Background Neurofilament light chain (NfL) represents a promising biomarker for axonal injury. We present the first exploratory study on serum NfL in patients with a clinically isolated syndrome (CIS) and healthy controls.
Methods We investigated serum NfL levels in 100 patients with CIS with a short conversion interval to clinically definite multiple sclerosis (MS) (fast converters (FC), median (IQR) conversion time: 110 days (79–139)); 98 patients with non-converting CIS (non-converters (NC), follow-up: 6.5 years (5.3–7.9)); and 92 healthy controls.
Results NfL levels were higher in FC (24.1 pg/mL (13.5–51.8)) and NC (19.3 pg/mL (13.6–35.2)) than in healthy controls (7.9 pg/mL (5.6–17.2)) (OR=5.85; 95% CI 2.63 to 13.02; p=1.5×10−5 and OR=7.03; 95% CI 2.85 to 17.34; p=2.3×10−5, respectively). When grouping FC and NC, increased serum NfL concentration was also associated with increasing numbers of T2 hyperintense MRI lesions (OR=2.36; 95% CI 1.21 to 4.59; p=0.011), gadolinium-enhancing lesions (OR=2.69; 95% CI 1.13 to 6.41; p=0.026) and higher disability scores (OR=2.54; 95% CI 1.21 to 5.31; p=0.013) at CIS diagnosis.
Conclusions If replicated in future studies, serum NfL may represent a reliable and easily accessible biomarker of early axonal damage in CIS and MS.
- MULTIPLE SCLEROSIS
Statistics from Altmetric.com
Biomarkers of disease activity in MS are urgently needed.1 Neurofilaments (Nf) are structural elements of neurons composed of three Nf chains (light (NfL)), medium and heavy) and α-internexin in the central nervous system or peripherin in the peripheral nervous system. Nf are released in the extracellular space following neuronal death and as such are considered a candidate biomarker of ongoing neurodegeneration. Levels of Nf are abnormally high in the cerebrospinal fluid (CSF) in patients with clinically isolated syndrome (CIS) and MS; levels also appear to correlate with measures of disease activity, including lesions seen on MRI and disability scores.2–4 Obtaining CSF via lumbar puncture is a relatively invasive procedure; this limits the potential use of Nf as biomarkers in MS, especially in longitudinal studies that would require repeated sampling. We recently developed a sensitive electrochemiluminescence (ECL)-based immunoassay for the quantification of NfL in serum.5 ,6 The aim of this study was to investigate whether serum NfL levels were: (1) higher in patients with CIS than in healthy individuals; (2) associated with conversion to clinically definite MS (CDMS); (3) associated with other markers of disease activity, including MRI lesions, disability scores and oligoclonal bands (OCB) in the CSF.
As part of our International Clinically Isolated Syndrome Study Group, a total of 1047 CIS cases were collected across 33 centres located in 17 different countries between November 1986 and December 2011.7 All patients were followed-up for a minimum of 2.0 years. The median follow-up was 4.3 years (IQR 2.9–6.4). During this time, 623 patients (59.5%) converted to CDMS (Poser criteria) after a median of 421 (IQR 212–853) days.7 ,8 This study was approved by the corresponding local ethics committees and participants gave written informed consent.
We selected 100 patients with the shortest time of conversion to CDMS (fast converters (FC)) and 100 with the longest follow-up time in the absence of conversion (non-converters (NC)). Expanded disability status scale (EDSS), number of T2 hyperintense and gadolinium-enhancing (Gd+) lesions on cranial MRI and presence of IgG OCB in the CSF at the time of CIS were assessed in each participating centre as part of the diagnostic workup.7 ,9 Serum sampling was performed at the time of CIS diagnosis and EDSS calculations. EDSS and Gd+ lesion data were available on 170 and 146 patients with CIS, respectively. Patients were grouped in categories based on EDSS scores (0.0–1.0 vs 1.5–2.0 vs >2.0), number of T2 lesions (0–1 vs 2–9 vs >9) and presence or absence of Gd+ lesions and OCB. We measured serum NfL concentration in CIS samples and in a cohort of British healthy controls (HC, n=92) using our previously validated ECL immunoassay.5 HC were recruited from university workers and external recruits, and none of them had a personal or family history of demyelinating disease. The investigators who conducted NfL measurements had no access to the clinical data. Intra-assay and inter-assay (low, medium and high concentration control sample) variability was below 15%.
Variables were described as median with IQR and counts with percentages. Normalised (log 10) NfL levels were treated as a continuous variable and used for all analyses. Logistic regression models were used to assess the ability of NfL levels to predict disease status (FC vs HC, NC vs HC and FC vs NC). Similarly, we tested the association between NfL levels and markers of disease activity using logistic regression (NfL predicting OCB and Gd+ status) and ordinal regression models (NfL predicting increase in T2 and EDSS categories). In all models, results were corrected for both age and sex. All analyses were performed using R (http://www.r-project.org/).
The time interval between date of CIS onset and serum sampling was similar between the two groups (FC: 24.0 days (14.0–39.0); NC: 31.5 (19.0–62.0)), and no correlation between serum NfL levels and time of serum sampling was observed (Pearson's r=0.04, p=0.50). Two NC patients had insufficient sample volume for serum NfL measurement and were, therefore, excluded from the analysis. The median time to conversion to CDMS in FC was 110 days (79–139), while the median follow-up time in NC was 6.5 years (5.3–7.9). The general characteristics of FC, NC and HC are provided in table 1.
No significant difference in serum NfL levels were observed across centres (analysis of variance (ANOVA) p=0.18). There was no correlation between age and NfL in either CIS (Pearson's r=−0.11, p=0.10) or HC (Pearson's r=−0.02, p=0.88). NfL levels were higher in FC (24.1 pg/mL (13.5–51.8)) and NC (19.3 pg/mL (13.6–35.2)) than in HC (7.9 pg/mL (5.6–17.2)) (figure 1). Increasing NfL levels were significantly associated with FC status (OR=5.85; 95% CI 2.63 to 13.02; p=1.5×10−5) and NC status (OR=7.03; 95% CI 2.85 to 17.34; p=2.3×10−5) compared to HC (table 2 upper panel). When comparing FC vs NC, NfL levels were not associated with fast conversion to CDMS. However, when FC and NC were grouped together, NfL concentration was positively associated with presence of Gd+ lesions (OR=2.69; 95% CI 1.13 to 6.41; p=0.026), increasing T2 lesion load (OR=2.36; 95% CI 1.21 to 4.59; p=0.011), increasing EDSS (OR=2.54; 95% CI 1.21 to 5.31; p=0.013), but not OCB status (OR=1.66; 95% CI 0.75 to 3.70; p=0.214) (table 2 lower panel, figure 2).
The number of patients with OCB (87.0% vs 61.2%), Gd+ (58.2% vs 37.3%) and >9 T2 lesions (57.0% vs 25.5%) was higher in FC than NC. EDSS scores were also overall higher in FC than NC (2.0 (1.5–3.0) vs 1.5 (1.0–2.0)). Presence of OCB, increasing number of T2 lesions, presence of Gd+ lesions and higher EDSS scores were all positively associated with FC status (table 2 upper panel).
This is the first exploratory study investigating serum NfL in patients with CIS. Our first aim was to assess whether NfL levels were higher in patients with CIS than in healthy individuals. Levels were significantly higher in FC as well as NC than in HC. This supports the presence of accumulating axonal injury in the very early phase of MS, and potential ongoing Wallerian degeneration from focal lesion(s) presenting as CIS. It is important to note that the high ORs obtained when comparing FC and NC vs HC should be interpreted with caution since NfL levels were log transformed to be normally distributed. We noted that some HC had higher NfL levels than patients with CIS. To what extent these high NfL levels in HC represent minor neurological insults occurring in day-to-day life (eg, minor head injuries or compression of peripheral nerves) rather than normal synaptic and neuronal turnover is unclear and needs further study.
We then tested the association of NfL levels with conversion to CDMS. We decided to use Poser rather than MRI criteria as evidence of dissemination in time and space because MRI protocols were not uniform across centres. The difference in serum NfL levels between FC and NC was not statistically significant. Therefore, while increased serum NfL levels appear to be a good indicator of axonal damage, this does not appear to be specific to those patients with CIS who undergo a relatively rapid conversion to CDMS. Increased serum NfL and Nf heavy chain levels have been found in neurological conditions other than MS, such as amyotrophic lateral sclerosis and Guillain-Barré syndrome, and have also been used as a marker of neurotoxicity after chemotherapy.5 ,6 ,10 ,11 Results from previous studies in CIS have been controversial with one study showing slightly higher CSF NfL levels in patients with CIS converting to CDMS than in NC,4 while other studies have not.12 ,13 It should be noted that even in this cohort, serum NfL levels were to some extent higher in FC than NC and it is plausible that by increasing the sample size the difference may become statistically significant. Serum samples are more easily accessible compared to CSF, but also more distant from the pathological process taking place in the central nervous system. This may affect the sensitivity of serum NfL to predict conversion to CDMS. Furthermore, we did not include additional factors which may influence the risk of conversion, such as MS-associated genetic variants, vitamin D status and smoking history.
Finally, we investigated the association of NfL with other markers of disease activity. We found that higher serum NfL concentration was associated with the presence of both T2 and Gd+ lesions; the former is a marker of burden of disease and the latter, a marker of acute focal inflammation. Furthermore, NfL levels increased with increasing EDSS at the time of CIS. These results are in agreement with previous studies showing that Nf levels in the CSF are correlated with both MRI and clinical markers of MS disease activity.3 ,4 ,14 ,15 Despite this and the relatively low number of tests that were performed, false-positive associations cannot be excluded and we did not correct our results for multiple testing. Replication in independent samples is, therefore, required. It is also important to determine if raised serum NfL levels at initial clinical presentation predict long-term disability outcomes in a similar manner to CSF NfL levels.16
In conclusion, serum NfL levels are abnormally high in patients with CIS, appear to correlate with MRI activity and disability scores, and may represent a promising and easily accessible biomarker deserving of further exploration.
The authors would like to thank the French Multiple Sclerosis Observatory and BioMSeu: Consortium for CSF biomarker research for supporting this study.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Collaborators For complete list of members see online supplementary appendix 1.
Contributors GD has made substantial contributions to conception and design, acquisition, analysis and interpretation of the data, drafting of the manuscript, critical revision of the manuscript for important intellectual content and statistical analysis. RA has made substantial contributions to the acquisition, analysis and interpretation of the data, critical revision of the manuscript for important intellectual content, administrative, technical or material support. RD has made substantial contributions to critical revision of the manuscript for important intellectual content. VM, GDC, TR, EE, ET, MT, NN and CT have made substantial contributions to critical revision of the manuscript for important intellectual content, administrative, technical or material support. LK has made substantial contributions to critical revision of the manuscript for important intellectual content. GG has made substantial contributions to conception and design, critical revision of the manuscript for important intellectual content, obtained funding and did supervision of the work. JK has made substantial contributions to conception and design, acquisition, analysis and interpretation of the data, critical revision of the manuscript for important intellectual content, statistical analysis, obtained funding and did supervision of the work.
Funding This work was supported by institutional funding, the Swiss MS Society and an unrestricted grant by Genzyme.
Competing interests EE has received and dedicated to research support fees for board membership, consultancy or speaking, or grants, in the last 2 years, from Biogen Idec, Novartis, Sanofi-Aventis, Genzyme, Pharmstandart, R-Pharm, Pharmsyntez, Genfa Medica, Takeda and Generium. ET has received honoraria, travel grants or research grants from Biogen, Genzyme, Merck Serono, Novartis and Teva pharma. MT has served on scientific Advisory Boards for Biogen Idec, Novartis, Roche and Merck Serono; has received speaker honoraria from Biogen Idec, Bayer-Schering, Sanofi Aventis, Merck-Serono, Teva and Novartis; has received research grants from Biogen Idec, Merck-Serono and Novartis. NN is employed by UmanDiagnostics AB, Sweden. LK reports, the University Hospital Basel as employer of LK has received and dedicated to research support fees for board membership, consultancy or speaking, or grants, in the last 3 years, from Actelion, Advancell, Allozyne, Bayer, Bayhill, Biogen Idec, BioMarin, CSL Behring, Eli Lilly, European Union, GeNeuro, Genmab, Gianni Rubatto Foundation, Glenmark, Merck Serono, MediciNova, Mitsubishi Pharma, Novartis, Novartis Research Foundation, Novonordisk, Peptimmune, Roche, Roche Research Foundation, Sanofi-Aventis, Santhera, Swiss MS Society, Swiss National Research Foundation, Teva, UCB and Wyeth. GG serves on scientific advisory boards for Merck Serono and Biogen Idec and Vertex Pharmaceuticals; served on the editorial board of Multiple Sclerosis; has received speaker honoraria from Bayer Schering Pharma, Merck Serono, Biogen Idec, Pfizer Inc, Teva Pharmaceutical Industries Ltd, Sanofi-Aventis, Vertex Pharmaceuticals, Genzyme Corporation, Ironwood and Novartis; has served as a consultant for Bayer Schering Pharma, Biogen Idec, GlaxoSmithKline, Merck Serono, Protein Discovery Laboratories, Teva Pharmaceutical Industries Ltd—Sanofi-Aventis, UCB, Vertex Pharmaceuticals, GW Pharma, Novartis and FivePrime; serves on the speakers’ bureau for Merck Serono; and has received research support from Bayer Schering Pharma, Biogen Idec, Merck Serono, Novartis, UCB, Merz Pharmaceuticals, LLC, Teva Pharmaceutical Industries Ltd—Sanofi-Aventis, GW Pharma and Ironwood. JK was supported by an ECTRIMS Research Fellowship Programme and by the ‘Forschungsfonds’ of the University of Basel, Switzerland, and has received research support from the Swiss MS Society, Swiss ALS Society, Protagen AG, Roche, Genzyme and Novartis, and served in scientific advisory boards for Genzyme/Sanofi-Aventis, Merck Serono and Novartis Pharma.
Patient consent Obtained.
Ethics approval The study was approved by the corresponding local ethics committees.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.