CSF neurofilament light chain and OCB predict conversion to MS in RIS

CSF neurofilaments are ready for prime time. Do you agree? Wouldn’t you as someone with MS not want to know what their CSF NFL levels were? 

Brain. 2018 Feb 14. doi: 10.1093/brain/awy021. [Epub ahead of print]

Neurofilament light chain and oligoclonal bands are prognostic biomarkers in radiologically isolated syndrome.

Matute-Blanch C, Villar LM, Álvarez-Cermeño JC, Rejdak K, Evdoshenko E, Makshakov G, Nazarov V, Lapin S, Midaglia L, Vidal-Jordana A, Drulovic J, García-Merino A, Sánchez-López AJ, Havrdova E, Saiz A, Llufriu S, Alvarez-Lafuente R, Schroeder I, Zettl UK, Galimberti D, Ramió-Torrentà L, Robles R, Quintana E, Hegen H, Deisenhammer F, Río J, Tintoré M, Sánchez A, Montalban X, Comabella M.


The prognostic role of cerebrospinal fluid molecular biomarkers determined in early pathogenic stages of multiple sclerosis has yet to be defined. In the present study, we aimed to investigate the prognostic value of chitinase 3 like 1 (CHI3L1), neurofilament light chain, and oligoclonal bands for conversion to clinically isolated syndrome and to multiple sclerosis in 75 patients with radiologically isolated syndrome. Cerebrospinal fluid levels of CHI3L1 and neurofilament light chain were measured by enzyme-linked immunosorbent assay. Uni- and multivariable Cox regression models including as covariates age at diagnosis of radiologically isolated syndrome, number of brain lesions, sex and treatment were used to investigate associations between cerebrospinal fluid CHI3L1 and neurofilament light chain levels and time to conversion to clinically isolated syndrome and multiple sclerosis. Neurofilament light chain levels and oligoclonal bands were independent risk factors for the development of clinically isolated syndrome (hazard ratio = 1.02, P = 0.019, and hazard ratio = 14.7, P = 0.012, respectively) and multiple sclerosis (hazard ratio = 1.03, P = 0.003, and hazard ratio = 8.9, P = 0.046, respectively). The best cut-off to classify cerebrospinal fluid neurofilament light chain levels into high and low was 619 ng/l, and high neurofilament light chain levels were associated with a trend to shorter time to clinically isolated syndrome (P = 0.079) and significant shorter time to multiple sclerosis (P = 0.017). Similarly, patients with radiologically isolated syndrome presenting positive oligoclonal bands converted faster to clinically isolated syndrome and multiple sclerosis (P = 0.005 and P = 0.008, respectively). The effects of high neurofilament light chain levels shortening time to clinically isolated syndrome and multiple sclerosis were more pronounced in radiologically isolated syndrome patients with ≥37 years compared to younger patients. Cerebrospinal fluid CHI3L1 levels did not influence conversion to clinically isolated syndrome and multiple sclerosis in radiologically isolated syndrome patients. Overall, these findings suggest that cerebrospinal neurofilament light chain levels and oligoclonal bands are independent predictors of clinical conversion in patients with radiologically isolated syndrome. The association with a faster development of multiple sclerosis reinforces the importance of cerebrospinal fluid analysis in patients with radiologically isolated syndrome.

Figure:Kaplan-Meier curves for time to CIS and time to multiple sclerosis in RIS patients according to high and low CSF CHI3L1 (A) and NfL levels (B) and (C) positive and negative oligoclonal bands.

Ideas are the next currency of the world, and I believe this is where biomarkers have the greatest impact to make in MS. At Barts Health we offer CSF neurofilament levels (NfL) on the NHS; the feedback from clinicians who have used the test has been incredible – one colleague even described it as revolutionary and the best invention in MS in the last 10 years.

Comabella’s team have examined three important biomarkers in the CSF (cerebrospinal fluid, obtained via a lumbar puncture) as predictors of developing the first clinical symptoms of MS (i.e. CIS) or MS itself (more than one clinical event) when you simply have an abnormal MRI (RIS or asymptomatic MS or non-specific symptoms):

  • Neurofilament light chain – proteins maintaining the neuronal framework

  • Oligoclonal bands – antibodies produced by the immune cells

  • Chitinase 3 like 1 protein (CHI3L1) – secreted by activated macrophages (part of the innate immune system)

It is known that high CSF levels of both NfL and CHI3L1 predict the conversion of CIS to MS. In addition, the presence of CSF OCBs also increase the risk of developing MS, so much so that it is now possible to diagnose MS at the first presentation simply by having a positive OCB result in addition to an abnormal MRI (i.e. instead of waiting for the next lesion to appear on MRI as demonstration dissemination in time [DIT], positive OCBs can now be substituted for DIT).

And not surprisingly as this is all one disease, the investigators found that not only were these biomarkers useful in predicting conversion to MS from CIS but also RIS.

Seventy-five RIS patients were studied and followed up on average 2.8y. Around 31% converted to CIS within 2y and of this 91% were later diagnosed as having multiple sclerosis. CHI3L1 was not very useful, but, both NFL and OCBs were independent risk factors for conversion to CIS and multiple sclerosis from RIS. Other factors which were not found to be useful predictors of conversion were age, number of T2 brain lesions on MRI, gender and treatment. Moreover, high NfL levels and a positive OCB were associated with shorter time to conversion to CIS and MS (see figure above). Only with CSF NfL was the age <37y was found to be a predictor for shortening the time of conversion of RIS to CIS and MS (whether this is because there is an increase in age-related axonal loss as you get older is not known).

In summary, although treatment of RIS outside of the US is controversial, the results from this study make it tantalizing enough to intervene early in those with raised CSF NFL and positive OCBs.

About the author

Neuro Doc Gnanapavan


  • In an ideal world, how would you treat a menopausal 50+ patient referred to you with RIS?
    And if a similar patient were referred to you with mild symptoms suggestive of MS?

  • 50 is the new 30’s! Jokes aside in an ideal world I would in both cases tailor management based on their current disease profile, both structural (MRI appearances) and biological (NFL, OCB status…). I would also if high risk for progression over the next 10y if the person is OCB positive try a B cell therapy preferentially. Having this information is key in planning future management – it’s the Star Trek of neurology!

  • I recently asked my neurologist whether I could have the lumbar puncture to test my CSF neurofilament levels/OCB status but was told emphatically "no" because the NHS didn't provide the test. I live in the Westcountry. Why are you allowed to do this test on the NHS at Barts? Would it be worth paying to have the procedure done privately and if so, do you know how much it would cost?

    • Hi, no need to pay for this out of your own pocket. Normally the lab in your hospital will liaise with our lab and pathway for a test is set up. I will upload our request form onto the blog.

  • NDG, do you think that at some point the presence of Neurofilaments in the CSF will enter as a criterion for the diagnosis of the progression of MS?

    Because I talked to my neurologist and said that I would like to do an analysis of my CSF to find out if I have neurofilament in my CSF, since although I am stable and apparently a "responder" to Copaxone, I have positive OCB. and said that "there is not yet a concession on the use or not of the presence of neurofilaments in CFS to determine the progression of the disease".



Recent Posts

Recent Comments