Tuesday, 14 June 2016

Diffusion of innovations: neurofilaments predict brain atrophy 15 years on

Mult Scler. 2016 May 11. pii: 1352458516645206. [Epub ahead of print]

Elevated CSF neurofilament proteins predict brain atrophy: A 15-year follow-up study.

Petzold A, Steenwijk MD, Eikelenboom JM, Wattjes MP, Uitdehaag BM.

BACKGROUND: Body fluid and structural imaging biomarkers give information on neurodegeneration. The relationship over time is not known in multiple sclerosis.

OBJECTIVE: To investigate the temporal relationship of elevated cerebrospinal fluid (CSF) neurofilament (Nf) protein levels, a biomarker for axonal loss, with magnetic resonance imaging (MRI) atrophy measures.

METHODS: In patients with multiple sclerosis, CSF Nf heavy chain (NfH) phosphoform levels were quantified at baseline and dichotomised into 'normal' and 'high'. Atrophy was assessed by MRI at baseline and 15-year follow-up using SIENAX and FreeSurfer software.

RESULTS: High baseline CSF NfHSMI35 levels predicted pronounced atrophy at 15-year follow-up (odds ratio (OR): 36, p < 0.01), in the absence of baseline brain atrophy (OR: 28, p < 0.05), for the averaged MRI normalised brain volume (1.44 L vs 1.33 L, p < 0.05), normalised grey matter volume (0.77 L vs 0.69 L, p < 0.01) and putamen (12.7 mL vs 10.7 mL, p < 0.05). Region-specific calculations including the spinal cord showed that a power of >80% is reached with 14-50 patients.

CONCLUSION: These data suggest that high CSF NfH levels are an early predictor of later brain and spinal cord atrophy using structural imaging biomarkers and can be investigated in reasonably sized patient cohorts.


This is an area of research that our group understand well; so much so, that we have translated this piece of work into our clinics. Our patients now undergo lumbar punctures at diagnosis to evaluate their risk of short-term and long-term disease progression based on neurofilament levels. Prof G often talks about the adoption curve, comprising of the innovators (the first group to use a product - the more risk oriented), followed by early adopters, and eventually the laggards (the very conservative). This concept is very pertinent to what we're doing currently and we hope to revolutionize the way MS is managed in real-life terms with this test. This is the beginning of personalized medicine.

Axel Petzold as early as 2005 demonstrated that baseline CSF neurofilament heavy chain levels (NfH, a marker of axonal damage) could predict irreversible disease progression at 3 years (Petzold A et al. Axonal damage accumulates in the progressive phase of multiple sclerosis: Three year follow up study.J Neurol Neurosurg Psychiatry 2005; 76:206211). Here, the group extend their 3-year follow-up data and test whether high baseline CSF NfH predicts more severe brain volume loss after 15 years? The short answer to this is a 'YES'! The Odds ratio (which is the risk of a particular outcome if a certain factor is present) of this happening is 36. To give you an idea of the importance of this statement, the odds ratio of low vitamin D predicting the development of MS is ~2.

Another piece of interesting information also arises from this work, which is that their 3 year study did not show this association with brain volume loss; the association only emerges at 15 years, indicating a serious lag to when axonal injury is manifest on MRI (see Fig below). Very annoying if your intention is to design a clinical trial using brain atrophy as the primary outcome measure!


Fig: Model suggesting that development of disability related to axonal degeneration directly causes an increase in CSF neurofilament levels. Elevated neurofilament levels precede global brain atrophy on MRI.

27 comments:

  1. So if a medication is working you will see a stabilization/reduction of neurofilament levels and if a medication is not working you will see an increase in neurofilament levels and a physician should switch medication? What if neurofilament levels continue to further increase despite use of most potent anti-inflammatory DMDs?

    How can a drug trial claim a medication is efficacious re: reduction of disability using MRI to detect brain atrophy when it lags so far behind and the trials are so short? Shouldn't CSF neurofilament be the gold standard to predict disability in trials, not MRI. Sorry for all the questions.

    ReplyDelete
    Replies
    1. Yet to understand how CSF results fit in the prescription and shifting strategy...

      It seems like "white noise" until then.

      Delete
    2. Well not "white noise" but clinical practice has to catch up with CSF neurofilament technology as NDG and colleagues expose it's relevance and refine it.

      Delete
    3. The evidence is pointing towards neurofilaments predicting neurodegeneration (at the most basic level this is mostly because their is axonal breakdown), which leads to long term disability progression. So anti-inflammatory treatments lead to a reduction in NFL (light chain levels) - natalizumab and fingolimod (only these two have been looked at) by reducing relapses, which we know lead to disability. NFH (heavy chain) has been shown to drop in the lamotrigine trial in SPMS, it hasn't been looked at with other treatments/ their isn't published evidence.
      The idea is to guide our treatment decisions, what we are are discovering is that we no longer have to simply rely on our clinical sixth sense that some has active disease - the result adds to it (and there has been a lot of 'well yes OK moments'!).
      The hope is that neurofilaments enter in as NEDA tick list, which is how we are using it. If the neurofilaments continue to increase despite conventional DMTs we need to step up and come up with new therapies - this is how you drive forward treatment developments.
      With regard to drug trials predicting reduction in atrophy, there is the conventional way of interpreting it, or the way which I and some of my colleagues see it. In an anti-inflammatory drug (say alamtuzumab) the brain atrophy rate plateaus at yr 4 - that's when the drug is working and achieving its effect.
      I was recently invited to present by the UK MS society clinical trials division about neurofilaments and many in the audience agreed that compared to brain atrophy and Optical coherence tomography, neurofilaments were the most robust measure. Again we come back to adoption theory.

      Delete
    4. "If the neurofilaments continue to increase despite conventional DMTs we need to step up and come up with new therapies"

      So natalizumab patients respond well to NFL from what I understand. What escalation treatment is available to them if they didn't despite being NEDA?

      Another one please: how do you breakdown drivers of brain atrophy between NFL and NFH?

      Thank you

      Delete
    5. From the AFFIRM study 37% of natalizumab achieve NEDA-4 at two years. Step up options are escalate or try a different class of therapy - stepping up = HSCT, different class = ocrelizumab or cladrabine.

      Your second question is more difficult to answer. By proportions alone there is more NFL than NFH in the CNS. NFH is found in large myelinated axons, whilst NFL is predominant in synapses. So there may be regional changes depending on the area involved - for example hippocampal NFL is topical in Alzheimer's disease. But in MS both are impacted on, it's the degree or intensity of the damage and duration of injury is important. Those with very high NFL levels also have raised NFH. Both have been shown to correlate with brain atrophy.

      Delete
  2. NDG- the vitamin D OR might be way more than 2, depends what you are quantifying. Likewise I suspect for neurofillies.

    ReplyDelete
    Replies
    1. True. We wait for an era where just looking at a simple graph should be self explanatory!

      Delete
  3. " Our patients now undergo lumbar punctures at diagnosis to evaluate their risk of short-term and long-term disease progression based on neurofilament levels"

    I am a patient of yours and was never offered an LP.

    How do you interpret the results: Do you guide low CSF patients to low risk/efficacy DMT and vice-versa?

    ReplyDelete
    Replies
    1. We've started in newly diagnosed at their first diagnostic LP - where OCB's are also done, and in patients where we think that there maybe subclinical disease which we're not getting a clear assessment of and need escalation. I've pretty much left it to the individual consultant managing the case. Some of our patients who avidly follow our blogs have asked! And we've done it and it has been useful to have. The treatment decision on DMT choice is already made before the test but may in the future change, in some patients it has made a difference and is swinging decisions.

      Delete
    2. Precisely what I had suspected - NFL seem to be a clinical trial biomarker rather than a DMT decision making variable on this stage...

      Yet to see how it will be deployed in clinical practice and DMT selection/switch.

      Delete
    3. lol this all makes me wonder about the value of my partner having a post diagnosis and treatment LP (she had one at diagnosis).

      of course, whatever the value may or may not be, my chances of actually talking my partner into it after the botched diagnosis LP which flattened her into a horizontal position for 3 weeks post LP, are not that great lol.

      still, one could try lol.

      Delete
    4. I wouldn't bother until someone responds to Anon 10:47 and articulate why it is beneficial in your partner's decision making process, particularly if they are on a high efficacy DMT already!

      Delete
    5. Perhaps if you posted as something other than anonymous it might be answered quicker. Just a thought.

      Delete
    6. For anyone who missed it, there's a post on how to use a name or pseudonym if you prefer here
      http://multiple-sclerosis-research.blogspot.com/2016/06/whats-in-name.html

      Delete
    7. I like being Anon :) I hope it's OK?

      Does being Anon dilute my point? if not, why not addressing it?

      Delete
    8. It makes it difficult to differentiate you from other Anons. Why not pick any name you fancy? it makes answering easier. Most other commenters seem happy to do this now.

      I'll leave it to NDG who will hopefully answer your point in the fullness of time (she's very busy).

      Delete
    9. Anon Wed June 15 8:43 and Anon Thurs June 16 8:40! I'm tempted to let any of the patients who have had the test to answer this one!

      Firstly, the data of this study is not about a clinical trial this is about picking up people at high risk early. It will be wrong of me to recommend on a decision which the individual has with their doctor to come to a conclusion on what it ultimately means to them. But information is power - if you have it, use it.

      Delete
    10. AnonymousThursday, June 16, 2016 8:40:00 am

      I'd bother and I'd be very interested to know her levels relative to levels at diagnosis, perhaps in two or three or four years after induction treatment. Even if no decision follows the information, I'm still interested to know.

      Delete
  4. What is the cutoff for a high CSF level?

    ReplyDelete
    Replies
    1. For an average young adult ~ 300pg/ml is acceptable for a normal result with NFL, we provide clinicians with the age related cut offs in the report. Let's say in high readings in MS I have seen 3000-7000pg/ml and in some cases higher than this.

      Delete
    2. Lower NFL means lower BVL, and lower inflammation (relapses or subclinical inflammation).
      What is your experience about Fingolimod or Natalizumab?

      Delete
  5. So it appears that neurofilament levels correlate with disease progression but what do you think is the pathological process behind this?

    ReplyDelete
    Replies
    1. Probably a number of different processes, demyelinated nerves having to work much harder to transmit impulses (by increasing the number of sodium channels) causing an energy drain on the nerve, activated microglia releasing toxic factors at low level over a long time coupled with the local environment giving a reduced level of support to the nerves.

      Delete
    2. I've even seen in a live microscopy model, that they also directly target the entire axon not just the myelin sheath.

      Delete

Please note that all comments are moderated and any personal or marketing-related submissions will not be shown.