Blood and Cerebrospinal Fluid Biomarkers in Amyotrophic Lateral Sclerosis: Current Landscape and Potential Utility in Clinical Practice

Amyotrophic lateral sclerosis (ALS) is a paralyzing neurodegenerative disease characterized by progressive motor neuron loss. With varying sites of onset, the disease results in progressive weakness of limb, bulbar, and respiratory muscles, leading to death.¹ ALS has a global incidence of 1 to 3 per 100,000 people/year and a mean survival time to death or invasive ventilation from 2 to 4 years.² Most cases are sporadic but ~10% are familial. A sequence variation in one of multiple genes, such as C9ORF72 or SOD1, is typically identifiable.¹ Many drugs have been trialed over the decades, with few successes and achievement of only slightly slowed disease progression in sporadic ALS.

Multiple pathophysiologic mechanisms are implicated in ALS, including glutamate excitotoxicity, neuroinflammation, oxidative stress, and mitochondrial dysfunction.³ The hallmark pathologic finding is cytoplasmic accumulation and nuclear depletion of transactive response DNA-binding protein 43 (TDP-43), primarily in the upper and lower motor neurons, which has been found in the majority of sporadic and familial ALS autopsy cases.^4,5

The lack of a clear understanding of ALS pathophysiology along with clinical heterogeneity and lack of reliable pharmacodynamic biomarkers are the main factors that have hindered the development of effective therapeutics. To better stratify individuals with ALS and more effectively monitor treatment and disease progression, there has been a renewed emphasis on the development of novel biomarkers for ALS.³ The BEST (Biomarkers, Endpoints, and Other Tools) Resource by the Food and Drug Administration (FDA)–National Institutes of Health Biomarker Working Group defines a biomarker as “a biologic molecule found in blood, other body fluids, or tissues that is measured as an indicator of normal biologic processes, pathogenic processes, or biologic responses to an exposure or intervention, including therapeutic interventions.”⁶ Categories of biomarkers include susceptibility, diagnostic, monitoring, and prognostic.⁶ In ALS clinical trials, blood and cerebrospinal fluid (CSF) neurofilament levels have served as surrogate pharmacologic biomarkers of efficacy⁷ and supported regulatory drug approvals by the FDA and Health Canada. In this review, we highlight the most relevant blood and CSF-based biomarkers in ALS and discuss their potential utility in clinical practice.

Neurofilaments

Neurofilaments are proteins or intermediate filaments exclusive to neurons that form the structure of axons. The 3 most relevant subunits are neurofilament light chain (NfL), neurofilament medium chain, and neurofilament heavy chain.⁸ As a marker of neuronal loss, higher concentrations of neurofilaments have been identified in the CSF of individuals with different neurologic diseases, including neurodegenerative, inflammatory, and traumatic disorders.^9-11 The most recent assays have demonstrated that serum NfL concentration is strongly correlated to CSF levels in people with ALS.⁸ Phosphorylated neurofilament heavy chain is also occasionally measured in studies, but there has been a growing preference to use blood NfL measurements due to their greater stability, stronger correlation with CSF NfL levels, and enhanced detectability at lower serum concentrations.¹² A better understanding of the natural history of neurofilament levels throughout the course of disease progression has improved its utility as a prognostic and treatment response biomarker in ALS.⁸

Presymptomatic individuals with known ALS-causing sequence variations experienced an increase in serum NfL levels months before symptom onset. In people with a highly penetrant and rapidly progressive SOD1 sequence variation, NfL levels increased 6 to 12 months before symptom onset. Serum NfL was found to demonstrate high sensitivity and specificity in predicting onset of ALS.⁸ These findings may not be generalizable to all forms of ALS but highlight the potential of NfL as a susceptibility and risk biomarker to predict the likelihood and timing of disease onset in presymptomatic carriers of monogenetic ALS-causing sequence variations.⁸

Neurofilament levels continue to rise for ~1 year after ALS symptom onset before reaching a plateau that persists throughout the disease course.⁸ Although elevated neurofilament levels are not specific to ALS, they are typically higher than levels seen in other neuromuscular diseases in the differential diagnosis.^13,14 Certain conditions with substantial lower motor neuron or peripheral nerve involvement (eg, multifocal motor neuropathy, spinobulbar muscular atrophy) can also present with increased serum NfL levels, although generally at lower concentrations than in ALS. In contrast, neuromuscular junction diseases such as myasthenia gravis or myopathies tend to have lower serum NfL levels than seen in those with ALS.^13,14

NfL has been investigated as a putative diagnostic biomarker and a recent study demonstrated strong discriminatory capability (area under the curve of .91 in a receiver operating characteristic plot) in distinguishing people with ALS from both healthy and disease controls.¹⁵ However, because most individuals with ALS are assessed only after clear signs and symptoms of the disease have emerged, its diagnostic utility for neurologists is limited. Serum NfL levels may be most beneficial as diagnostic screening tools for people in the earliest disease stages when the diagnosis remains uncertain.⁸

The most promising clinical use of NfL in ALS is as a prognostic biomarker. Higher NfL levels at the plateau stage have been associated with faster disease progression and worse prognosis. Whereas NfL levels are not highly correlated with the extent of clinical deficit or disability, they are more strongly associated with the rate of neuronal loss and disease progression.⁸ Differentiating fast- from slow-progressing ALS is challenging in the earliest disease stages but is imperative in the assessment of prognosis and for stratification in clinical trials. Neurofilament measurement in serum or CSF is not widely available in clinical practice due to the lack of standardization of assay methodologies and cutoff values.

Neurofilament levels have been included as an outcome measure in research settings. Clinical trials have used neurofilaments for monitoring of pharmacodynamic biomarkers, confirming target engagement, and assisting in dose selection.⁶ The VALOR clinical trial (NCT02623699) demonstrated that tofersen (Qalsody; Biogen, Cambridge, MA), an intrathecal antisense oligonucleotide, induced a reduction in serum NfL concentrations compared with placebo in individuals with ALS and SOD1 sequence variations.⁷ Reduced superoxide dismutase 1 (SOD1) protein concentrations in the CSF served as a target engagement measure after intrathecal tofersen administration.⁷ Whereas VALOR failed to achieve its primary clinical efficacy outcome, lower levels of both CSF SOD1 protein and serum NfL were sufficient to support unprecedented FDA and Health Canada approval, underscoring the importance of these biomarkers in the current landscape of ALS clinical trials.¹⁶

A limitation associated with the use of neurofilaments as biomarkers is the relatively long period required to detect a meaningful change in NfL levels given its half-life, with most clinical trials requiring at least 6 months of follow-up to observe significant NfL reduction.⁸

The Figure summarizes the behavior of neurofilaments in the natural history of ALS and its use as a biomarker.

Figure. Representation of neurofilament light chain (NfL) concentration in a person with faster amyotrophic lateral sclerosis (ALS) progression (orange line) and one with slower ALS progression (blue line). NfL levels start to increase in the presymptomatic stage of the disease and then reach a plateau after ~1 year of symptoms. Higher NfL levels indicate faster rate of neurodegeneration or disease progression rather than higher degree of accumulated deficits. To illustrate the potential of NfL as a monitoring biomarker, the red dashed line indicates initiation of a new hypothetical therapy that successfully targeted neurodegeneration and slowed disease progression. NfL levels decreased after ~6 months of therapy initiation. The individual with slower ALS progression was not treated.

Inflammatory Biomarkers

Growing evidence suggests that immune system dysregulation is a key component of ALS pathophysiology, which has generated interest in inflammatory biomarkers for this disease. Numerous studies have examined C-reactive protein, cytokines, and inflammatory cell subpopulations in ALS. Whereas reports on CSF and serum cytokine levels have been inconsistent, a random-effects meta-analysis found elevated pooled levels of tumor necrosis factor–α, interleukin (IL)-6, IL-1, and IL-8 in individuals with ALS.¹⁷ Although studies have demonstrated an association between worse disease prognosis and higher C-reactive protein and IL-6 levels, these associations have not been consistently replicated which limits their utility as reliable biomarkers for research or clinical use.¹⁸

Studies assessing the potential role of chitinases (ie, a family of enzymes that degrade chitin and play a role in the immune system in humans) as a diagnostic biomarker in ALS have demonstrated more consistent results. For example, a study by Vu et al¹⁹ demonstrated an association between ALS disease progression and elevated levels of chitotriosidase (Chit-1) and chitinase-3-like protein 1 (CHI3L1). Another study suggested that combining Chit-1 or CHI3L1 (sometimes referred to as YKL-40) with NfL may prove a better predictor of survival than any of these markers separately.²⁰

A higher blood count of regulatory T cells (Tregs)—a subpopulation of T cells with immunomodulatory functions—has been associated with better prognosis in ALS.²¹ Studies with therapies targeting T cells have used Tregs as a monitoring biomarker and primary outcome measure.²²

Glial Fibrillary Acidic Protein

Glial fibrillary acidic protein (GFAP)—a cytoskeletal protein expressed in astrocytes—serves as a marker of astrocyte activation or damage.²³ Studies have reported elevated GFAP levels in people with ALS compared with healthy controls.²³ However, GFAP alone has shown limited discriminative performance in distinguishing people with ALS from healthy or disease controls.^15,23 Notably, one study found a significant correlation between higher GFAP levels and greater cognitive impairment in people with ALS.²³

Markers of Muscle Injury and Metabolites

Various routinely measured metabolites and muscle injury markers have been studied in ALS. Serum creatinine, a metabolite derived from muscle activity that reflects both muscle mass and kidney function, was reduced in people with ALS compared to healthy controls. Higher creatinine levels have been associated with lower risk of all-cause mortality in people with ALS according to a meta-analysis.²⁴ Higher baseline albumin levels have been associated with better survival in one study, whereas other studies suggested that their longitudinal decrement may be a better predictor of survival.²⁵ These results likely reflect the influence of muscle mass and nutritional status on survival. Conversely, levels of creatine kinase, a muscle enzyme indicative of muscle damage, is often elevated in individuals with ALS due to lower motor neuron damage and denervation atrophy. However, a recent meta-analysis found no independent association between creatine kinase levels and survival.²⁶

Abnormal iron metabolism has been implicated in neurodegenerative processes through oxidative stress–related mechanisms and the phenomenon of ferroptosis.²⁵ Some studies have found increased ferritin and decreased transferrin levels in people with ALS vs controls. Other studies have highlighted the potential of ferritin as a prognostic biomarker, but results have not been replicated consistently.²⁵

Biomarkers Specific to Sequence Variations Causing ALS

Proteins related to ALS pathology have been investigated as more specific biomarkers for the disease. These include proteins related to each familial form of the disease (eg, SOD1, C9ORF72, FUS) and TDP-43. The utility of SOD1 protein levels was discussed previously. In people with intronic hexanucleotide repeat expansions in the C9ORF72 gene, there is both a loss of normal chromosome 9 open reading frame 72 (C9ORF72) protein function and the production of 5 dipeptide repeat proteins, such as poly(GP). Similar to SOD1, these proteins have been studied as target engagement biomarkers for genetic therapies for the C9ORF72 familial form of the disease.²⁷

Efforts have been made to identify a reliable biomarker for TDP-43 proteinopathy which could have broader use, including in people with sporadic ALS. However, attempts to measure abnormal TDP-43 protein levels in the blood or CSF have been largely unsuccessful because of its low concentrations and the difficulty of distinguishing pathologic from normal TDP-43.²⁸ As such, researchers attempted to measure TDP-43 pathology indirectly. When TDP-43 is functionally absent from the nucleus, a cryptic exon is included in the messenger RNA of the UNC13 homologue A (UNC13A) protein, with loss of the functional protein. Depletion of the UNC13A protein impairs neurotransmission and appears to contribute to ALS pathophysiology. UNC13A is, therefore, a potential target for new therapies, and its associated abnormal RNA or protein (cryptic peptide) could serve as a target engagement or diagnostic biomarker. TDP-43 abnormalities have also been linked to the incorporation of cryptic exons in other genes, forming other aberrant proteins with similar biomarker potential as UNC13A.^29,30

The extent to which reductions in these ALS-related proteins correlate with prolonged survival and other clinical benefits remains uncertain. As a result, they are often paired with NfL, which has more established clinical correlations. These monitoring and target engagement biomarkers have reshaped the ALS drug development landscape because exclusive reliance on clinical measures (eg, the Amyotrophic Lateral Sclerosis Functional Rating Scale–Revised) has proven limiting over the years.

The most important biomarkers discussed in this review are summarized in the Table. 

Conclusion

Biomarkers in ALS are increasingly shaping the development of new treatments as demonstrated by a recent drug approval granted solely on biomarker-based outcomes. Therefore, the clinician must understand this evolving landscape. Among the biomarkers discussed in this review, neurofilaments are the most promising for clinical translation given their potential as prognostic and diagnostic tools. Other biomarkers that could aid clinicians in the treatment and monitoring of individuals with ALS are likely to emerge in the coming years.