Predicting a relapse


The longer I work in MS the more I realize how unpredictable it is. There are complications in diagnosis, in assessing treatment efficacy, in prognosticating, and in achieving homogeneity in clinical practice. The complications are complicated.

But what if you have a test done, albeit quite randomly and it tells you something bad is going to happen in the near future – would you ignore it?

Probably not.

This is exactly what the Edwards group in New York found. As part of another study to determine the drug activity of dimethyl fumarate (or Tecfidera), 16 PwMS had lumbar punctures performed.

However, one person in the group demonstrated a rise in their neurofilament levels (PwMS No 4 in figure below). This person, subsequently went onto experience a disabling MS relapse.

NfL levels of CSF at baseline, 6-weeks, 28-weeks in the 16 patients with
case report showing spike of NfL at 6-weeks.

There are two key findings to consider with this person: 1) the rise in neurofilament preceded the relapse, and 2) the MRI didn’t demonstrate any changes.

And, then there is the niggling point, which is, once you’re faced with such a test result, can you afford to sit and wait?

Mult Scler Relat Disord. 2019 Mar 23;31:59-61. doi: 10.1016/j.msard.2019.03.016. [Epub ahead of print]

Neurofilament light chain as an indicator of exacerbation prior to clinical symptoms in multiple sclerosis.
Edwards KR, Garten L, Button J, O’Connor J, Kamath V, Frazier C.

Background: Biomarkers may be a sensitive measure of disease activity in patients with multiple sclerosis (pwMS).

Objective: A pwMS had a marked increase of neurofilament light chain (NfL) in CSF 9-weeks prior to a clinical exacerbation.

Methods and Results: Brain MRI, CSF, EDSS were measured at baseline, 6 weeks and 28 weeks. The patient had an exacerbation at week 15 of study but the NfL measured at week 6 were found to show a nearly 3-fold increase of CSF NfL levels prior to symptoms when the NfL levels were later measured.

Conclusion: This is an example supporting the usefulness of NfL in monitoring disease activity in pwMS which may predict disease activity prior to a clinical exacerbation.

About the author

Neuro Doc Gnanapavan


  • So, you are suggesting a 6 or 12 monthly LP program to proactively switch drugs. I like it. Non invasive, cost-effective and highly practical.
    You may want to start stocking up on epidural blood patches before the 31st of October.

    • I would have said that if that was what I meant! This article is pointing out that those with very high levels of NFL are at risk of relapse sooner rather than later. I have witnessed this before and am now cautious with my monitoring of these individuals and treating the relapse sooner, even if the MRI is normal.

      • How do you treat a relapse, or prevent it if you suspect one is about to happen? Are steroids any value at that point?

        • This is tricky, firstly we don’t know if steroids can prevent a relapse. However, since we are using NFL in our practice those with very high levels can be recruited into a prospective analysis to see if IV MP brings down NFL levels and prevents relapses, clinical disability versus no steroids. This is just an idea.

  • Nice post

    1º Question
    What is a relapse (pathophysiology)?

    Nfl levels (lumbar puncture)
    should replace mri monitoring in the future?


  • I think what is being suggested is that Nfl can be measured in blood from a blood test rather than from a lumbar puncture.

    • Doesn’t look like it. But the clinical symptoms are consistent with a brain relapse than one that can be localised to the spinal cord.

  • Boy does it take long to have Nfl test in clinical practice (and its not even for MS only). Nice post, thanks!

  • Still seems hit or miss (even with serum NFL). But it would be extremely helpful to have this tool.
    Can an MRI predict a relapse? If so, how long can a lesion show active before a visible relapse?

    • Many have looked at this and the answer is that you’re more likely to have new lesions that are asymptomatic.

      David Miller did a small study with serial GAD enhancing scans and demonstrated that more likely than not there is GAD enhancement during an acute relapse – this seems to last roughly 6weeks, and it can be on average more than just a single lesion enhancing. It provides a structural perspective.

  • Just found this

    Longitudinal analysis of serum neurofilament
    light chain: A potential therapeutic monitoring
    biomarker for multiple sclerosis

    (no lumbar puncture)

    The sNfL levels in MS patients with or without
    disease activity and in healthy controls
    After initiation of alemtuzumab therapy, the sNfL
    levels from 83 samples of 11 patients with EDA
    (26.7 ± 33.7 pg/mL) were significantly higher than
    those from 44 samples of 6 patients with NEDA
    (10.4 ± 4.7 pg/mL, p < 0.001) as well as those from
    35 healthy controls (5.9 ± 2.2 pg/mL, p < 0.001)
    (Figure 1(a)). In longitudinal analysis, the sNfL levels
    were stably low in patients with NEDA, while
    they were more variable in patients with EDA
    (Figure 1(b)).

    35% neda at 21 months

    6 out 17 patients

    Some patients exhibited elevation of sNfL levels as
    well as MRI activity preceding clinical relapses, and
    the period in which these features were displayed
    ranged between 3 weeks to 3 months. Retrospective
    nature of this study and using available preserved
    serum samples resulted in measurement of sNfL levels
    at irregular time points. In addition, all enrolled
    patients were Korean; therefore, possible racial differences
    in the role of sNfL could not be investigated.
    Future prospective studies with longitudinal analysis
    of sNfL levels at regular time points including diverse
    racial groups are needed to investigate appropriate
    time interval for therapeutic monitoring and potential
    population-specific roles of sNfL. In conclusion, our
    results support that sNfL can be an additional therapeutic
    monitoring biomarker reflecting disease activity
    for individualized therapeutic approaches in
    patients with MS

    More or less in line with the study above

    Interesting is the fact that the in some non responders sNfl rose dispite alemtuzumab infusions

    (eligible for hsct)

    DOI: 10.1177/


    • “Interesting is the fact that the in some non responders sNfl rose dispite alemtuzumab infusions”

      You can develop antibodies for all MABs and thus render the drug less effective or even useless. Maybe this was the case here with these patients. So NFL checks could actually show if the drug is even working or not.

      • Nabs (or neutralising antibodies) are a completely different ball game and should be tested for separately when there is a query about treatment failure.

    • There have been similar studies. It’s the cut offs that concern me. For instance in the above study they report the average level (26.7 ± 33.7 pg/mL). Some of the subjects had levels in the negative integer if the above was to be believed. But these individuals still have nfl levels that overlap with those that have NEDA and healthy controls!



Recent Posts

Recent Comments