Relationship Between Serum Neurofilament Light and Multiple Sclerosis Disability Progression

Neurofilament light (NfL) is a scaffolding cytoskeletal protein that is expressed exclusively in central and peripheral neurons and is released when there is damage to the neuron, including its axon. NfL can be measured in the CSF and serum, with modest correlations between the two.^1,2 This makes serum NfL (sNfL) an attractive potential biomarker for a variety of neurodegenerative and neuroinflammatory conditions, including multiple sclerosis (MS). Potential applications to neurologic conditions include a prognostic biomarker to predict disease progression and a pharmacodynamic/treatment response biomarker to be used in clinical trials and possibly clinical practice.³ A number of studies have looked into these potential applications.^4,-,8

Because sNfL directly reflects neuronal injury, it is not surprising that sNfL levels increase at the time of MS inflammatory activity, such as when there is a clinical relapse or an enhancing lesion on MRI.^1,3 Inflammatory injury predisposes to disability progression, so it is no surprise that elevation in sNfL associated with active inflammation predicts future disability.^4,5 It is less clear how sNfL levels relate to disability progression that develops independent of inflammatory activity. Could sNfL be used to predict the clinical course in a patient with progressive MS? Could changes in sNfL be used to identify potential therapies for progressive disease? A large proportion of the progression of disability occurs independent of relapse activity. It would be useful to have a biomarker that predicts and correlates with this progression.⁹

In this issue of Neurology®, Bridel et al.¹⁰ look at the value of sNfL as such a biomarker using a longitudinal cohort of natalizumab-treated patients. Because patients were on the highly effective therapy natalizumab, there was very little inflammatory activity, and thus active inflammation was not a potential confounder. The authors compared sNfL levels in patients with and without disability progression and, surprisingly, found no association between sNfL levels at baseline or longitudinally with disability progression over time. Levels of sNfL were neither predictive of future disability nor did they track with disability over time.

The results of this study contrast with those from the phase 3 natalizumab trial (ASCEND), where sNfL levels at follow-up were significantly higher in those with progression compared to those without.⁶ Similarly, results from phase 3 trials of fingolimod (INFORMS) and siponimod (EXPAND) found that baseline sNfL levels were associated with disability progression.⁷ All of these studies were different from that reported by Bridel et al. because they included placebo patients who had ongoing inflammation. A large natural history cohort study published last year showed that elevated sNfL levels were associated with disability progression,⁸ but that study, too, included patients with active inflammation. The results by Bridel et al.¹⁰ suggest that when active inflammation is removed from the picture, sNfL may not relate so closely with the pathologic processes that drive the gradual, insidious disability progression reported by our patients with progressive MS. This aligns with a recent clinical trial report of ibudilast, which is not an anti-inflammatory therapy. Despite ibudilast treatment benefiting MRI measures of tissue injury in progressive MS, sNfL was not reduced by treatment with ibudilast.²

The story of sNfL in MS parallels the complexities of the disease: inflammatory disease activity is easily identifiable, has a well-established pathophysiology, and is clearly associated with temporal increases in sNfL, while progressive disease is more elusive in its pathophysiology, quantification, and relationship to biomarkers such as sNfL. This story also tracks with the dichotomous history of disease-modifying therapy in MS, with massive advances in relapsing MS therapies but little success in progressive MS. The authors aptly point out that progression contributes significantly to long-term disability, yet treatments to slow disability progression are scarce.

There is no doubt that to make strides in drug development, more research is needed to understand the complex pathophysiology of neurodegeneration that leads to disability progression in progressive MS. sNfL appears well-suited to predict inflammatory activity in MS and treatment response to anti-inflammatory disease-modifying therapy. Its role in relation to disability progression in the absence of inflammation remains less clear, and we need more analyses of natural history and clinical trial datasets, particularly those focusing exclusively on patients treated with anti-inflammatory therapies.

• © 2021 American Academy of Neurology