Tuesday, 27 June 2017

Blood NFL as marker of no evidence of disease activity with fingolimod

Plasma neurofilament light chain levels in patients with MS switching from injectable therapies to fingolimod

Fredrik Piehl, Ingrid Kockum, Mohsen Khademi, Kaj Blennow, Jan Lycke, Henrik Zetterberg, Tomas Olsson

Multiple Sclerosis Journal, First Published 19 Jun 2017.doi: 10.1177/1352458517715132


Neurofilament light chain (NFL) is a cerebrospinal fluid (CSF) marker of neuroaxonal damage in multiple sclerosis (MS).

To determine the correlation of NFL in CSF and serum/plasma, and in plasma after switching from injectable MS therapies to fingolimod.

A first cohort consisted of MS patients (n = 39) and neurological disease controls (n = 27) where CSF and plasma/serum had been collected for diagnostic purposes. A second cohort (n = 243) consisted of patients from a post-marketing study of fingolimod. NFL was determined with Single Molecule Array (Simoa™) technology (detection threshold 1.95 pg/mL).

Mean NFL pg/mL (standard deviation (SD)) was 341 (267) and 1475 (2358) in CSF and 8.2 (3.58) and 17.0 (16.94) in serum from controls and MS, respectively. CSF/serum and plasma/serum levels were highly correlated (n = 66, rho = 0.672, p < 0.0001 and n = 16, rho = 0.684, p = 0.009, respectively). In patients starting fingolimod (n = 243), mean NFL pg/mL (SD) in plasma was reduced between baseline (20.4 (10.7)) and at 12 months (13.5 (7.3), p < 3 × 10−6), and levels remained stable at 24 months (13.2 (6.2)).

NFL in serum and CSF are highly correlated and plasma NFL levels decrease after switching to highly effective MS therapy. Blood NFL measurement can be considered as a biomarker for MS therapy response.

Yesterday, I spoke at the European Academy of Neurology (EAN) Congress 2017 in Amsterdam about the utility of neurofilament analysis in neurological disorders. My point was to get across to the audience that neurofilaments as a biomarker was here to stay, but also to make it known that as far as MS management is concerned the goal post is shifting. We are looking to a future in which long-term remission (tantamount to a cure) is achievable, and not simply a woefully inadequate hype that recedes at the end of a conference. So how do the current biological fare in slowing down nerve loss?
I know of two therapies that have already demonstrated this; natalizumab (Tysabri) and fingolimod in the cerebrospinal fluid (CSF). Here the authors present more work on fingolimod, but this time looking at blood neurofilament levels (serum/plasma). The question is whether we are able to substitute blood neurofilament measures effectively for CSF neurofilament measures?
When reading through this paper, a couple of things stand out: 1) the levels in the CSF are ~100-fold higher in the CSF than in the blood, whether it be in PwMS or in controls; 2) the relative ratio's of difference between PwMS and controls in the CSF and blood is ~4 and ~2, respectively based on mean results. In a nutshell, there is more room to detect real changes in neurofilament levels in the CSF than in blood. Well big deal you say, but it matters. Maybe not in a clinical trial with 100+ participants, but at an individual level. If I'm a clinician strongly considering using this test in practice, I'll want to know these relative numbers as I may be basing treatment decisions on these numbers. Interestingly, another group (Disanto et al. JNNP 2016; 87:126-129) previously reported a three fold difference between CIS subjects and controls in serum neurofilaments, but using a different platform (mesoscale, as opposed to Simova used in this study). Maybe a more sensitive method is not all that its cracked up to be?!
All is not lost, as far as swapping from interferons or copaxone to fingolimod there appears to be a significant drop in blood NfL levels at 12 months, which is sustained at 24 months (see figure below). Not all subjects demonstrated a reduction in NfL levels after switching to fingolimod, 49 demonstrated a rise in NFL levels. They tended to be older (see figure below) and had a higher age of onset of disease among other factors. 

There are a lot of archived samples sitting in freezers around the world from clinical trials, now may be the time to start analysing them...I say London is open!

Figure: (a) NfL levels in those switching from injectables to fingolimod sampled at time 0, 12 and 24 months. Mean NfL levels reduced by 34% at 12 months; (b) blood NfL levels correlate with age.


  1. And the levels of neurofilaments in both blood and CSF in Alemtuzumab and Cladribine, have anyone looked at it?

    1. We're on the case. I have a study starting on alemtuzumab soon...


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