Neurofilament light chain (NfL) is an important marker of neuroaxonal injury. Today, high-sensitivity platforms allow reliable measurement of NfL in the blood, facilitating its use in clinical practice.
Damage to neuronal axons in the central nervous system (CNS) – resulting from ageing, trauma, or disease – can determine the release of neurofilaments into the cerebrospinal fluid (CSF) and blood [1.] In people with multiple sclerosis (MS), elevated levels of neurofilament light chain (NfL) reflect neuroaxonal damage occurring over the preceding three months [2.] Ultra-sensitive assays have enabled the simple and reliable quantification of NfL levels in blood, which facilitates its increasing application into routine clinical practice for monitoring disease activity and progression, and treatment response [3, 4.]
What are neurofilaments? We ask Prof. Paul Matthews of the Imperial College London: “Neurofilaments are a fundamental structural component of neurons, found mostly in axons, but also in synapses. There are different forms, including heavy, medium, and light chain. They share a common core structure and are not restricted to a single neuronal subtype. Neurofilaments are broadly expressed across neurons throughout both the CNS and the peripheral nervous system. They seem to be quite ubiquitous and are transported very slowly.”
The Perfect Biomarker
“NfL is almost the perfect biomarker,” Prof. Jan Lycke of the University of Gothenburg tells us. “It’s robust. We can measure it repeatedly in the blood, keep the samples at room temperature for several days, freeze and thaw them at least three times without affecting the levels. You can even shake the samples or send them by post. This makes it exceptionally useful for clinics outside university hospitals. Moreover, combining NfL with other biomarkers during the diagnostic workup allows for a more comprehensive evaluation. Baseline measurements at diagnosis enable regular follow-up – say every six months – to evaluate the efficacy of the treatment and monitor patient stability. If NfL levels rise, it can flag a new lesion, which you can then confirm with a magnetic resonance imaging (MRI).”
NfL and MRI. MRI offers essential insights into lesion burden, stage, and spatial distribution [2.] Although NfL cannot provide this information, blood-based NfL measurements offer several practical advantages over MRI, including greater accessibility and lower costs. Therefore, NfL can also be measured more frequently [2.]
NfL levels are increasingly used to monitor treatment response. They offer predictive value: elevated levels are associated with a more severe disease course [5, 6.] While high-efficacy disease modifying therapies (DMTs) substantially reduce these levels, residual elevations can indicate ongoing subclinical disease activity leading to progression [3, 5.]
As Prof. Lycke notes, “People with MS presenting with active, relapsing disease typically exhibit high baseline NfL levels, which normalise during treatment. This reduction may reflect minimal inflammatory activity. Consequently, combining NfL monitoring with the clinical absence of relapses, and a lack of new lesions on MRI provide a comprehensive evaluation of treatment efficacy. On the contrary, when patients are not doing as well as expected, NfL levels must be reassessed. If there is a suspicion of relapse, we measure NfL to check whether there is an increase. Occasionally, we find that some patients may have relapses or new MRI lesions without a rise in NfL. A possible reason is that certain lesions cause myelin loss without damaging the axons. However, in most cases, relapses or new MRI lesions are associated with an increase in NfL levels. And most importantly, across all our patients treated with immunomodulatory therapies, NfL levels decrease. With high-efficacy treatments the NfL levels may reach the same levels observed in the normal population.”
NfL concentrations can also be used to investigate the effect of extended interval dosing [7.] For example, Prof. Lycke and his team extended natalizumab dosing intervals from 4- to 6-week interval in people with relapsing remitting MS and measured NfL levels at every infusion of natalizumab over 12 months. The baseline mean concentration in this cohort remained unchanged for 12 months and individual concentrations did not change significantly following the interval extension [7.]
NfL Over Time
How fast do neurofilaments move within the neuron? Neurofilaments are synthesised and assembled in the neural soma and then, transported along axons. In vitro studies of growing axons show that neurofilaments are transported very fast, reaching speeds up to 2 micrometres per second [8.] However, the transport is interrupted by prolonged pauses. This results in an overall slow transport rate of 0.1-1 mm per day [8.] Therefore, the neurofilaments rank among the slowest proteins transported within the axon. This may explain why axons grow so slowly [8.]
How do they behave following neuroaxonal damage? Prof. Paul Matthews of Imperial College London tells us, ” When neurofilaments are released due to neuroaxonal damage, they diffuse through the extracellular space before entering the CSF and bloodstream. They are then cleared from the blood. This overall turnover lasts months. So, let’s say there is a stroke. A stroke occurs in a few seconds. That’s when the tissue starts to get damaged, followed by secondary inflammation over days. After a stroke, the damage stops at some point, but neurofilament release and clearance stretch over weeks to a couple of months, creating a temporal lag that cannot yet be defined precisely.”
Professor Tomas Olsson of the Karolinska Institute adds, “This is why measurements must be taken at baseline, during the workup, and then repeated multiple times over time, with complementary NfL measurements every 3 to 6 months thereafter. By tracking repeated measurements in individual patients, clinicians can follow the development of the disease. And also, we need to remember that NfL is not specific to MS. Elevated levels can occur across various clinical scenarios, including traumatic brain injuries or comorbidities such as diabetes mellitus.”
Ageing and BMI. NfL levels are influenced by some confounders, such as age and body mass index (BMI) in healthy people. Concentrations increase by 2.2% per year between the ages of 18 and 70 [9.] To account for these variables, an international normative database makes it possible to calculate z-scores adjusted for age and BMI [10.]
Nfl in the Clinic. Prof. Matthews adds an important element to the discussion: “There is a very interesting aspect to consider when we talk about assays that look at neurofilament light peptide. We believe that even high-sensitive, automated clinical platform primarily measure peptide fragments, rather than the full-length protein, which is about 70 kilodaltons. There is even the possibility that the precise mix of these fragments can vary across different diseases. This is a fascinating hypothesis to explore.”
While researchers continue to investigate the various applications of NfL, the biomarker’s clinical utility remains clear. Prof. Lycke concludes, “I think that NfL should be included in the diagnostic workup, as well as the evaluation of disease activity and treatment efficacy. It is a reliable marker to see if patients are on the right therapy. It should certainly be part of routine MS care.“
Written by Stefania de Vito
Special thanks to Prof. Jan Lycke (University of Gothenburg), to Prof. Paul Matthews (Imperial College London), and to Prof. Tomas Olsson (Karolinska Institute) for their insights.
References
[1] Bittner S et al Brain 2021; 144.10: 2954-2963.
[2] Freedman MS et al EBioMedicine 2024; 101.
[3] Khalil M et al Nature Reviews Neurology 2024; 20.5: 269-287.
[4] Novakova L et al Neurology 2017; 89.22: 2230-2237.
[5] Filippi M et al Journal of Neurology 2025; 72.9: 565.
[6] Bittner S et al EBioMedicine 2020; 56.
[7] Johnsson M et al Multiple Sclerosis Journal 2022; 28.13: 2070-2080.
[8] Gafson AR et al Brain 2020; 143.7: 1975-1998.
[9] Mattsson N et al JAMA Neurology 2017; 74.5: 557-566.
[10] Benkert P et al The Lancet Neurology 2022; 21.3: 246-257.