: fcad063.

43.

Andersen

Trewin

Dale

, et al. Biomarkers to predict relapse in myelin oligodendrocyte glycoprotein antibody-associated disease: a systematic review and meta-analysis. J Neurol Neurosurg Psychiatry 2026; 97(2): 109–119.

44.

Takano

Misu

Takahashi

, et al. Astrocytic damage is far more severe than demyelination in NMO: a clinical CSF biomarker study. Neurology 2010; 75: 208–216.

45.

Peng

Xia

, et al. Increased cerebrospinal fluid neurofilament light chain in central nervous system inflammatory demyelinating disease. Mult Scler Relat Disord 2019; 30: 123–128.

46.

Wang

Qiu

, et al. Cerebrospinal fluid light and heavy neurofilaments in neuromyelitis optica. Neurochem Int 2013; 63: 805–808.

47.

Watanabe

Nakamura

Michalak

, et al. Serum GFAP and neurofilament light as biomarkers of disease activity and disability in NMOSD. Neurology 2019; 93: E1299–E1311.

48.

Chatanaka

Avery

Pasic

, et al. The relationship between serum astroglial and neuronal markers and AQP4 and MOG autoantibodies. Clin Proteomics 2024; 21: 28.

49.

Liu

Zhao

Fan

, et al. High serum neurofilament levels among Chinese patients with aquaporin-4-IgG-seropositive neuromyelitis optica spectrum disorders. J Clin Neurosci 2021; 83: 108–111.

50.

Lee

Lim

Kim

, et al. Clinical implication of serum biomarkers and patient age in inflammatory demyelinating diseases. Ann Clin Transl Neurol 2020; 7: 992–1001.

51.

Sechi

Mariotto

McKeon

, et al. Serum neurofilament to magnetic resonance imaging lesion area ratio differentiates spinal cord infarction from acute myelitis. Stroke 2021; 52(2): 645–654.

52.

Kim

Gomes

ABAGR

Schindler

, et al. Blood-based biomarkers for identifying disease activity in AQP4-IgG-positive neuromyelitis optica spectrum disorder. JAMA Neurol 2025; 82: 168–175.

53.

Kim

Lee

Kim

, et al. Longitudinal follow-up of serum biomarkers in patients with neuromyelitis optica spectrum disorder. Mult Scler J 2022; 28(4): 512–521.

54.

Hyun

Kim

, et al. Investigating the presence of interattack astrocyte damage in neuromyelitis optica spectrum disorder: longitudinal analysis of serum glial fibrillary acidic protein. Neurol Neuroimmunol Neuroinflamm 2021; 8: e965.

55.

Wang

Cui

, et al. Therapeutic response and possible biomarkers in acute attacks of neuromyelitis optica spectrum disorders: a prospective observational study. Front Immunol 2021; 12: 720907.

56.

Zhang

T-X

Chen

J-S

, et al. Longitudinal treatment responsiveness on plasma neurofilament light chain and glial fibrillary acidic protein levels in neuromyelitis optica spectrum disorder. Ther Adv Neurol Disord 2021; 14: 1–13.

57.

Aktas

Smith

Rees

, et al. Serum glial fibrillary acidic protein: a neuromyelitis optica spectrum disorder biomarker. Ann Neurol 2021; 89: 895–910.

58.

Aktas

Hartung

Smith

, et al. Serum neurofilament light chain levels at attack predict post-attack disability worsening and are mitigated by inebilizumab: analysis of four potential biomarkers in neuromyelitis optica spectrum disorder. J Neurol Neurosurg Psychiatry 2023; 94: 757–768.

59.

Liu

Tan

, et al. Serum aquaporin 4-immunoglobulin G titer and neuromyelitis optica spectrum disorder activity and severity: a systematic review and meta-analysis. Front Neurol 2021; 12: 746959.

60.

Yalachkov

Schäfer

Jakob

, et al. Effect of estimated blood volume and body mass index on GFAP and NfL levels in the serum and CSF of patients with multiple sclerosis. Neurol Neuroimmunol Neuroinflamm 2023; 10: e200045.

61.

Schindler

Aktas

Ringelstein

, et al. Glial fibrillary acidic protein as a biomarker in neuromyelitis optica spectrum disorder: a current review. Expert Rev Clin Immunol 2023; 19: 71–91.

62.

Mariotto

Sechi

Ferrari

Serum neurofilament light chain studies in neurological disorders, hints for interpretation. J Neurol Sci 2020; 416: 116986.

63.

Villacieros-Álvarez

Sepulveda

Valls-Carbó

, et al. Cerebrospinal 14-3-3 protein levels as a neuroaxonal biomarker in aquaporin-4 antibody-positive neuromyelitis optica spectrum disorder. Neurol Neuroimmunol Neuroinflamm 2025; 12: e200432.

64.

Hyun

Kim

, et al. CSF GFAP levels in double seronegative neuromyelitis optica spectrum disorder: no evidence of astrocyte damage. J Neuroinflammation 2022; 19(1): 86.

65.

Johnsson

Eriksson

Rosenstein

, et al. The value of CSF diagnostic and prognostic biomarkers in NMOSD and MOGAD in real-life use. Mult Scler Relat Disord 2025; 94: 106302.

66.

Kleerekooper

Herbert

Kuiperij

, et al. CSF levels of glutamine synthetase and GFAP to explore astrocytic damage in seronegative NMOSD. J Neurol Neurosurg Psychiatry 2020; 91: 605–611.

67.

Carta

Dinoto

Capobianco

, et al. Serum biomarker profiles discriminate AQP4 seropositive and double seronegative neuromyelitis optica spectrum disorder. Neurol Neuroimmunol Neuroinflamm 2023; 11: e200188.

68.

Hebert

Riche

Rabilloud

, et al. Epidemiology of paraneoplastic neurologic disorders and autoimmune encephalitides in France. Neurol Neuroimmunol Neuroinflamm 2020; 7: e883.

69.

Satyadev

Tipton

Martens

, et al. Improving early recognition of treatment-responsive causes of rapidly progressive dementia: the STAM3P score. Ann Neurol 2024; 95: 237–248.

70.

Dinoto

Flanagan

EP.

Autoimmune dementia. Curr Opin Psychiatry 2025; 38: 101–111.

71.

Gilligan

McGuigan

McKeon

Autoimmune central nervous system disorders: antibody testing and its clinical utility. Clin Biochem 2024; 126: 110746.

72.

Papi

Milano

Spatola

Mechanisms of autoimmune encephalitis. Curr Opin Neurol 2024; 37: 305–315.

73.

Lancaster

Dalmau

Neuronal autoantigens-pathogenesis, associated disorders and antibody testing. Nat Rev Neurol 2012; 8: 380–390.

74.

Gresa-Arribas

Ariño

Martínez-Hernández

, et al. Antibodies to inhibitory synaptic proteins in neurological syndromes associated with glutamic acid decarboxylase autoimmunity. PLoS One 2015; 10: 1–14.

75.

Dalmau

Armangué

Planagumà

, et al. An update on anti-NMDA receptor encephalitis for neurologists and psychiatrists: mechanisms and models. Lancet Neurol 2019; 18: 1045–1057.

76.

Irani

Michell

Lang

, et al. Faciobrachial dystonic seizures precede Lgi1 antibody limbic encephalitis. Ann Neurol 2011; 69: 892–900.

77.

Thompson

Murchison

, et al. The importance of early immunotherapy in patients with faciobrachial dystonic seizures. Brain 2018; 141: 348–356.

78.

Sabater

Gaig

Gelpi

, et al. A novel non-rapid-eye movement and rapid-eye-movement parasomnia with sleep breathing disorder associated with antibodies to IgLON5: a case series, characterisation of the antigen, and post-mortem study. Lancet Neurol 2014; 13: 575–586.

79.

Abboud

Probasco

Irani

, et al. Autoimmune encephalitis: proposed recommendations for symptomatic and long-term management. J Neurol Neurosurg Psychiatry 2021; 92: 897–907.

80.

Abboud

Probasco

Irani

, et al. Autoimmune encephalitis: proposed best practice recommendations for diagnosis and acute management. J Neurol Neurosurg Psychiatry 2021; 92: 757–768.

81.

Dinoto

Ferrari

Mariotto

Treatment options in refractory autoimmune encephalitis. CNS Drugs 2022; 36: 919–931.

82.

Graus

Vogrig

Muñiz-Castrillo

, et al. Updated diagnostic criteria for paraneoplastic neurologic syndromes. Neurol Neuroimmunol Neuroinflamm 2021; 8: e1014.

83.

Graus

Dalmau

Paraneoplastic neurological syndromes in the era of immune-checkpoint inhibitors. Nat Rev Clin Oncol 2019; 16: 535–548.

84.

Farina

Birzu

Elsensohn

M-H

, et al. Neurological outcomes in immune checkpoint inhibitor-related neurotoxicity. Brain Commun 2023; 5(3): fcad169.

85.

Graus

Saiz

Dalmau

GAD antibodies in neurological disorders—insights and challenges. Nat Rev Neurol 2020; 16: 353–365.

86.

Budhram

Sechi

Flanagan

, et al. Clinical spectrum of high-titre GAD65 antibodies. J Neurol Neurosurg Psychiatry 2021; 92: 645–654.

87.

Guasp

Martín-Aguilar

Sabater

, et al. Neurofilament light chain levels in anti-NMDAR encephalitis and primary psychiatric psychosis. Neurology 2022; 98: E1489–E1498.

88.

Nissen

Ryding

Nilsson

, et al. CSF-neurofilament light chain levels in NMDAR and LGI1 encephalitis: a national cohort study. Front Immunol 2021; 12: 719432.

89.

Mariotto

Gajofatto

Zuliani

, et al. Serum and CSF neurofilament light chain levels in antibody-mediated encephalitis. J Neurol 2019; 266: 1643–1648.

90.

Brenner

Mariotto

Bastiaansen

AEM

, et al. Predictive value of serum neurofilament light chain levels in anti-NMDA receptor encephalitis. Neurology 2023; 100: E2204–E2213.

91.

Aboseif

Mangioris

Yang

, et al. Cytokine, chemokine, and neurofilament light chain signatures in LGI1 autoimmune encephalitis. Ann Clin Transl Neurol 2025; 12(11): 2258–2270.

92.

Businaro

Masciocchi

Barnabei

, et al. Serum autoantibody titers and neurofilament light chain levels in CASPR2/LGI1 encephalitis: a longitudinal study. Neurol Neuroimmunol Neuroinflamm 2025; 12: e200431.

93.

Grüter

Möllers

Tietz

, et al. Clinical, serological and genetic predictors of response to immunotherapy in anti-IgLON5 disease. Brain 2023; 146: 600–611.

94.

Eisenhut

Faber

Engels

, et al. Early neuroaxonal damage in neurologic disorders associated with GAD65 antibodies. Neurol Neuroimmunol Neuroinflamm 2023; 11: e200176.

95.

Kammeyer

Mizenko

Sillau

, et al. Evaluation of plasma neurofilament light chain levels as a biomarker of neuronal injury in the active and chronic phases of autoimmune neurologic disorders. Front Neurol 2022; 13: 689975.

96.

Segal

Mangioris

Lennon

, et al. CSF cytokine, chemokine and injury biomarker profile of glial fibrillary acidic protein (GFAP) autoimmunity. Ann Clin Transl Neurol 2025; 12: 855–860.

97.

Mazowiecki

Flet-Berliac

Roux

, et al. Long-term clinical and biological prognostic factors of anti-NMDA receptor encephalitis in children. Neurol Neuroimmunol Neuroinflamm 2025; 12: e200346.

98.

Chang

Day

Graff-Radford

, et al. Alzheimer’s disease cerebrospinal fluid biomarkers differentiate patients with Creutzfeldt–Jakob disease and autoimmune encephalitis. Eur J Neurol 2022; 29: 2905–2912.

99.

Bastiaansen

AEM

Van Steenhoven

De Bruijn

MAAM

, et al. Autoimmune encephalitis resembling dementia syndromes. Neurol Neuroimmunol Neuroinflamm 2021; 8: e1039.

100.

Lardeux

Fourier

Peter

, et al. Core cerebrospinal fluid biomarker profile in anti-LGI1 encephalitis. J Neurol 2022; 269: 377–388.

101.

Constantinescu

Krýsl

Bergquist

, et al. Cerebrospinal fluid markers of neuronal and glial cell damage to monitor disease activity and predict long-term outcome in patients with autoimmune encephalitis. Eur J Neurol 2016; 23: 796–806.

102.

Abu-Rumeileh

Erhart

Barba

, et al. CSF beta-synuclein, SNAP-25, and neurogranin in infectious and autoimmune inflammatory neurologic diseases. Neurol Neuroimmunol Neuroinflamm 2025; 12: e200491.

103.

Day

Yarbrough

Körtvelyessy

, et al. Prospective quantification of CSF biomarkers in antibody-mediated encephalitis. Neurology 2021; 96: E2546–E2557.

104.

Macher

Zrzavy

Höftberger

, et al. Longitudinal measurement of cerebrospinal fluid neurofilament light in anti-N-methyl-D-aspartate receptor encephalitis. Eur J Neurol 2021; 28: 1401–1405.

105.

Borko

Winters

Money

, et al. Long-term outcomes in leucine-rich glioma inactivated-1 autoimmune encephalitis and associated biomarkers of inflammation and neuronal and glial injury. Front Neurol 2025; 16: 1583892.

106.

Muñoz-Lopetegi

Guasp

Prades

, et al. Neurological, psychiatric, and sleep investigations after treatment of anti-leucine-rich glioma-inactivated protein 1 (LGI1) encephalitis in Spain: a prospective cohort study. Lancet Neurol 2024; 23: 256–266.

107.

Farina

Villagrán-García

Abichou-Klich

, et al. Serum neurofilament light chain correlates with clinical severity and predicts mortality in anti-IgLON5 disease. Neurol Neuroimmunol Neuroinflamm 2025; 12: e200498.

108.

Cortese

Carrasco

Tur

, et al. Differentiating multiple sclerosis from AQP4-neuromyelitis optica spectrum disorder and MOG-antibody disease with imaging. Neurology 2023; 100: E308–E323.

109.

Peng

, et al. Serum NfL associated with anti-NMDA receptor encephalitis. Neurol Sci 2022; 43: 3893–3899.

110.

Piepgras

Steffen

, et al. Glial fibrillary acid protein reflects disease activity in autoimmune encephalitis. Eur J Neurol 2025; 32: e70207.

111.

McKeon

Shelly

Zivelonghi

, et al. Neuronal intermediate filament IgGs in CSF: autoimmune axonopathy biomarkers. Ann Clin Transl Neurol 2021; 8: 425–439.

112.

Pranzatelli

Tate

McGee

, et al. CSF neurofilament light chain is elevated in OMS (decreasing with immunotherapy) and other pediatric neuroinflammatory disorders. J Neuroimmunol 2014; 266: 75–81.

113.

Vakrakou

Tzartos

Strataki

, et al. Neuronal and neuroaxonal damage cerebrospinal fluid biomarkers in autoimmune encephalitis associated or not with the presence of tumor. Biomedicines 2022; 10: 1262.

114.

Constantinescu

Krýsl

Andrén

, et al. Cerebrospinal fluid markers of neuronal and glial cell damage in patients with autoimmune neurologic syndromes with and without underlying malignancies. J Neuroimmunol 2017; 306: 25–30.

115.

Farina

Villagrán-García

Fourier

, et al. Diagnostic and prognostic biomarkers in immune checkpoint inhibitor-related encephalitis: a retrospective cohort study. Lancet Reg Health Eur 2024; 44: 101011.

116.

Schmitt

Müller

Tiedt

, et al. Increased serum NfL and GFAP levels indicate different subtypes of neurologic immune-related adverse events during treatment with immune checkpoint inhibitors. Int J Cancer 2025; 156: 1961–1971.

117.

Bjursten

Zhao

Al Remawi

, et al. Concentrations of S100B and neurofilament light chain in blood as biomarkers for checkpoint inhibitor-induced CNS inflammation. EBioMedicine 2024; 100: 104955.

118.

Farina

Villagrán-García

Benaiteau

, et al. Opsoclonus-ataxia syndrome in a patient with small-cell lung cancer treated with immune checkpoint inhibitors. Neurol Neuroimmunol Neuroinflamm 2024; 11: e200287.

119.

Bjursten

Pandita

Zhao

, et al. Early rise in brain damage markers and high ICOS expression in CD4+ and CD8+ T cells during checkpoint inhibitor-induced encephalomyelitis. J Immunother Cancer 2021; 9: e002732.

120.

Piepgras

Müller

Steffen

, et al. Neurofilament light chain levels reflect outcome in a patient with glutamic acid decarboxylase 65 antibody-positive autoimmune encephalitis under immune checkpoint inhibitor therapy. Eur J Neurol 2021; 28: 1086–1089.

121.

Gottiparthy

Lam

Kundu

, et al. Neurofilament light chain in serum of cancer patients with acute neurological complications. CNS Oncol 2024; 13: 2386233.

122.

Zahid

Tummala

Cerebral clues: serum neurofilament light chain (sNfL) as a novel biomarker for immune check point inhibitor (ICI) mediated seronegative encephalitis. Oxf Med Case Reports 2024; 2024: omae058.

123.

Mizenko

Bennett

Owens

, et al. A longitudinal, observational analysis of neuronal injury biomarkers in a case report of a patient with paraneoplastic anti-CRMP5 antibody-associated transverse myelitis. Front Neurol 2021; 12: 691509.