NEUROLOGY

Berger T and Stuve O

Neurology. 2019 Feb 8. pii: 10.1212/WNL.0000000000007022

DOI: 10.1212/WNL.0000000000007022

Multiple sclerosis (MS) is a chronic progressive disorder of the CNS that leads to neuro-degeneration and accumulation of neurologic disability. At disease onset, it is challenging to predict whether individual patients will have a benign or aggressive disease course, and whether individual patients will respond to disease-modifying  herapies. MS phenotypes1 and diagnostic criteria2 are defined on the basis of clinical and MRI characteristics. They are limited in their ability to guide treatment decisions.

In this issue of Neurology®, Kuhle et al.3 used a single molecule array (SIMOA) digital ELISA to evaluate blood neurofilament light chain (NfL) as a biomarker of recent, ongoing, and future disease activity and tissue damage in patients with MS on pharmacotherapies. Furthermore, these investigators tested the ability of their assay to monitor treatment response in patients with relapsing-remitting MS (RRMS). Blood samples were obtained from 589 patients with RRMS from 2 phase 3 studies that compared fingolimod to placebo4 or subcutaneous in- terferon-β-1a.5 Thirty-five healthy individuals were utilized as controls. Five major observations were made: (1) blood NfL levels at baseline were higher in patients with RRMS than in controls; (2) blood NfL levels at baseline correlated with T2-weighted lesion load and the number of gadolinium-enhancing T1-weighted lesions on brain MRI; (3) blood NfL levels at baseline, pharmacologic treatment, and the number of new or enlarging T2-weighted lesions predicted blood NfL levels at the end of the study; (4) high NfL levels at baseline were associated with increasing number of new or enlarging T2-weighted lesions, higher number of relapses, greater brain atrophy, and a higher risk of confirmed disability worsening; and (5) fingolimod reduced NfL blood levels at 6 months and over the duration of the studies.

These findings are encouraging and may suggest NfL as a feasible biomarker in patients with RRMS. Perera and Weinstein6 classified biomarkers based on their ability to detect pathogenic events during different disease stages: (1) induction (risk factors), (2) latency (screening and diagnosis), and (3) disease (prognosis). When simplified, this classification leaves scientists with 2 major classes of biomarkers: biomarkers of exposure, which inform on the pathogenesis of a disorder and predict de novo risk, and biomarkers of disease, which are useful in screening and monitoring disease progression.7

There is good biological plausibility for neurofilaments as biomarkers in MS. One of the earliest descriptions of neurofilaments was made by Palay and Palade8 in 1955. They are neuronal intracytoplasmic proteins that participate in the cytoskeleton and structural axonal support. In humans, neurofilaments are heteropolymers that are classified based on their molecular weights and anatomical distribution as NfL, NlM, NlH, internexin, and peripherin. Current thinking is that neuronal and axonal damage results in release of NfL from the CNS into the CSF and peripheral blood, where the proteins are detectable and quantifiable by sensitive detection assays, including the SIMOA immunoassay used in the present study. Studies by  these investigators demonstrated a positive correlation between NfL in these compartments and clinical and paraclinical MS disease activity.

According to the criteria described above, NfL is hardly a bio-marker of exposure in MS.7 Neurofilament release from theCNS and humoral immune responses toward them are likely to occur in any disorder that affects neurons and axons in the CNS and peripheral nervous system. In fact, CSF and blood NfL levels were substantially higher in patients with amyotrophic lateral sclerosis9 and Huntington disease than in controls.10 The true value of the study by Kuhle et al. over published reports on NfL as a disease marker in MS is that the data were obtained in the controlled setting of 2 clinical trials. This allowed the determination of treatment effects of fingolimod and subcutaneous interferon-β-1a. Also, the authors were able to demonstrate in this experimental setting that blood NfL levels are closely related to clinical disease and MRI disease burden in patients with RRMS. However, it is important to note that the value of NfL as a biomarker is so far based on observations made at the group level, therefore limiting its potential use as an outcome or prognostic factor in clinical trials. Further research is needed to understand whether NfL may predict or guide treatment  hoice in individual patients.

In summary, blood NfL appears to be a very promising bio- marker of disease7 that can easily be obtained serially to monitor disease activity and treatment response, once age- adjusted and comorbidity-adjusted normative values have been established and the analytical performance of the different neurofilament assays, which can influence data analysessubstantially, is fully investigated.