Multiple Sclerosis Journal - October 2017 - 1447

N Dubuisson, F Puentes et al.
in demyelination and remyelination in EAE.111 CSF
OPN levels are higher than plasma OPN levels suggesting that the OPN is preferentially secreted in the
CSF.112,113
Strength and specificity of association. OPN levels
are increased in MS cases compared to healthy
controls and non-inflammatory disorders.113 OPN
levels closely correlate with clinical exacerbations
in CIS and RRMS,114,115 and precede an increase in
gadolinium-enhancing lesion number.116 However,
changes in OPN levels are not specific for MS, and
due to its diverse roles, it is also increased in other
neurological and non-neurological disorders, particularly those with an underlying inflammatory
pathophysiology.112,113 Furthermore, despite the link
with active disease, there appears to be no correlation with disability.112
Consistency of findings. In a majority of cases, OPN
levels are elevated in active MS, a finding that is
more consistent when sampling the CSF than plasma.
However, Comabella et al.117 did note that plasma
OPN levels were also elevated in SPMS cases
compared with healthy controls, suggesting a role in
chronic disease activity, although not consistent.115 In
addition, CSF OPN levels have also been noted to be
elevated in PPMS cases, suggesting possible ongoing
inflammation.114 Timing of sampling may therefore
be key, with significantly higher levels during relapse
than stable disease.117,118 However, in a longitudinal
evaluation, Runia et al.119 did not find an association
between plasma OPN levels and relapse risk, or a difference in OPN levels when sampled during active
disease versus stable disease. There may also be
other confounders to consider, for example, a trend
toward higher plasma OPN levels in those who
reported infections and other autoimmune disorders,
and in those who were pregnant or postpartum during
sampling.120
Temporal sequence of association. Plasma OPN
levels have been shown to rise 1 month before a rise
in gadolinium-enhancing lesions,116 and CSF OPN
levels have been shown to return to baseline within
3 weeks of an acute relapse.114 The rise in OPN
levels therefore seems to precede and coincide with
MS disease activity.
Biological gradient. Although there appears to be no
association between OPN levels and disability,119,120
they are modified by anti-inflammatory therapies. In
RRMS, plasma OPN levels returned to remission
levels following treatment with IFNβ,117 glatiramer
acetate and natalizumab treatment.120,121 CSF OPN
journals.sagepub.com/home/msj

levels have also been demonstrated to decrease after
1 year natalizumab therapy in RRMS.122
Biomarker technology
The divide between hypothesis-driven biomarker
analysis and hypothesis-generating biomarker discovery programmes is blurring, partly driven by the
availability of new technologies over the last
decade. Prior to this, analysis was reliant on laborious benchwork and the development of enzymelinked immunosorbent assays (ELISAs). Meso Scale
Discovery's (MSD) electrochemiluminescence (ECL)based assays and Luminex bead technologies have
led to the development of high-throughput multiplex
assays with greater sensitivity, which is ideal for
CSF biomarker analysis (where volume is scarce),
particularly for low abundant proteins. Recently,
Quanterix's Simoa analyser offers a standardized
way of analysing neurodegenerative biomarkers
such as NfL in sera and CSF with better correlation
between the two and even more sensitivity than ECL
or ELISA.123 The latest introduction of aptamerbased technology (Somalogic), which uses singlestranded oligonucleotides to bind protein targets
with high affinity, now allows quantification of at
least 1300 analytes in a single sample. While proteomics remain the gold standard for discovery-level
analysis, the addition of labelled peptides is a first
step toward semi-quantitation of analytes. Recently,
the introduction of isobaric tags now allows multiplex capability.124 Proteomic technologies have the
added advantage of allowing analysis of proteins
where antibodies are not available.
Conclusion
In this review, we have critically evaluated the
most promising neurodegenerative biomarkers in MS
according to sufficient biological plausibility, strength
and specificity of association, consistency of findings,
temporal sequence and demonstration of a biological
gradient. The neurofilament proteins, neurofilament
antibodies, tau, NAA, chitinase and chitinase-like
proteins and OPN with their inherent advantages and
disadvantages have progressed our understanding of
the neurodegenerative process in MS better than
before. With the added advantage of the revolution in
biomarker technology, it is hoped that complex and
large scale data-driven models of neurodegeneration
can be built in the near future. In the meantime, biomarker researchers should carefully evaluate any
methodological shortcomings and progress the most
promising biomarkers to the next level, that is, the
ability to make predictions in individuals.
1447
https://journals.sagepub.com/home/msj
Table of Contents for the Digital Edition of Multiple Sclerosis Journal - October 2017
