Guestpost: Neurofilaments vs. ‘The straight-line instinct’!



I am new to the blog and Barts-MS. I am Dr. Ide Smets and I am an ECTRIMS clinical fellow working in the Barts-MS team. My country of origin is Belgium which is famous for having six governments, for having a national dish that is claimed by another country (French Fries) and for going wild about cyclocross (bring it on Chris Froome ;-)). In 2018, I obtained my PhD in biomedical sciences at the Catholic University of Leuven on the topic of multiple sclerosis. Me and my research lab were able to demonstrate how genetic risk factors for multiple sclerosis shape an individual’s immune system, and in particular the notorious B cell pool. In 2019, I completed my training as a general neurologist in University Hospitals Leuven and I decided – after a journey on the Trans-Siberian express – to head off to London to gain additional hands-on experience in one of the top-notch multiple sclerosis centres in the world. 

Blog post

If you are regular readers of this blog you will have heard about neurofilament light chains already. 

Some people consider them one of the most promising biomarkers in multiple sclerosis and other neurological diseases. Neurofilaments are important for the three-dimensional structure of neurons and, when neurons die, they are released into the cerebrospinal fluid and blood. By measuring neurofilament levels in the blood we can estimate the damage to somebody’s brain and spinal cord as a result of MS and how fast their brain function is declining as a result of – unfortunately – time (older age = higher neurofilament levels). This would be of great benefit to many people with MS as neurofilaments would not only reflect the visible inflammation (i.e. the bright white spots on your brain MRI) but also the damage accumulating because of (currently) invisible inflammation in your entire brain and spinal cord. 

However, prof. Chris Whitty’s famous words are appropriate here: “no test is better than a bad test”. The study of Khalil et al. has namely investigated the levels of blood neurofilament light chain in a large group of healthy individuals. They have shown that neurofilaments not only increase with increasing age but they also go firmly against our ‘straight-line instinct’. This means that we see an acceleration of brain damage in people after 60 years old, with the levels of neurofilaments increasing every year four times as fast as when you are below 60 years old. Khalil and his team have so far only looked at healthy individuals but as cognitive symptoms and brain atrophy are hallmarks of multiple sclerosis it is not inconceivable that this non-linear increase in blood neurofilaments may start even more early in people with multiple sclerosis. 

This has huge implications on how to interpret blood neurofilament light levels in people with MS. Would we be satisfied with a one-time measurement of neurofilaments, similar to other biomarkers you might be familiar with such as oligoclonal bands in the cerebrospinal fluid? Or do we need serial measurements over time, similar to for example CRP which the GP will measure every time you present with an infection? 

Personally, I think based on this article it has become unlikely that one-time measurements will do the job unless we are satisfied with a bad test of course. The non-linear increase at older ages makes it treacherous to interpret a single time point measurement, and there will always remain reasonable doubt about whether an individual’s value is within the normal or abnormal range. Too many shades of grey! What would be more informative is to track the neurofilament levels as they evolve over time, similar as to what is done with the height and weight of young babies. If the increase of an individual’s neurofilament level is increasing at a rate that we would expect for a given age, we can all sleep on both ears! 


Serum neurofilament light levels in normal ageing and their association with morphologic brain changes

Michael Khalil, Lukas Pirpamer, Edith Hofer, Margarete M. Voortman, Christian Barro, David Leppert, Pascal Benkert, Stefan Ropele, Christian Enzinger, Franz Fazekas, Reinhold Schmidt & Jens Kuhle 

Nature Communications volume 11, Article number: 812 (2020) 

Neurofilament light (NfL) protein is a marker of neuro-axonal damage and can be measured not only in cerebrospinal fluid but also in serum, which allows for repeated assessments. There is still limited knowledge regarding the association of serum NfL (sNfL) with age and subclinical morphologic brain changes and their dynamics in the normal population. We measured sNfL by a single molecule array (Simoa) assay in 335 individuals participating in a population-based cohort study and after a mean follow-up time of 5.9 years (n = 103). Detailed clinical examination, cognitive testing and 3T brain MRI were performed to assess subclinical brain damage. We show that rising and more variable sNfL in individuals >60 years indicate an acceleration of neuronal injury at higher age, which may be driven by subclinical comorbid pathologies. This is supported by a close association of sNfL with brain volume changes in a cross-sectional and especially longitudinal manner.

About the author

The MS Bloggger


  • I would like to measure the value every six months plus whenever I change my treatements. DMT but also other things like food, sports and nutritions like vitaminD.

    With such a timeline I could understand better which treatements had a positive effect and which not. So the absolut values are not so interesting for me as a patient, but the change over time would give me valuable hint about how it goes

  • Ide,

    Many thanks for your informative post.

    I’m a big fan of Belgium for many reasons (Bruges, waffles…). But “Pump up the Jam” by Technotronic is your best export and brings back many good memories.

      • MD2,
        New Romantic was my thing (followed by The Smiths). While I was attending 90s rave you were attending the eisteddfod meets [Ide – an eisteddfod is an event which Welsh druids attend and listen to boring poems with lots of gg and ff sounds].

        • Thought you were more of a Vera Lynn aficionado. New Romantic,ugh. Ah, the Smiths, how I loved them, way abck when, such a shame how Brexiter Morrisey turned out. 90’s raves= off your face on E (only way to make the music acceptable) in a muddy field, but each to their own.
          Not cool to diss other people’s culture, Sid, Welsh is a beautiful language, as is the poetry, and our national anthem beats yours hands down 😉
          Diversity and unconconscious/conscious bias training for you.
          Pob hwyl!

    • I was not aware this was a Belgian export product :-), should definitely include that next time!

  • As a 48 year old with progressive MS with an edss of 7.5 would the neurofilament level be expected to be on the high side?

    • It is possible, but much less likely than in relapsing patients that still manifest with relapses and MRI activity. The problem will however be to determine what is elevated. Only in case of very high values we could be really certain.

      • So if it is very low am I right in saying it could mean a number of things……

        1 the patient has plateaued so could remain in their current state

        2 their most recent treatment has worked

        3 there Ms deterioration is no longer easy to find i.e. is smouldering MS although I thought prof g had said neurofilament levels could determine smouldering Ms or am I wrong?

        • 1. Yes maybe
          2. Yes maybe
          3. Yes probably….I think when we first started it was viewed as a marker for damage, but most of this is inflammation induced damage and if you get rid of that the neurofilament is much harder to find if you are loking for neurofilament light.

          We saw this in the animals a few years ago…during the relapsing stage it was easy to find in the blood in the secondary progressive phase it was hard to find

          • If a low neurofilament level is likely explained by 3 do you believe any drugs will help smouldering?

            Is this where the chariot trial comes in?

            Without any tangible markers how can you monitor improvement?

  • Is it known what happens to neurofilament particles over time once they enter the blood stream? Are they forever there or do they break down further over time and get adsorbed or excreted by the body?

  • Hello. In England we also had a dish stolen from us but by the French. We used to eat frogs legs, long before the French.

  • It’s great to see somebody working in Neurology who’s obviously as enthusiastic and able as you are. I’m not entirely sure how this is going to prevent accumulation of disability for me. Scenario: pwMS – I can’t move my right leg, so I think I’m having a relapse. Doctor – your sNfLs are elevated so you’re having a relapse. pwMS – yes I just told you that…

    • Hi Mark! Very good remark. I think you are right, we do not need NfL to diagnose relapses, as a relapse diagnosis is essentially a clinical judgement. However, the majority of the new lesions that appear in the brain are not associated with a clinical manifestation, and this is where the biggest window of opportunity for NfL will be: detection of ongoing inflammation in pwMS that do not have overt relapses.

  • Thank you for the interesting post. It would be of benefit to monitor MS progression with additional sensitive and less invasive techniques than conventional MRI. New biomarkers are needed for this purpose. OCT was all the rage it seems only a few years ago but has never really gained a foothold in clinical practice. Below is a paper comparing serum neurofilament light chain levels and OCT in MS patients. Thoughts on utilizing these for progression and monitoring current drug efficacy?

    • Hi Steve! Thanks for your comment. I think you make a fair point there that nothing excludes that NfL will join the list of other promising biomarkers that in the end never reach clinical practice. As the scientific evidence for both is strong, I think it will be mainly dependent on how easily you can sort out the logistics. NfL finds itself probably at the better end in terms of logistics as it is a lab test which could be requested together with other blood work whereas OCT requires collaboration with the ophthalmologists and separate appointments. However, also the Simoa technology is not used in clinical laboratories (yet) and relatively labour intensive in terms of calibration etc. TBC

        • Superinteresting! One of the few that has prospective data. But again: good example of how associations on group level and statistical significance don’t mean anything for the clinical practice. The authors: “Based on this work, a one-off baseline sNfL level alone does not seem to be particularly helpful as a specific test to highlight individuals at greatest risk who would benefit the most from early and intensive treatment.” It’s a validation of the ideas in the blogpost 😉



Recent Posts

Recent Comments