TYPE Review

PUBLISHED 13 July 2023

DOI 10.3389/fimmu.2023.1162340

Cerebrospinal fluid inflammatory

OPEN ACCESS biomarkers for disease
EDITED BY
Rudolph Tanzi,
Harvard Medical School, United States
progression in Alzheimer’s
REVIEWED BY
Peter Kosa,
disease and multiple sclerosis:
National Institutes of Health (NIH),
United States
Giovanni Di Liberto,
a systematic review
Centre Hospitalier Universitaire Vaudois
(CHUV), Switzerland
Joke Temmerman 1,2,3, Sebastiaan Engelborghs 1,2,3,
*CORRESPONDENCE
Miguel D’haeseleer Maria Bjerke 1,2,3,4† and Miguel D’haeseleer 1,3,5*†
[email protected] 1
Vrije Universiteit Brussel, Center for Neurosciences (C4N), Jette, Brussels, Belgium, 2 Universiteit
† Antwerpen, Department of Biomedical Sciences and Institute Born-Bunge, Reference Center for
These authors have contributed
equally to this work and share Biological Markers of Dementia (BIODEM), Wilrijk, Antwerp, Belgium, 3 Universitair Ziekenhuis Brussel,
senior authorship Department of Neurology, Jette, Brussels, Belgium, 4 Universitair Ziekenhuis Brussel, Department of
Clinical Biology, Laboratory of Clinical Neurochemistry, Jette, Brussels, Belgium, 5 National MS Center
RECEIVED 09 February 2023 (NMSC), Neurology, Melsbroek, Steenokkerzeel, Belgium
ACCEPTED 12 June 2023
PUBLISHED 13 July 2023

CITATION
Temmerman J, Engelborghs S, Bjerke M
Inflammatory processes are involved in the pathophysiology of both Alzheimer’s
and D’haeseleer M (2023) Cerebrospinal
fluid inflammatory biomarkers for disease disease (AD) and multiple sclerosis (MS) but their exact contribution to disease
progression in Alzheimer’s disease and progression remains to be deciphered. Biomarkers are needed to define
multiple sclerosis: a systematic review.
Front. Immunol. 14:1162340.
pathophysiological processes of these disorders, who may increasingly co-exist
doi: 10.3389/fimmu.2023.1162340 in the elderly generations of the future, due to the rising prevalence in both and
COPYRIGHT ameliorated treatment options with improved life expectancy in MS. The purpose
© 2023 Temmerman, Engelborghs, Bjerke of this review was to provide a systematic overview of inflammatory biomarkers, as
and D’haeseleer. This is an open-access
article distributed under the terms of the
measured in the cerebrospinal fluid (CSF), that are associated with clinical disease
Creative Commons Attribution License progression. International peer-reviewed literature was screened using the
(CC BY). The use, distribution or PubMed and Web of Science databases. Disease progression had to be
reproduction in other forums is permitted,
provided the original author(s) and the measured using clinically validated tests representing baseline functional and/or
copyright owner(s) are credited and that cognitive status, the evolution of such clinical scores over time and/or the
the original publication in this journal is
cited, in accordance with accepted
transitioning from one disease stage to a more severe stage. The quality of
academic practice. No use, distribution or included studies was systematically evaluated using a set of questions for
reproduction is permitted which does not clinical, neurochemical and statistical characteristics of the study. A total of 84
comply with these terms.
papers were included (twenty-five for AD and 59 for MS). Elevated CSF levels of
chitinase-3-like protein 1 (YKL-40) were associated with disease progression in
both AD and MS. Osteopontin and monocyte chemoattractant protein-1 were
more specifically related to disease progression in AD, whereas the same was true
for interleukin-1 beta, tumor necrosis factor alpha, C-X-C motif ligand 13, glial
fibrillary acidic protein and IgG oligoclonal bands in MS. We observed a broad
heterogeneity of studies with varying cohort characterization, non-disclosure of
quality measures for neurochemical analyses and a lack of adequate longitudinal
designs. Most of the retrieved biomarkers are related to innate immune system
activity, which seems to be an important mediator of clinical disease progression in

Frontiers in Immunology 01 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

AD and MS. Overall study quality was limited and we have framed some
recommendations for future biomarker research in this field.
Systematic review registration: https://www.crd.york.ac.uk/prospero/, identifier
CRD42021264741.

KEYWORDS

inflammation, biomarkers, Alzheimer’s disease (AD), multiple scleorsis (MS),

disease progression

1 Introduction AD is classically viewed as a pure neurodegenerative disorder,

characterized by the deposition of amyloid beta (Ab)-containing
Alzheimer’s disease (AD) and multiple sclerosis (MS) are two plaques and neurofibrillary tangles composed of hyperphosphorylated
frequently occurring and debilitating disorders of the central tau protein in the brain. The earliest biomarker sign representing AD
nervous system (CNS). Whereas these conditions exhibit pathology is a reduced concentration of Ab1-42 and a decreased Ab1-42/
fundamental epidemiological and pathobiological differences, they Ab1-40 ratio in the cerebrospinal fluid (CSF), of which the latter is the
also share a number of striking commonalities, making them most specific for plaque pathology (16). Neuropathology may develop up
suitable for an integrative approach in translational research. to 20 years before the first manifestation of cognitive symptoms (17) and
Cognitive decline is the core clinical feature of AD, usually supports the concept that AD is a continuum with a long preclinical
starting with memory loss, language difficulties and visuospatial phase, followed by a prodromal phase of mild cognitive impairment
deficits (1). In MS, physical symptoms are often more visible (2), (MCI) that eventually leads to dementia (18). Positron emission
but impaired cognition (typically affecting information processing tomography (PET) studies have demonstrated an increased density of
speed, episodic memory, attention and executive function), which translocator protein (TSPO), a marker for neuroinflammation that is
may be subtle but therefore not necessarily less troublesome, has predominantly present in microglia, in patients with AD, as compared to
been reported in up to 70% of patients and can already be present healthy individuals (19). Disease-associated microglia (DAMs), a
from the early stages of the disease (3, 4). Neurodegeneration is the context-dependent microglia state identified by novel technologies,
ultimate pathological mechanism leading to functional decline in such as single-cell RNA sequencing (20), have been found in the
both entities, accounting as the main determinant of global long- vicinity of the amyloid plaques (21), while several well-established AD
term prognosis in MS (5), and highly correlating with cognitive risk genes closely relate to microglial functions and appear to be highly
decline in individuals with AD (6). Inflammatory responses are enriched in these cells (22). As such, the innate inflammatory response
involved as well, to varying degrees and as explained more into may be implicated in the early pathological process in AD. However, the
detail below, but their precise role in the respective heterogenous presentation of AD has previously split the field
neurodegenerative cascades remains to be elucidated (7). concerning the role of inflammation and hindered the establishment
Interestingly, the life expectancy of individuals with MS has of inflammatory biomarkers to monitor disease progression. It remains
significantly improved over the past decades, probably due to to be determined whether the inflammatory changes are actually a
ameliorated disease modifying treatments (DMT) and/or general driving force behind the pathology or merely represent a bystander effect,
medical care. As a consequence, and in combination with a rising and whether the inflammatory state is beneficial, by clearing pathological
prevalence of both disorders (8, 9), there is a growing likelihood of aggregates, or detrimental, by acting provocative for the degenerative
co-existence of MS with classic age-related dementias such as AD process (23). Conversely, an inadequate immune response in AD may
(10). Some studies have even reported an increased risk of AD equally lead to disease progression, as microglia surrounding the plaques
diagnosis in patients with MS, as compared to controls (11–13). In apparently fail to take part in phagocytosis (24). Anyhow, current
order to enable high-quality precision medicine, it will become literature does suggest that inflammatory CSF biomarkers are altered
increasingly important to unravel the unique and (possibly) shared in AD (25), but the impact on progression across the disease continuum
pathophysiological processes of both disorders. Biomarkers are vital needs further elaboration.
when investigating disease mechanisms as they are the in vivo proxy MS is a neuroinflammatory and -degenerative disorder of the CNS
of neuropathology, used in both medical routine and clinical trials with a causative mechanism that is incompletely understood but
to achieve formal (and early) diagnosis/patient stratification, predict presumably autoimmune in origin. Most patients (85%) start with a
disability progression and/or monitor treatment response (14, 15). relapsing-remitting (RR) pattern during which abrupt and at least
In addition, biomarker research may lead to the identification of partially recovering exacerbations of neurological dysfunction (termed
new druggable targets for disease modification, potentially leading relapses) are interchanged with periods of clinical stability. Many will
to prevention of functional decline in affected individuals. eventually transit into a secondary progressive (SP) phase characterized

Frontiers in Immunology 02 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

by a slower but sustained downhill course that may still be accompanied 2 Methods
by some relapses, whereas about 15% of patients experience a similar
primary progressive (PP) decline from disease onset (2). Focal 2.1 Research question
demyelinating lesions, resulting from acute inflammatory activity with
blood-brain barrier (BBB) disruption and perivenular infiltrates of We formulated our research question and conceptualized the
peripheral immune cells (including T-cells, B-cells, plasma cells and search terms by using the ‘Patient – Intervention – Comparator –
macrophages), form the pathological substrate for relapses (2). Recent Outcome – Time – Setting (PICOTS) framework’ (41). Our aim was
insights have learned that progressive MS is more likely to rely, among to provide a structured overview of the inflammatory biomarkers in
other mechanisms, on innate immune processes behind an intact BBB. the CSF (I) of patients belonging to the AD and/or MS continuum
Microglia - more specifically microglia inflamed in MS (MIMS; whose (P) that have been associated with disease progression (O), as
transcriptional profile partially overlaps with DAMs) (20, 26) - have been opposed to situations of clinical stability (C), in cross-sectional
found diffusely throughout the normal-appearing white matter and/or longitudinal (T) clinical studies (S). Papers were only found
(NAWM) and at the border of a particular subset of chronic slowly suitable for inclusion if disease progression was measured using (a)
expanding lesions (also called chronically active or smoldering lesions), clinically validated tests representing baseline functional and/or
both of which are features increasingly associated with the progressive cognitive status, (b) the evolution of such clinical scores over time
phase of the disease (27–29). Increased TSPO expression in the brain of and/or (c) the transitioning from one (possibly prodromal) disease
patients with MS was found to predict future disability progression and stage to a more severe stage (i.e., MCI to AD; CIS or RIS to clinically
several MS susceptibility genes are enriched in microglia, suggesting an definite MS; RR MS to SP MS). These criteria were also used as the
important role of the innate immune response in MS (30, 31). However, core structure for grouping the retrieved papers, which allows a
similar to the potential double-edged role in AD, microglia also facilitate consistent synthesis of the results. Our review was conducted
neuronal repair by clearing debris and stimulating remyelination (32). following the guidelines of the Preferred Reporting Items for
Indications of MS pathology presumably appear long before clinical Systematic Review and Analyses (PRISMA) (42), we refer to the
diagnosis, as evidenced by various alterations at the clinical, biochemical Supplementary Material for the related checklist (Table S1). The
and/or radiological level (33–37). Some of these prodromal protocol was submitted to the PROSPERO database (Internal
circumstances are well-defined by the terms ‘clinically and Prospective Register of Systematic Reviews; registration number:
radiologically isolated syndrome’ in case of first inflammatory CRD42021264741), maintained by the Centre for Reviews and
demyelinating episode without proof of dissemination in time (CIS) or Dissemination at the University of York (Heslington, UK), to
when magnetic resonance imaging (MRI) findings are strongly help avoid duplicate efforts.
suggestive for MS but without clinical repercussions (RIS), respectively.
Despite the identification of the aforementioned inflammatory players in
MS pathology, it remains to be seen how they may translate to disease 2.2 Search strategy
progression biomarkers. Thus far, only oligoclonal bands (OCB), which
are the product of intrathecal antibody secretion by activated B-cells, International peer-reviewed literature relevant to our PICOTS
have been included in the official criteria for MS diagnosis (38). research question was screened up to March 24, 2022 using the
So far, drug development mainly focused on disease-specific PubMed and Web of Science databases. To be more inclusive, we
characteristics, such as the removal of Ab plaques in AD and developed separate search strategies for AD and MS. Medical
altering the peripheral immune response in MS, but fails to fully Subject Heading search terms were entered in all fields of
cease disease progression (39, 40). Improving our understanding of the publication (e.g., abstract, title, keywords); the final search queries
complex relationship between neuroinflammation and are presented in Table 1. Title and abstract from retained papers
neurodegeneration in both disorders, which may have more in were subsequently evaluated for eligibility, with additional reading
common than hitherto thought, can be valuable to uncover of the entire text in case of non-exclusion. Studies were considered
therapeutic targets that may truly halt disease progression. suitable for inclusion and data collection only if they were written in
Notwithstanding the progress that recently has been made in the English, were performed in humans and investigated the
field of body fluid-based inflammatory biomarker research in AD and relationship between CSF inflammatory biomarkers and disease
MS, its implementation in the real-world clinical practice turns out to progression in AD and/or MS patients; the latter in accordance with
be a slow process. With this systematic review, we have listed all CSF the above described definition. Reviews, case reports and case series
inflammatory biomarkers showing an association with clinical disease were rejected; there were no other exclusion criteria.
progression in AD and/or MS patients, in an attempt to identify the Search and selection procedures were performed by a single
most promising candidates for further validation in clinical practice reviewer (J.T.); for flowchart see Figure 1. Title and abstract
and to gain a deeper understanding in the inflammatory pathways screening resulted in 165 candidate papers. Following full-text
underlying disease progression, which may ultimately culminate in evaluation, 84 papers (all published between 1991 and 2022)
ameliorated clinical care for patients affected by these chronic complied to our criteria and were included for data extraction
disorders. Moreover, we have found several methodological and and quality assessment. More than double the number of papers
conceptual commonalities in biomarker research for AD and MS, were found in the field of MS (n = 59) compared to AD (n = 25).
and summarize some general key points to enhance the quality of Uncertainties during the search were resolved by consensus after a
future efforts studying the relationship with disease progression. discussion involving all authors.

Frontiers in Immunology 03 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

TABLE 1 Search query for systematic review.

MS/AD Search query Results

(“CSF” OR “cerebrospinal fluid” OR “Biomarkers/cerebrospinal fluid”[Mesh]) AND (“multiple
sclerosis” OR “MS” OR “Multiple Sclerosis”[Mesh] OR “CIS” OR “clinically isolated syndrome” OR
“RIS” OR “radiologically isolated syndrome” OR “RRMS” OR “relapsing?remitting multiple sclerosis”
OR “SPMS” OR “secondary progressive multiple sclerosis” OR “PPMS” OR “primary progressive
MS multiple sclerosis” OR “Converters” OR “Non-converters”) AND (neuro?inflammat* OR 364
“Inflammation”[Mesh]) AND (“predictive value” OR “prognostic value” OR “progression” OR
“Predictive Value of Tests”[Mesh] OR “disease progression” OR “Disease Exacerbation” OR “Disease
Progression”[Mesh] OR “functional decline” OR “cognitive decline” OR “cognitive dysfunction” OR
PubMed “Cognitive Dysfunction”[Mesh] OR “neurodegeneration” OR “Nerve Degeneration”[Mesh])

AD (“CSF” OR “cerebrospinal fluid” OR “Biomarkers/cerebrospinal fluid”[Mesh]) AND (“alzheimer

disease” OR “alzheimer’s disease” OR “Alzheimer Disease”[Mesh] OR “mild cognitive impairment” OR
“MCI” OR “Prodromal” OR “Preclinical” OR “AD”) AND (neuro?inflammat* OR
“Inflammation”[Mesh]) AND (“predictive value” OR “prognostic value” OR “progression” OR 258
“Predictive Value of Tests”[Mesh] OR “disease progression” OR “Disease Exacerbation” OR “Disease
Progression”[Mesh] OR “functional decline” OR “cognitive decline” OR “cognitive dysfunction” OR
“Cognitive Dysfunction”[Mesh] OR “neurodegeneration” OR “Nerve Degeneration”[Mesh])

MS TS= (‘multiple sclerosis’ OR ‘MS’ OR ‘RRMS’ OR ‘relapsing?remitting multiple sclerosis’ OR ‘SPMS’

OR ‘secondary progressive multiple sclerosis’ OR ‘PPMS’ OR ‘primary progressive multiple sclerosis’
OR ‘CIS’ OR ‘clinically isolated syndrome’ OR ‘RIS’ OR ‘radiologically isolated syndrome’) AND TS= 491
(cerebrospinal fluid) AND TS=(‘prognostic value’ OR ‘disease progression’ OR ‘progression’ OR
Web of Science ‘predictive value’) AND TS=(inflam*)

AD TS=(‘alzheimer* disease’ OR ‘AD’ OR ‘MCI’ OR ‘mild cognitive impairment’ OR ‘Prodromal’ OR

‘Preclinical’) AND TS=(cerebrospinal fluid) AND TS=(‘prognostic value’ OR ‘disease progression’ OR 344
‘progression’ OR ‘predictive value’) AND TS=(inflam*)

AD, Alzheimer’s disease; MS, multiple sclerosis.

2.3 Data extraction from 50% to 80%, and above 80% of the potential maximum score
were categorized as “Low”, “Moderate” or “High” in quality,
Data extraction was performed in a standardized manner by J.T. respectively. The quality assessment for all included papers (n =
and M.B. and consisted of the following items: first author, year of 84) can be found in the Supplementary Material (Table S3).
publication, journal, study aim, characteristics of the studied Although most papers were of low quality, none were excluded
cohorts (diagnostic criteria, subgroups, age, clinical status), based on this criterion, in order to be able to provide a complete
neurochemical quality measures (assay, intra-assay coefficient of overview of the literature.
variation, limit of detection, detectability rate, blinding of the
researcher), inflammatory biomarkers of interest, observed
median and/or mean levels of biomarkers in the studied cohorts, 3 Results
measures of disease progression (with follow-up data in case of
longitudinal studies), statistical analyses and main findings. A 3.1 Inflammatory biomarkers related to
composite evidentiary Excel spreadsheet (Microsoft Corp, clinical status
Redmond, WA, USA) compiling all variables was created to
ensure reproducibility and completeness of the dataset. The 3.1.1 General overview
complete overview of the collected data can be found in the Of the 84 papers selected for data extraction, 43 focused on the
Supplementary Material (Table S2). cross-sectional relationship between inflammatory biomarkers and
clinically validated scores representing functional status and/or
cognition collected at baseline (AD/MS = 14/29). The Mini-
2.4 Quality assessment Mental State Examination (MMSE) and Expanded Disability
Status Scale (EDSS), which are measures of global cognitive
We developed a set of questions to systematically evaluate the impairment and general disability, respectively (43, 44), were the
methodological quality of each included study, focusing on most frequently used scales in AD (MMSE: 85.7%) and MS (EDSS:
diagnostic criteria, neurochemical analysis, clinical scoring, cohort 96.6%) studies. The vast majority of relationships were assessed
size, statistics, validation of findings and, in case of longitudinal using correlation analyses; a detailed overview of these results can
studies, duration of follow-up. Separate questions for AD and MS be found in the Supplementary Material (Table S4).
papers were formulated for aspects that were considered disease-
specific. The quality assessment was conducted independently by 3.1.2 AD literature
J.T. and M.B., and variations were discussed between both authors Cognitive testing in AD studies included MMSE, the Clinical
until a consensus was reached. Publications with a grade below 50%, Dementia Rating (CDR), the Alzheimer’s Disease Assessment Scale

Frontiers in Immunology 04 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

FIGURE 1
PRISMA 2020 flow chart diagram for systematic reviews. AD, Alzheimer’s disease; MS, multiple sclerosis. * From the 165 papers, 55 raised doubts for
inclusion after full-text assessment. These papers were double-checked by all the authors. In the end, 19 papers were included after reaching a
consensus, while the remaining 36 were excluded. Illustration adapted from Page et al., 2021 (42).

– Cognitive Subscale (ADAS-Cog), Dementia Rating Scale-2 (DRS- Disease Severity Scale, MS-DSS; Bayesian Risk Estimate for
2), the Frontal Assessment Battery (FAB), the Consortium to Multiple Sclerosis, BREMS) and fatigue (Fatigue Severity Scale,
Establish a Registry for Alzheimer’s Disease (CERAD) FSS) (54–59). The cohorts used for correlation analyses included
neuropsychological battery and the Montreal Cognitive between 16 and 244 subjects, but RR, SP and PP MS patients were
Assessment (MoCA) (45–50). The examined cohorts consisted often grouped together and considered as representative for the
either of MCI or dementia due to AD patients, with a sample size entire MS spectrum. Only 15 biomarkers demonstrated significant
varying from 19 to 163 individuals. clinical relevance, with the majority describing a positive yet weak
From the 33 inflammatory biomarkers that were studied, only correlation with one or more of the clinical scores (Table 2B). More
eight were significantly associated with cognitive impairment: specifically, anti-neurofilament light chain (NF-L) antibodies,
activated helper and cytotoxic T-cells, monocyte chemoattractant chitinase 3-like protein 1 (CHI3L1 or YKL-40), complement
protein-1 (MCP-1 or CCL2), C-X-C motif chemokine ligand 8 and factor 3 (C3), glial fibrillary acidic protein (GFAP) and tumor
10 (CXCL8 and CXCL10, respectively), interleukin-1 beta (IL-1b), necrosis factor alpha (TNF-a) were investigated by various
macrophage inflammatory protein-1 alpha (MIP-1a) and independent studies, yet none of these biomarkers consistently
osteopontin (OPN) (Table 2A). However, consistent results across showed significant results.
multiple studies were only found for OPN (which is a cytokine with
regulatory effects on the function of microglia (51)), demonstrating
an overall moderate correlation with baseline MMSE scores in 3.2 Inflammatory biomarkers related to
subjects with AD (52, 53). These findings were mainly based on changes in clinical status
cohorts with lower mean MMSE values and older age, likely
representative for a more advanced disease stage. 3.2.1 General overview
In total, 33 papers (AD/MS = 10/23) investigated the
3.1.3 MS literature relationship between inflammatory biomarkers and disease
In total, 93 inflammatory biomarkers were evaluated for cross- progression, with the latter based on the longitudinal evolution of
sectional associations with clinical scores, including measures of validated clinical scores similar to the ones used in the previous
general disability (EDSS), rate/severity of disability (Multiple section. Statistical methodology mainly relied on correlation
Sclerosis Severity Score, MSSS; Age-Related Multiple Sclerosis analyses and/or regression modelling for both disease entities, but
Severity, ARMSS; Progression Index, PI; Multiple Sclerosis we observed a notable distinction between the AD and MS field for

Frontiers in Immunology 05 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

TABLE 2A Overview inflammatory biomarkers significantly associated with baseline clinical scores in AD publications.

Biomarker Study Quality Cohort (n) Age (y) MMSE Clinical measure Correlation coefficient
Activated CD4+ T-cells Lueg, 2015 L AD (54) 68.7 21 Verbal learning -0.441
(n = 46)

Verbal retrieval -0.456

(n = 46)

Visuospatial skills -0.364

(n = 46)

Activated CD8+ T-cells Lueg, 2015 L MCI (19) 68.0 26 Verbal learning -0.545
(n = 15)

Verbal retrieval -0.298

(n = 15)

Visuospatial skills -0.523

(n = 15)

AD (54) 68.7 21 Verbal learning -0.371

(n = 46)

Verbal retrieval -0.744

(n = 46)

Visuospatial skills -0.334

(n = 46)

MCP-1 (CCL2) Kimura, 2018 M AD (69) 73.6 23 MMSE -0.245

FAB -0.306

Correa, 2011 L AD (22) 74.7 16 MMSE NR

Galimberti, 2006 L EOAD (22) 59.4 16 MMSE 0.350

LOAD (14) 72.4 19 (EO+LO)

CXCL8 (IL-8) Kimura, 2018 M AD (69) 73.6 23 FAB -0.293

CXCL10 Kimura, 2018 M AD (69) 73.6 23 MMSE -0.183

FAB -0.255

Correa, 2011 L AD (22) 74.7 16 MMSE NR

Galimberti, 2006 L EOAD (22) 59.4 16 MMSE 0.370

LOAD (14) 72.4 19 (EO + LO)

IL-1b Rui, 2021 M aMCI (33) 63.6 18 MMSE -0.357

MoCA -0.373

AD (33) 65.7 13 MMSE -0.486

MoCA -0.499

Rizzi, 2017 L aMCI (33) 68.0 NR CERAD 0.299

Hesse, 2016 L AD (NR) 68.0 22 MMSE -0.330

Tarkowski,2003 L MCI (56) 72.0 29 MMSE 0.460

(n = 6)

MIP-1b Taipa, 2019 M AD (32) 62.7 18 DRS-2 0.429

OPN Sun, 2013 M Newly AD (17) 73.0 23 MMSE 0.530

Chronic AD (18) 74.0 19 (Newly + Chronic)

Comi, 2010 L AD (67) 70.0 22 MMSE 0.580

AD, Alzheimer’s disease; (a)MCI, (amnestic) Mild Cognitive Impairment; CERAD, Consortium to Establish a Registry for Alzheimer’s disease neuropsychological battery; DRS-2, Dementia
Rating Scale-2; EO(AD), Early Onset Alzheimer’s disease; FAB, Frontal Assessment Battery; H, High quality; L, Low quality; LO(AD), Late Onset Alzheimer’s disease; M, Moderate quality;
MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment; NR, not reported. Significant p-values are marked in bold.

Frontiers in Immunology 06 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

the determination of disease progression. After all, in AD several neurodegenerative diseases (61)) showed promising results
publications, the majority of clinical scores were reported as a in more than one study (62, 63).
change in score over a given follow-up period (e.g., dMMSE defined
as the change in MMSE score from baseline to follow-up) with only 3.2.3 MS literature
a few reports of raw scores obtained throughout the study period, MS papers presented longitudinal associations for 29
whereas for MS the opposite was true with the majority of clinical inflammatory biomarkers with functional disability measured by
scores only being cross-sectionally collected at the time of follow-up clinical scoring instruments such as the EDSS, MSSS, PI and the
(e.g., EDSS score at follow-up) and merely a few studies reporting Multiple Sclerosis Functional Composite Scale (MSFC), but also
an actual change in scores over time. In case of prediction analyses, with cognitive performance, based on the Brief Repeatable Battery
a predefined threshold was used when determining the probability of Neuropsychological Tests (BRB) or Symbol Digit Modalities Test
of reaching a certain grade of disability with the inflammatory (SDMT) (64–66). Cohorts were once again mainly representing MS
biomarker as a predictor (e.g., the probability to reach an EDSS in general and included individuals with CIS, CIS that converted to
score of 4.0). Once again, both the MMSE and EDSS were the most clinically definite MS during the study, RR and/or progressive MS,
frequently used scores for AD (MMSE: 80.0%) and MS (EDSS: with a sample size ranging from 15 to 6,398 participants and a
87.0%), respectively. A detailed overview of all the results from the follow-up period between 1 and 10 years. The majority of
individual papers can be found in the Supplementary Material biomarkers (n = 20) showed significant results, though most were
(Table S5). evaluated in single studies. OCB, YKL-40 (a glycoprotein secreted
by glial cells (67)) and IL-1b (a cytokine involved in the regulation
3.2.2 AD literature of the innate immune response via microglia and astrocytes (68))
AD literature described 41 distinct inflammatory biomarkers were significantly associated with longitudinal measures of clinical
that were investigated for their relationship with disease status based on multiple publications and may reflect underlying
progression, represented by various clinical scores including the pathways driving disease progression (Table 3B).
MMSE, DRS-2, ADAS-Cog and Clinical Dementia Rating – Sum of
Boxes (CDR-SB), as well as composites of memory and/or executive
function (60). Cohorts consisted of patients with MCI, MCI due to 3.3 Inflammatory biomarkers related to
AD (based on having an AD biomarker profile and/or clinical AD transition to a more severe disease stage
diagnosis at follow-up; MCI-AD), and dementia due to AD. Sample
sizes varied from 32 up to 174 participants and follow-up periods 3.3.1 General overview
ranged from 9 months to 5 years. Similar to what we observed in Thirty (AD/MS = 8/22) of the 84 distinct publications
cross-sectional analyses, cohorts were mainly representative of the investigated the clinical relevance of inflammatory biomarkers
later stages of AD, characterized by lower MMSE scores and higher with regard to transitioning from one (possibly prodromal)
age. Twenty-two biomarkers were significantly associated with disease stage to a more severe disease stage. The progression rate
cognitive decline, many of which were also studied in the (i.e., the ratio of participants progressing over those remaining
previous section describing cross-sectional relationships stable) varied from 35 to 68% in AD publications and from 28 to
(Table 3A). Only MCP-1 (a crucial mediator of the innate 93% in MS reports. Numerous statistical methods were used for the
immune response that has been implicated in the pathology of subsequent data analyses, including between-group comparisons,

TABLE 2B Overview inflammatory biomarkers significantly associated with baseline clinical scores in MS publications.

Biomarker Study Quality Cohort (n) Age (y) EDSS Clinical measure Correlation
coefficient
15(S)-PGF2a / 20:4 µg Lam, 2016 L RRMS (23) 35.9 1.5 EDSS 0.313
SPMS (24) 49.4 5.8 (RR + SP + PP)
PPMS (15) 55 4.0

15(S)-PGF2a / CSF µg Lam, 2016 L RRMS (23) 35.9 1.5 EDSS 0.280
SPMS (24) 49.4 5.8 (RR + SP + PP)
PPMS (15) 55 4.0

Anti-NF-L antibodies Ehling, 2004 L MS (130) 42.1 3.7 EDSS NR

Silber, 2002 L RRMS (38) 29 NR EDSS 0.620

SPMS (18) 28 NR (RR + SP + PP: n = 49)
PPMS (10) 41 NR

Anti-NF-H antibodies Silber, 2002 L RRMS (38) 29 NR EDSS 0.510

SPMS (18) 28 NR (RR + SP + PP: n = 49)
PPMS (10) 41 NR

(Continued)

Frontiers in Immunology 07 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

TABLE 2B Continued

Biomarker Study Quality Cohort (n) Age (y) EDSS Clinical measure Correlation
coefficient
Anti-tubulin antibodies Silber, 2002 L RRMS (38) 29 NR EDSS 0.340
SPMS (18) 28 NR (RR + SP + PP: n = 49)
PPMS (10) 41 NR

B-cell / monocyte ratio Cepok, 2001 L RRMS (21) SPMS NR NR EDSS NR

(6) PPMS (4) NR NR (RR + SP + PP)
NR NR
PI 0.570
(RR + SP + PP)

PI 0.590
(RR + SP)

PI 0.630
(RR)

YKL-40 (CHI3L1) Huss, 20 20 M RRMS (47) 34 2.0 EDSS 0.140

PMS (39) 53 6.0 (RR)

EDSS -0.050
(P)

Gil-Perotin, 2019 M RRMS (99) 35 2.0 EDSS 0.210

SPMS (35) 45 5.5 (RR + SP + PP)
PPMS (23) 51 5.0

Novakova, 2017 M RRMS (59) 37 2.5 EDSS 0.274

MSSS 0.297

Complement factor C3 Aeinehband, 2017 L MS (48) 41.2 3.0 EDSS 0.170

MSSS 0.130

Sladkova, 2011 L CIS (20) 34.9 NR EDSS NR

GFAP Azzolini, 2022 M RRMS (51) 36.5 1.5 EDSS NR

BREMS 0.501

Abdelhak, 2018 H RRMS+ (24) 36 2.3 EDSS 0.400

RRMS- (18) 28 2.0 (RR + SP + PP)
SPMS (13) 50 6.5
PPMS (25) 53 4.5 EDSS NR
(RR)

EDSS NR
(SP+PP)

Novakova, 2017 M RRMS (59) 37 2.5 EDSS NR

MSSS NR

IgG oligoclonal bands Sladkova, 2011 L CIS (20) 34.9 NR EDSS “positive”
(OCB)

miR-142-3p De Vito, 2021 M MS (151) 39.5 2.0 EDSS NR

PI 0.270

sBCMA + IgG + IgG index Milstein, 2019 L RRMS (118) 42.3 NR MSSS 0.240
PMS (173) 55.3 NR (RR + P: n = 191)

MSSS 0.190
(RR: n = 71)

MSSS 0.260
(P: n = 120)

ARMSS 0.220
(RR + P: n = 190)

(Continued)

Frontiers in Immunology 08 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

TABLE 2B Continued

Biomarker Study Quality Cohort (n) Age (y) EDSS Clinical measure Correlation
coefficient
ARMSS 0.080
(RR: n = 70)

ARMSS 0.250
(P: n = 120)

MS-DSS 0.230
(RR + P: n = 244)

MS-DSS 0.070
(RR: n = 98)

MS-DSS 0.240
(P: n = 146)

sCD14 + sCD163 + YKL-40 Milstein, 2019 L RRMS (118) 42.3 NR MSSS 0.150
PMS (173) 55.3 NR (RR + P: n = 171)

ARMSS 0.220
(RR + P: n = 170)

MS-DSS 0.180
(RR + P: n = 211)

sCD27 Milstein, 2019 L RRMS (118) 42.3 NR MSSS 0.180

PMS (173) 55.3 NR (RR + P: n = 171)

MSSS 0.200
(RR: n = 63)

MSSS 0.230
(P: n = 108)

MS-DSS 0.200
(RR + P: n = 210)

MS-DSS 0.060
(RR: n = 86)

MS-DSS 0.240
(P: n = 124)

TNF-a Stampanoni-Bassi, L RRMS (205) 34.8 1.5 EDSS NR

2018

Malekzadeh, 2017 L RRMS (23) 45.2 2.0 EDSS NR

SPMS (22) (MS) 5.6 (RR + SP + PP)
PPMS (11) 5.0
FSS NR
(RR + SP + PP)

Obradovic, 2012 L MS (60) 43.5 3.8 EDSS NR

Sharief, 1991 L PPMS (32) 37.4 3.7 EDSS 0.834

(n = 17)

PI 0.741
(n = 17)

ARMSS, Age-Related Multiple Sclerosis Severity; BREMS, Bayesian Risk Estimate for Multiple Sclerosis; CIS, Clinically Isolated Syndrome; EDSS, Expanded Disability Status Scale; FSS, Fatigue
Severity Scale; H, High quality; L, Low quality; M, Moderate quality; MS, Multiple Sclerosis; MS-DSS, Multiple Sclerosis Disease Severity Scale; MSSS, Multiple Sclerosis Severity Score; NR, not
reported; PI, Progression Index; P(MS), Progressive Multiple Sclerosis; PP(MS), Primary Progressive Multiple Sclerosis; RR(MS), Relapsing-remitting Multiple Sclerosis; RR(MS)+, acute
exacerbation; RR(MS)-, remission; SP(MS), Secondary Progressive Multiple Sclerosis. Significant p-values are marked in bold.

prediction (such as Cox proportional hazard and regression 3.3.2 AD literature

modelling) and discrimination analyses with receiver operating Cohorts consisted of MCI subjects, with a sample size varying
characteristic measures. A comprehensive overview of all from 31 to 174 individuals and an observation period ranging from
statistical analyses and the corresponding results is provided in 9 months to 5 years. Sixteen inflammatory biomarkers were
the Supplementary Material (Table S6). investigated for their clinical relevance with regard to progression

Frontiers in Immunology 09 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

TABLE 3A Overview inflammatory biomarkers associated with longitudinal clinical measures in AD publications.

Biomarker Study Quality Cohort (n) Age (y) MMSE Associated clinical outcome measure Result
(follow-up time): statistical analysis statistical
analysis
MCP-1 Pillai, 2020 L MCI-AD (48) 68.1 24.8 dMMSE (9m)* b = -1.540
(CCL2)
dMMSE (15m)* b = -2.360

dCDR-SB (9m)* b = 1.230

dCDR-SB (15m) R = 0.540

(MCI-AD: n = 40)

dCDR-SB (15m)* b = 2.820

MCI (134) 74.9 26.9 dMMSE (12m)* b = -1.450

dMMSE (24m)* b = -1.460

dMMSE (36m)* b = -2.870

dCDR-SB (24m)* b = 0.860

dCDR-SB (36m)* b = 1.430

dCDR-SB (36m) R = 0.240

dCDR-SB (36m) R = 0.270

(MCI-AD: n = 97)

Westin, L MCI-AD (47) 74.0 26.7 Annual decline MMSE (60m) R = 0.420
2012
Annual decline MMSE (60m)° b = 0.390

CCL4 Pillai, 2020 L MCI (134) 74.9 26.9 dMMSE (36m)* b = -1.360

dCDR-SB (36m)* b = 0.760

CCL5 Pillai, 2020 L MCI (134) 74.9 26.9 dMMSE (24m)* b = -0.430

dCDR-SB (24m) R = 0.190

(MCI: n = 118)

dCDR-SB (24m)* b = 0.410

YKL-40 Kester, M MCI (61) 68.0 27.0 MMSE decline: predictor (32m) b = -0.320
(CHI3L1) 2015
AD (65) 65.0 22.0 MMSE decline: predictor (46m) b = 0.650

Complement Toledo, M MCI (163) 74.5 27.0 ADAS-Cog: C3xTime (42m) b = -0.120
factor C3 2014
AD (83) 74.8 23.7 ADAS-Cog: C3xTime (22m) b = -0.009

Factor H Toledo M MCI (163) 74.5 27.0 ADAS-Cog: FHxTime (42m) b = -0.075
(FH) 2014
AD (83) 74.8 23.7 ADAS-Cog: FHxTime (22m) b = -0.005

FGF basic Taipa, M AD (32) 62.7 18.2 dDRS-2 (12m) R = -0.664

2019

Fibrinogen Pillai, 2020 L MCI-AD (48) 68.1 24.8 dCDR-SB (15m)* b = -0.270
(FGA)
MCI (134) 74.9 26.9 dMMSE (12m)* b = 0.490

dCDR-SB (24m) R = -0.180

dCDR-SB (24m)* b = -0.230

Gas-6 Sainaghi, M AD (50) 65.0 22.0 MMSE decline (24m) R = -0.800

2017

G-CSF Taipa, M AD (32) 62.7 18.2 dDRS-2 (12m) R = -0.521

2019

GM-CSF Taipa, M AD (32) 62.7 18.2 dDRS-2 (12m) R = -0.479

2019

(Continued)

Frontiers in Immunology 10 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

TABLE 3A Continued

Biomarker Study Quality Cohort (n) Age (y) MMSE Associated clinical outcome measure Result
(follow-up time): statistical analysis statistical
analysis
Tarkowski, L MCI (56) 72.0 28.9 MMSE (9m) R = NR
2003 (n = 0)

IFN-g Taipa, M AD (32) 62.7 18.2 dDRS-2 (12m) R = -0.495

2019

IL-1b Taipa, M AD (32) 62.7 18.2 dDRS-2 (12m) R = -0.576

2019

Tarkowski, L MCI (56) 72.0 28.9 MMSE (9m) R = 0.340

2003 (n = 6)

dMMSE (9m) R = 0.300

(n = 6)

IL-4 Taipa, M AD (32) 62.7 18.2 dDRS-2 (12m) R = -0.507

2019

IL-6 Taipa, M AD (32) 62.7 18.2 dDRS-2 (12m) R = -0.553

2019

IL-9 Taipa, M AD (32) 62.7 18.2 dDRS-2 (12m) R = -0.578

2019

IL-17 Taipa, M AD (32) 62.7 18.2 dDRS-2 (12m) R = -0.577

2019

MIP-1b Taipa, M AD (32) 62.7 18.2 dDRS-2 (12m) R = -0.577

2019

MMP3 Pillai, 2020 L MCI-AD (48) 68.1 24.8 dMMSE (9m) R = 0.380
(MCI-AD: n = 40)

dMMSE (9m)* b = 2.050

dCDR-SB (9m) R = -0.36

(MCI-AD: n = 39)

dCDR-SB (9m)* b = -0.970

dCDR-SB (15m)* b = -0.940

MCI (134) 74.9 26.9 dCDR-SB (36m)* b = -0.530

sTNFR1 Hu, 2021 H MCI (174) 75.2 NR dCDR-SB (60m) b = -0.026

score
dCDR-SB (60m): sTNFR1 score x Months b = -0.020

Time to CDR > 4.0: High AD + High sTNFR1 scores Longer than
Low sTNFR1

Risk for CDR > 4.0: High AD + High sTNFR1 scores HR = 0.454

dADNI-Mem-EF (60m) b = -0.010

dADNI-Mem-EF (60m): sTNFR1 score x Months b = 0.005

sTREM2 Pillai, 2021 L MCI (67) 74.1 26.8 CDR-SB decline (60m): sTREM2 x Visit number b = 0.016

AD (42) 74.2 23.5 CDR-SB decline (60m): sTREM2 x Visit number b = -1.334

sTREM2 Hu, 2021 H AD (97) 75.1 NR dCDR-SB (36m) b = 0.004

score
dCDR-SB (36m): sTREM2 score x Months b = -0.040

Time to CDR-SB > 7.8 (36m): High AD + High Longer than

sTREM2 score Low sTREM2

Time to CDR-SB > 7.8 (36m): High p-Tau + High Longer than
sTREM2 score Low sTREM2

(Continued)

Frontiers in Immunology 11 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

TABLE 3A Continued

Biomarker Study Quality Cohort (n) Age (y) MMSE Associated clinical outcome measure Result
(follow-up time): statistical analysis statistical
analysis
Risk to CDR-SB > 7.8 (36m): High AD + High HR = 0.412
sTREM2 scores

dADNI-Mem-EF (36m) b = -0.027

dADNI-Mem-EF (36m): sTREM2 score x Months b = 0.006

b, regression coefficient; AD, Alzheimer’s disease; ADAS-Cog, Alzheimer’s Disease Assessment Scale – Cognitive Subscale; ADNI-Mem-EF, Average of composite Alzheimer’s Disease
Neuroimaging Initiative Memory and Executive Functioning Scores; CDR-SB, Clinical Dementia Rating – Sum of Boxes; DRS-2, Dementia Rating Scale-2; H, High quality; HR, Hazard ratio; L,
Low quality; m, months; M, Moderate quality; MCI, Mild Cognitive Impairment; MCI-AD, MCI with AD biomarkers and/or AD diagnosis at follow-up; MMSE, Mini-Mental State Examination;
n, number; R, Spearman or Pearson correlation coefficient. Significant p-values are marked in bold. *Multivariable analyses; °Multivariate analyses.

in the AD continuum; however, MCI cohorts showed or MS. YKL-40 was the only biomarker showing relevance for both
characteristics suggestive for late-onset AD, based on a relatively disorders. In addition, alterations in MCP-1 and OPN were
older age and higher MMSE value. Eight biomarkers demonstrated associated with clinical deterioration in patients with AD
significant relevance for patients that progressed from MCI to exclusively, while the same was true for IL-1b, TNF-a, CXCL13,
dementia due to AD, including MCP-1 (63), neuronal pentraxin 2 GFAP and OCB in subjects affected by MS. These combined results
(NPTX2), OPN, YKL-40 and several TNF receptors (with related point towards a prominent role for innate immune response in the
proteins that were also part of a composite score) (Table 4A). Only CNS, in which microglia are considered to be key mediators (72), in
YKL-40 was assessed in two independent studies but conflicting the progression of both disorders. However, we must acknowledge
results were reported (69, 70). Both of them showed similar cohort that the studies retrieved by our search were highly heterogenous in
characteristics, but the study with the highest progression rate (i.e., study design and methodology, largely complicating the
higher number of MCI patients that developed dementia due to interpretation of their findings. The ‘Standard for Reporting of
AD), what in itself could be due to the longer follow-up period (32.4 Diagnostic Accuracy’ initiative states: “To comprehend the results
versus 24 months), favored YKL-40 as a predictive biomarker (70). of diagnostic accuracy studies, readers must understand the study
Although OPN was only investigated in one study, it was the sole design, conduct, analysis, and results of such studies. That goal can
biomarker evaluated at two distinct time points within the same be achieved only through complete transparency from authors”
cohort. With this approach, authors could measure the intra- (73). This statement nicely summarizes the main difficulties
individual change of the CSF levels throughout the disease course. encountered when reviewing the literature. In the following
Results demonstrated increased levels in MCI patients that paragraphs, we sum up the most important findings for both
progressed to AD at follow-up, compared with those who disorders, as well as some recommendations that, in our vision,
remained stable (53). may contribute to an enhanced clinical relevance and improved
quality of result reporting when incorporated in future biomarker
3.3.3 MS literature studies in this field (Table 5).
The vast majority of publications collected from the MS Is neuroinflammation beneficial or detrimental in AD? To
literature investigated the transition from CIS to clinical definite answer this question, we need to take a closer look at the
MS, with cohort sizes varying from 18 to 139 participants and a complex biology and temporal behavior of microglia in this
follow-up duration between 10 months and 10 years. The clinical disorder. These residential CNS macrophages are responsible for
relevance of 62 distinct inflammatory biomarkers was assessed. Half maintaining local homeostasis and can elicit an inflammatory
of the studies presented significant results, yet only ten biomarkers response against potentially harmful agents, both endo- and
were evaluated in multiple studies (Table 4B). YKL-40 and CXCL13 exogenous. Microglia are highly dynamic cells that were
(a chemokine implicated in B cell aggregates that develop in the historically described as being activated into a pro- (M1) or anti-
inflamed meninges of PMS patients (71)) showed the most inflammatory phenotype (M2). In fact, these cells exists in various
consistent results, with increased CSF levels at baseline in CIS states depending on the specific context (age, spatial location,
subjects who received a MS diagnosis during follow-up, as environment) and are active in both health and disease. In line
compared with those who did not. These studies highlight the with this, neuroinflammation cannot be generalized as being either
importance of both glial cells as well as the innate immune response detrimental or beneficial, but will, also depending on the context,
in the early stages of the disease. exert adaptive or maladaptive effects (20, 74). A recent PET imaging
study described increased TSPO expression in MCI and dementia
due to AD patients, as compared with healthy controls. However, at
4 Discussion follow-up, a decreased expression in MCI and increased expression
in dementia due to AD was observed. These results prompted the
With this review, we provide an overview of the inflammatory two-peak hypothesis, stating that across the AD spectrum microglia
CSF biomarkers that are related to disease progression in AD and/ may undergo two major morphological changes, related with

Frontiers in Immunology 12 frontiersin.org

Frontiers in Immunology

Temmerman et al.
TABLE 3B Overview inflammatory biomarkers associated with longitudinal clinical measures in MS publications.

Biomarker Study Quality Cohort (n) Age (y) EDSS Associated clinical outcome measure Result statistical analysis
(follow-up time): statistical analysis
k-FLC Voortman, 2017 L CIS+RRMS (61) 28.8 2.0 dEDSS (57.6m) R = NR

Makshakov, 2015 L Converters (98) 32.0 NR EDSS (24m) R = 0.418

k-FLC / L -FLC ratio Rathbone, 2018 L CIS (43) 45.0 NR EDSS (60m) R = -0.370
RRMS (50) 40.5 NR (n = 29)

Voortman, 2017 L CIS+RRMS (61) 28.8 2.0 dEDSS (57.6my) R = NR

Adiponectin (Adp) Signoriello, 2021 M MS (66) 43.9 2.0 EDSS (55.2m): Adp > 9.91 mg/ml Higher than Adp < 9.91

EDSS increase (55.2m): Adp > 9.91 mg/ml ° OR = 0.62

MSSS (55.2m): Adp > 9.91 mg/ml Higher than Adp < 9.91

MSSS increase (55.2m): Adp > 9.91 mg/ml ° OR = 1.72*

PI (55.2m): Adp > 9.91 mg/ml Higher than Adp < 9.91

AF Decker, 2016 M MS (30) 31.2 NR Annual dEDSS (>12m) R = 0.810

BDNF Sarchielli, 2002 L SPMS (15) 38.4 3.7 EDSS increase (24m) Lower BDNF values
(n = 8)
13

CCL3 Puthenparempil, 2020 L RRMS (30) 35.7 NR Disease reactivation$(36.4m): CCL3 > 0.736 pg/mL OR = 4.9

YKL-40 (CHI3L1) Gil-Perotin, 2018 M RRMS (25) 35.0 2.0 EDSS increase: predictor * b = 1.097

HR = 2.996

Comabella, 2010 L Converters (48) 26.8 NR EDSS (12m) R = 0.340

(n = 44)

EDSS (24m) R = 0.400

(n = 39)

EDSS (36m) R = 0.470

(n = 41)

EDSS (48m) R = 0.380

(n = 37)

10.3389/fimmu.2023.1162340
EDSS (60m) R = 0.290
(n = 29)

CXCL8 (IL-8) Stampanoni-Bassi, 2018 L RRMS (150) NR NR EDSS (12m) R = NR

EDSS (36m) R = 0.242

frontiersin.org

PI (36m) R = 0.246

(Continued)
Frontiers in Immunology

Temmerman et al.
TABLE 3B Continued

Biomarker Study Quality Cohort (n) Age (y) EDSS Associated clinical outcome measure Result statistical analysis
(follow-up time): statistical analysis
CXCL12 Farina, 2017 M MS (90) 42.5 1.5 EDSS (120m post onset) R = 0.679
(n = 21)

GFAP Norgren, 2004 M RRMS (58) 34 1.5 EDSS (48m) R = 0.210

SPMS (21) 42 3.5 (RR + SP + PP + PR)
PPMS (15) 44 3
PRMS (5) 31 6 PI (48m) R = 0.240
(RR + SP + PP + PR)

GM-CSF Farina, 2017 M MS (90) 42.5 1.5 EDSS (120m post onset) R = 0.626
(n = 21)

IgG oligoclonal bands (OCB) Giedraitiene, 2021 M RRMS (49) 47.3 2.8 EDSS (60m): OCB+ Similar to OCB-

dSDMT(60m) = -3.1 – 1.0x(dRNFL_T) + 3.3xOCB R² = 0.599

dSDMT(60m) = -8.8 – 1.1x(dRNFL_PMB) + 4.4xOCB R² = 0.480

+ +
Karrenbauer, 2021 L MS OCB (6494) 38.1 NR Reach EDSS 3.0 (NR): OCB Similar to OCB-
MS OCB- (828) 40.8 NR
Risk EDSS 3.0 (NR): OCB+ HR = 1.29
(n = 5055)

Reach EDSS 4.0 (NR): OCB+ Similar to OCB-

14

Risk EDSS 4.0 (NR): OCB+ HR = 1.38

(n = 5802)

Reach EDSS 6.0 (NR): OCB+ Similar to OCB-

Risk EDSS 6.0 (NR): OCB+ HR = 1.20

(n = 6398)

Oechtering, 2021 L MS+CIS (530) 35.7 NR MSSS (61.2m): OCB+ vs. OCB- (n = 529) b = 0.860
+ - - -
MSSS (61.2m): OCB IgG IgM (n = 114) vs. OCB IgG IgM (n = 46) - -
b = 0.940
+ - - -
MSSS (61.2m): OCB IgG IgM vs. OCB IgG IgM ° - -
b = 0.730
+ + -
MSSS (61.2m): OCB IgG IgM (n = 228) vs. OCB IgG IgM - - -
b = 0.730

MSSS (61.2m): OCB+IgG+IgM- vs. OCB-IgG-IgM- ° b = 0.860

10.3389/fimmu.2023.1162340
MSSS (61.2m): OCB+IgG+IgM+ (n = 111) vs. OCB-IgG-IgM- b = 1.030
+ + + -
MSSS (61.2m): OCB IgG IgM vs. OCB IgG IgM ° - -
b = 1.110
+
Farina, 2017 M MS (90) 42.5 1.5 EDSS (120m post onset): OCB Higher than OCB-
frontiersin.org

(n = 21)

(Continued)
Frontiers in Immunology

Temmerman et al.
TABLE 3B Continued

Biomarker Study Quality Cohort (n) Age (y) EDSS Associated clinical outcome measure Result statistical analysis
(follow-up time): statistical analysis
BRB (120m post onset): OCB+ test failed Higher than OCB-
(n = 21)

Koch, 2007 L MS (143) 39.0 3.0 EDSS increase vs. stable (60m) Similar OCB+ / OCB- ratio

Similar OCB count

PPMS (50) 45.5 4.0 EDSS increase vs. stable (60m) Similar OCB+ / OCB- ratio

Similar OCB count

IgM oligoclonal bands (OCMB) Capuano, 2021 M RRMS (78) 32.9 2.0 Reach EDSS > 3.0 (115.2m) HR = 4.39
(n = 20)

Reach EDSS > 3.0 (115.2m)° HR = 2.93

(n = 20)

Reach EDSS > 4.0 (115.2m) HR = 5.39

(n = 8)

IL-1b Farina, 2017 M MS (90) 42.5 1.5 EDSS (120m post onset) R = 0.625
(n = 21)

EDSS (60m): IL-1b+ Higher than IL-1b-

15

Rossi, 2014 L RRMS (170) 36.3 2.2

EDSS increase (60m): %RRMS IL-1b+ Higher than IL-1b-

Reach EDSS 3.0 (60m): IL-1b+ OR = 4.34

+
Reach EDSS > 3.0 (60m): IL-1b OR = 3.38
+
Reach EDSS 4.0 (60m): IL-1b OR = 4.12
+
Reach EDSS > 4.0 (60m): IL-1b OR = 3.32

MSSS (60m): IL-1b+ Higher than IL-1b-

MSFC worsening (60m): %RRMS IL-1b+ Higher than IL-1b-

Disability progression based on MSFC (60M): IL-1b+ OR = 2.13

+
MSFC worsening (60M): IL-1b OR = 2.21

10.3389/fimmu.2023.1162340
+
PI increase (60m): IL-1b Higher than IL-1b-

BREMS (60m): IL-1b+ OR = 26.61

Rossi, 2014 L RRMS (170) 36.3 2.2 MSSS (48m): IL-1b+ Higher than IL-1b-
frontiersin.org

PI (48m): IL-1b+ Higher than IL-1b-

(Continued)
Frontiers in Immunology

Temmerman et al.
TABLE 3B Continued

Biomarker Study Quality Cohort (n) Age (y) EDSS Associated clinical outcome measure Result statistical analysis
(follow-up time): statistical analysis
IL-6 Stampanoni-Bassi, 2018 L RRMS (150) NR NR EDSS (12m) R = NR

EDSS (24m) R = 0.194

EDSS (36m) R = 0.274

PI (36m) R = 0.311

IL-9 Ruocco, 2015 M RRMS High-IL9 (50) 32.2 1.3 EDSS (48m): IL-9 > 100 pg/ml Lower than IL-9 < 100
RRMS Low IL9 (57) 31.8 2.0
Reach EDSS 4.0 (48m): IL-9 > 100 pg/ml OR = 0.24

MSSS (48m): IL-9 > 100 pg/ml Lower than IL-9 < 100

PI (48m): IL-9 > 100 pg/ml Lower than IL-9 < 100

IL-17 Ruocco, 2015 M RRMS Low IL9 (57) 31.8 2.0 Increased PI (48m): IL-17+ Higher than IL17-

miR-142-3p Mandolesi, 2017 L Active RRMS (18) 32.1 1.8 PI (24-60m) R = 0.500
Nonactive RRMS (12) 43.2 1.7 (n = 21)

Quotient k-FLC Makshakov, 2015 L Converters (98) 32.0 NR EDSS (24m) R = 0.410

TNF-a Sharief, 1991 L PMS (17) 37.4 3.7 EDSS (24m) R = 0.873
16

PI (24m) R = 0.851

b, regression coefficient; BREMS, Bayesian Risk Estimate for Multiple Sclerosis; CIS, Clinically Isolated Syndrome; Converters, CIS patients whom received a MS diagnosis; EDSS, Expanded Disability Status Scale; H, High quality; HR, Hazard ratio; L, Low quality; m, months;
M, Moderate quality; MS, Multiple Sclerosis; MSFC, Multiple Sclerosis Functional Composite; MSSS, Multiple Sclerosis Severity Score; NR, Not reported; OR, Odds ratio; PI, Progression Index; P(MS), Progressive Multiple Sclerosis; PP(MS), Primary Progressive Multiple
Sclerosis; PR(MS), Progressive Relapsing Multiple Sclerosis; RNFL_PMB, Papillomacular bundle retinal nerve fibre layer thickness; R, Spearman or Pearson correlation coefficient; RNFL_T, Temporal segment retinal nerve fiber layer thickness; RR(MS), Relapsing-Remitting
Multiple Sclerosis; SDMT, Symbol Digit Modalities Test; SP(MS), Secondary Progressive Multiple Sclerosis. Significant p-values are marked in bold; *Multivariable analyses; °Multivariate analyses; $ Not specified whether clinical or radiological disease reactivation.

10.3389/fimmu.2023.1162340
frontiersin.org
Frontiers in Immunology

Temmerman et al.
TABLE 4A Overview inflammatory biomarkers associated with progression across the AD spectrum.

Biomarker Study Quality Cohort (n) Age (y) MMSE Follow-up cohorts Associated clinical outcome measure Result
(follow-up time): statistical analysis
MCP-1 (CCL2) Westin, 2012 L MCI (119) 74.0 26.7 MCI-AD (47) Time to progression MCI-AD (60m): MCP-1 > 757 pg/ml Shorter than MCP-1 < 757 pg/ml
64.0 27.3 sMCI (52)
Time to progression MCI-AD (60m): MCP-1° Shorter

YKL-40 Swanson, L MCI (135) 74.6 74.8 26.9 MCI-AD (47) BL CSF levels in MCI-AD (24m) Similar to sMCI
(CHI3L1) 2016 27.0 sMCI (82)

Kester, 2015 M MCI (53) 70.0 64.0 26.0 MCI-AD (36) BL CSF levels in MCI-AD (32.4m) Higher than sMCI
28.0 sMCI (17)
Risk of progression (32.4my) HR = 1.003

NPTX2 Swanson, L MCI (135) 74.6 74.8 26.9 MCI-AD (47) BL CSF levels in MCI-AD (24m) Lower than sMCI
2016 27.0 sMCI (82)

OPN Sun, 2013 M MCI (31) 72.0 27.9 MCI-AD (13) BL CSF levels in MCI-AD (36m) Similar to sMCI
73.0 27.1 sMCI (18)
BL CSF levels in MCI-AD (36m) Lower than FUP CSF levels MCI-
AD

sTNFR1 score Hu, 2021 H MCI (174) 75.2 NR MCI-AD (99) Risk of progression (60m): High-AD + High sTNFR1 score HR = 0.541
sMCI (75)
Time to progression (60m): High-AD + High sTNFR1 score Longer than Low sTNFR1 score
181
ysTNFR1 score Hu, 2021 H MCI (174) 75.2 NR MCI-AD (99) Likelihood progression (60m): High-pTau + High ysTNRF1 Lower than Low ysTNFR1 score
17

sMCI (75) score

MCI (49) 69.3 NR MCI-AD (18) Likelihood progression (43m): High-pTau181 + High ysTNRF1 Lower than Low ysTNFR1 score
sMCI (31) score

TNFR1 Zhao, 2020 L MCI (116) 74.2 26.8 MCI-AD (64) Progression free survival time (30.2m): High TNFR1$ Shorter than Low TNFR1
sMCI (52)
- $
Progression free survival time (30.2m): TN , High TNFR1 Shorter than Low TNFR1

Progression free survival time (30.2m): TN+, High TNFR1$ Similar to Low TNFR1

TNFR2 Zhao, 2020 L MCI (116) 74.2 26.8 MCI-AD (64) Progression free survival time (30.2m): High TNFR2$ Longer than Low TNFR2
sMCI (52)
Progression free survival time (30.2m): TN-, High TNFR2$ Longer than Low TNFR2
+ $
Progression free survival time (30.2m): TN , High TNFR2 Similar to Low TNFR2

AD, Alzheimer’s Disease; BL, Baseline; CSF, Cerebrospinal fluid; FUP, Follow-up; H, High quality; HR, Hazard Ratio; L, Low quality; m, months; M, Moderate quality; MCI, Mild Cognitive Impairment; MCI-AD, MCI with AD diagnosis at follow-up; sMCI, stable MCI at

10.3389/fimmu.2023.1162340
follow-up. TN, Tau pathology. Significant p-values are marked in bold; °Multivariate analyses; $ Cut-off values not reported.
frontiersin.org
Frontiers in Immunology

Temmerman et al.
TABLE 4B Overview inflammatory biomarkers investigated in multiple studies and found to be associated with progression across the MS spectrum.

Biomarker Study Quality Cohort (n) Age (y) EDSS Follow-up Associated clinical outcome measure Result
cohorts (follow-up time): statistical analysis
k-FLC Voortman, 2017 L CIS (48) NR NR Converters (23) BL CSF levels in Converters (57.6m) Similar to Non-converters
Non-converters (25)

Makshakov, 2015 L CIS (139) 32.0 NR Converters (98) BL CSF levels in Converters (24m) Higher than Non-converters
33.0 NR Non-converters (41)

k-FLC index Vecchio, 2020 M RIS+CIS (18) 36.3 NR Converters (6) BL CSF levels in Converters (43.2m) Higher than Non-converters
Non-converters (12)
Risk of conversion (43.2m) HR = 1.070

Gaetani, 2020 L CIS (23) 41.8 1.7 Converters (12) Time to conversion (39.1m): k-FLC index > 10.6 C-statistic = 0.630
Non-converters (7) (n = 19)

k-FLC / L-FLC ratio Rathbone, 2018 L CIS (43) 45.0 NR Converters (NR) BL CSF levels in Converters (60m) Similar to Non-converters
Non-converters (NR) (n = 29)

Voortman, 2017 L CIS (48) NR NR Converters (23) BL CSF levels in Converters (57.6m) Lower than Non-converters
Non-converters (25)
Risk of conversion 57.6m): k / L ratio ≤ 3.38 HR = 2.890

Likelihood of conversion (57.6m): k / L ratio ≤ OR = 4.860

3.38
18

L-FLC Voortman, 2017 L CIS (48) NR NR Converters (23) BL CSF levels in Converters (57.6m) Similar to Non-converters
Non-converters (25)

Makshakov, 2015 L CIS (139) 32.0 NR Converters (98) BL CSF levels in Converters (24m) Higher than Non-converters
33.0 NR Non-converters (41)

YKL-40 (CHI3L1) De Fino, 2019 L CIS (25) 37.4 1.3 Converters (10) BL CSF levels in Converters (24m) Higher than Non-converters
Non-converters (14) (n = 24)

ROC (24m): Converters vs. Non-converters AUC = 0.790

(n = 24)

Gil-Perotin, 2019 M RRMS (99) 35.0 2.0 Converters (14) Risk of conversion (52.8m): SPMS ° HR = 18.040
Non-converters (85)

Thouvenot, 2019 L RIS (71) 38.0 NR Converters (20) BL CSF levels in Converters (16m) Similar to Non-converters
Non-converters (51)
Risk of conversion (16m): YKL-40 > 154.6 mg/ml HR = 2.130

10.3389/fimmu.2023.1162340
Risk of conversion (16m): YKL-40 > 154.6 mg/ml ° HR = 1.560

Borras, 2016 L CIS (50) 33.0 NR Converters (25) BL CSF levels in Converters (49.2m) Higher than Non-converters
38.0 NR Non-converters (25)
frontiersin.org

Comabella, 2010 L CIS (84) 26.8 NR Converters (48) BL CSF levels in Converters (66m) Higher than Non-converters
28.7 NR Non-converters (36)

(Continued)
Frontiers in Immunology

Temmerman et al.
TABLE 4B Continued

Biomarker Study Quality Cohort (n) Age (y) EDSS Follow-up Associated clinical outcome measure Result
cohorts (follow-up time): statistical analysis
Time to conversion (66m): YKL-40 > 287.9 mg/ml Shorter than YKL-40 < 287.9 mg/
ml

Risk of conversion (66m): YKL-40 > 287.9 mg/ml HR = 2.500

CIS (52) 33.6 NR Converters (26) BL CSF levels in Converters (58.8m) Higher than Non-converters
36.5 NR Non-converters (26)

CXCL13 Olesen, 2019 L ON (40) 41.5 NR Converters (16) BL CSF levels in Converters (28.2m) Higher than Non-converters
39.3 NR Non-converters (24)
Likelihood of conversion (28.2m): CXCL13 > 37 OR = 1.050
pg/ml

Ferraro, 2015 L CIS (110) 35.0 2.0 Converters (94) BL CSF levels in Converters (40m) Higher than Non-converters
Non-converters (16)
Risk of conversion (40m): CXCL13 ≥ 15.4 pg/ml HR = 2.900

Risk of conversion (40m): CXCL13 ≥ 15.4 pg/ml ° HR = NR

Brettschneider, L CIS (91) 34.0 2.0 Converters (45) BL CSF levels in Converters (24m) Higher than Non-converters
2010 Non-converters (46)

IgG index Olesen, 2019 L ON (40) 41.5 NR Converters (16) BL CSF levels in Converters (28.2m) Higher than Non-converters
19

39.3 NR Non-converters (24)

Likelihood of conversion (28.2m): IgG index > 0.64 OR = 11.090

Cinar, 2018 L CIS (41) NR NR Converters (35) Risk of conversion (12.8m): IgG index > 0.7 HR = 0.990
Non-converters (6)
Time to conversion (12.8m): IgG index > 0.7 Longer than IgG index < 0.7
+
IgG oligoclonal bands Gaetani, 2020 L CIS (19) 41.8 1.7 Converters (12) Time to conversion (39.1m): OCB Similar to OCB-
(OCB) Non-converters (7)

Kolcava, 2020 L CIS (64) 36.5 2.0 Converters (45) BL CSF levels in Converters (27m) Higher than Non-converters
Non-converters (19)
Risk of conversion (27m): OCB+ HR = 2.898

Risk of conversion (27m): OCB+ ° HR = 2.348

+
Likelihood of conversion (27m): OCB OR = 8.306
+
Likelihood of conversion (27m): OCB ° OR = 26.599

10.3389/fimmu.2023.1162340
Olesen, 2019 L ON (40) 41.5 NR Converters (16) BL CSF levels in Converters (28.2m): Higher than Non-converters
39.3 NR Non-converters (24)
+
Likelihood of conversion (28.2m): OCB OR = 13.000

Cinar, 2018 L CIS (41) NR NR Converters (35) Time to conversion (12.8m): OCB+ Shorter than OCB-
frontiersin.org

Non-converters (6) (n = 26)

(Continued)
Frontiers in Immunology

Temmerman et al.
TABLE 4B Continued

Biomarker Study Quality Cohort (n) Age (y) EDSS Follow-up Associated clinical outcome measure Result
cohorts (follow-up time): statistical analysis
Risk of conversion (12.8m): OCB+ HR = 2.160
(n = 26)

Farina, 2017 M RRMS (90) 42.5 1.5 Converters (19) BL CSF levels in Converters (120m) Higher than Non-converters
Non-converters (71)
+
Time to conversion (120m): OCB Shorter than OCB-

Brettschneider, L CIS (91) 34.0 2.0 Converters (45) Truth table (24m): Converters vs. Non-converters Sensitivity = 0.91
2010 Non-converters (46) Specificity = 0.36
PPV = 0.59
NPV = 0.81

Koch, 2007 L RRMS (78) 34.0 2.0 Converters (19) BL CSF levels in Converters (60M) Similar to Non-converters
Non-converters (59)

Tumani, 1998 L ON (36) 32.3 NR Converters (18) Truth table (48m): Converters vs. Non-converters Sensitivity = 0.83
Non-converters (18) Specificity = 0.33
PPV = 0.56
NPV = 0.67

OPN De Fino, 2019 L CIS (24) 37.4 1.3 Converters (10) BL CSF levels in Converters (24m) Similar to Non-converters
Non-converters (14)
ROC (24m): Converters vs. Non-converters AUC = 0.72
20

Borras, 2016 L CIS (50) 33.0 NR Converters (25) BL CSF levels in Converters (49.2m) Similar to Non-converters
38.0 NR Non-converters (25)

Positive CSF$ Thouvenot, 2019 L RIS (71) 38.0 NR Converters (20) BL CSF levels in Converters (16m) Higher than Non-converters
Non-converters (51)
Risk of conversion (16m) HR = 2.900
(n = 70)

Risk of conversion (16m) ° HR = 1.220

(n = 65)

Cinar, 2018 L CIS (41) NR NR Converters (35) BL CSF levels in Converters (12.8m) Higher than in Non-converters
Non-converters (6) (n = 26)

AUC, Area Under the Curve; BL, Baseline; CIS, Clinically Isolated Syndrome; Converters, ON/CIS/RIS subjects whom received MS diagnosis or RRMS subjects whom received SPMS diagnosis; CSF, Cerebrospinal fluid; H, High quality; HR, Hazard Ratio; L, Low quality; M,
Moderate quality; MS, Multiple Sclerosis; NPV, Negative Predictive Value; Non-converters, stable ON/CIS/RIS/RRMS subjects; NR, Not reported; ON, Acute Optic Neuritis; OR, Odds Ratio; PPV, Positive Predictive Value; RIS, Radiologically Isolated Syndrome; ROC,
Receiver Operator Characteristics; RRMS, Relapsing Remitting Multiple Sclerosis. Significant p-values are marked in bold. °Multivariate analyses; $ Defined as presence of OCB and/or IgG index > 0.7.

10.3389/fimmu.2023.1162340
frontiersin.org
Temmerman et al. 10.3389/fimmu.2023.1162340

TABLE 5 Recommendations for future research, based on the 'Standard for Reporting of Diagnostic Accuracy’ initiative for diagnostic accuracy tests.

Topic Items Relevance

MCI characterization MCI is a complex disease entity characterized by different subtypes and not only indicative for AD
AD Cohort
AD biomarker confirmation Verification of AD pathology based on well-established cut-offs for Ab, T-tau and P-Tau.

MS subtypes separated Due to important differences between CIS, RR, SP and PP MS with regard to inflammation

Relapsing versus Remitting Due to important differences between relapse/remission and active/non-active with regard to
MS Cohort
Active versus Non-active inflammation

Influence of DMT Due to the altering effect that may affect the interpretation of inflammatory biomarkers

Neurochemical Transparent reporting quality

To ensure reliability of results based on proper analyses with the selected assay
analyses measures

Statistical analyses Clinical versus statistical significance Ensure clinical relevance of the results or translate the results to clinical context

Longitudinal studies Appropriate follow-up duration Including repeated clinical evaluation and diagnostic confirmation of the disease

Ab, Amyloid-beta; AD, Alzheimer’s disease; CIS, Clinically Isolated Syndrome; DMT, Disease-modifying treatment; MCI, Mild Cognitive Impairment; MS, Multiple Sclerosis; RR(MS), Relapsing
Remitting Multiple Sclerosis; SP(MS), Secondary Progressive Multiple Sclerosis; PP(MS), Primary Progressive Multiple Sclerosis; P-tau, Phospho-tau; T-tau, Total-tau.

changes in microglia function (75). This hypothesis may explain the microglia across the AD spectrum and motivate a continued effort
distinction in microglia states between MCI and dementia due to in studying the earliest stages of AD.
AD observed at follow-up, despite both of them showing increased In MS, there seems to be a notable distinction between the types
TSPO expression when compared with healthy controls. However, of inflammation predominantly underlying the relapsing versus
increased TSPO expression in the diseased brain is considered a progressive phase of the disease. Whereas the former is mostly
neuroinflammatory marker, not a microglia specific marker (19). driven by recurrent invasion of peripheral immune cells into the
Furthermore, equating microglia morphology with function is CNS with formation of classic active demyelinating lesions, the
considered wrong, as different morphologies can execute similar latter has been associated with seemingly more subtle yet
functions (20). We found promising results for YKL-40, MCP-1 and chronically perseverating and compartmentalized immune-related
OPN in reflecting disease progression, particularly at the time of phenomena, such as (a) meningeal collections of lymphoid cells,
transition from the MCI to the dementia stage in the AD some of which are organized into follicle-like structures, with
continuum (53, 63, 70). These proteins are mainly produced by underlying subpial cortical demyelination and atrophy, (b)
microglia (51, 76, 77), and have shown increased expression near smoldering lesions characterized by a border zone of iron-rich
amyloid plaques or in the cytoplasm of AD brains (78, 79). MCP-1 MIMS and (c) widespread microglial inflammation and astrogliosis
(which is the most potent activator of the CCR2 receptor) and OPN throughout the NAWM, sometimes also referred to as ‘dirty white
have been implicated in several neurodegenerative diseases through matter’ (84, 85). Smoldering lesions can be visualized in vivo by
their respective role as chemoattractant and regulator of microglia susceptibility-weighted MRI as paramagnetic rim lesions and were
and DAMs (77, 80). In vivo studies using AD mouse models suggest found to be more abundant in patients with progressive MS (86),
that the MCP-1/CCR2 axis has varying effects on AD pathogenesis correlating with concurrent and forthcoming clinical disability and
across the stage of the disease. In the early stages, it may enhance brain volume loss (87–90). In addition, such lesions have been
accumulation of microglia and clearance of Ab, whereas in the late linked to (presumably intracellular) sodium accumulation and
stage, it may cause exaggerated inflammatory responses and the neurofilament release (91, 92), which are considered as very early
development of insoluble Ab within microglia (81). Perivascular pathobiological features of the neurodegenerative cascade in MS
macrophages (PVMs) are the main source of increased OPN in AD (93). Recent data have demonstrated that ongoing inflammation in
mouse models and postmortem AD brains, and seem to regulate smoldering lesions is mediated by a microglial-astrocytic axis in
phagocytic activity of microglia. Moreover, the notion that the which complement component C1q seems to be crucial for MIMS
timing of microglial phagocytosis across the disease stages may be activation (26). In general, T- and B-cells do not seem to be directly
crucial to determine whether the outcome will be beneficial or causing neuronal damage in MS but interact with local microglia,
harmful has been lifted forward recently (82). As the CNS immune hereby mediating tissue injury via a complex cascade that is still not
response seemingly changes depending on the stage of the disease, fully understood (94, 95). At the histological level, microglia
this may in part explain the inconsistency we found in our results as account for approximately 40% of the phagocytic cell population
biomarker findings would then highly depend on the selected in classic active lesions (96), while slowly expanding chronic lesions
cohort and microglia state at the time of sampling. Moreover, demonstrate a preferential accumulation of MIMS at the lesion’s
worsening of AD pathology may not necessarily follow a linear edge (97). In fact, it is hypothesized that “true MS” is best
course (83), further complicating the interpretation of biomarker represented by such chronic inflammatory state, since relapses
results. Nonetheless, together these observations point towards a (representative of focal inflammation) do not predict disease
complex and dynamic role of the innate immune response and outcome in untreated patients (98), and confirmed disability

Frontiers in Immunology 21 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

worsening occurring independently from relapse and/or MRI We were unable to perform a meta-analysis due to the
activity has frequently been described in cohorts of RR MS (99– heterogeneous nature of the included studies. The observed
101). Microglia seem to be characterized by a similar expression diversity in AD and MS study designs is in agreement with other
pattern in active and chronic lesions (95), suggesting a parallel reviews (15, 115, 116) and long known to be questionable for
function in both types. Since there are less active lesions during evaluating candidate biomarkers. Cohort discrepancies were one of
progressive MS, microglia become more prominent compared with the major limiting factors for the interpretability and
peripheral immune cell infiltration (102), as the disease evolves. reproducibility of the findings yielded by our search. In AD
Beside the association between increased TSPO expression in the studies, we observed a high variety in age and disease stage of
NAWM and an enhanced likelihood of experiencing progression participants at the time of inclusion. MCI cohorts, for example,
independent of relapse activity, TSPO is also associated with higher were rarely described by subtype (e.g., amnestic versus non-
EDSS and MSSS scores and found to be more common in SP amnestic, single- versus multiple-domain, etc.), by follow-up
compared with RR MS (30, 103). Interestingly, recent work using diagnosis (or even lacking follow-up) or by verification of AD
this technique showed that TSPO expression was most intense in pathology via CSF biomarkers (117). In MS papers, individuals were
the periventricular region, suggesting a diffusion of inflammatory clearly subdivided based on established diagnostic criteria but
CSF-derived factors into surrounding tissues (104), while the eventually often grouped together as “general MS” for analytical
thalamic activation pattern appeared to be more sensitive than purposes. Inflammation is highly fluctuating and/or situational in
conventional grey matter atrophy measures in predicting future MS, even in the RR group alone as, for instance, increased levels of
clinical disability worsening (105). These findings, together with our GFAP were found in a cohort of MS patients (compared with
results, support the hypothesis of microglia driving disease controls) but once you start to depict the subgroups you may find
progression across the entire MS spectrum. that the results are due to an increase at relapse in RR MS, as
Due to the role microglia seem to be playing in disease compared with stable or even progressive MS (118). The effects of
progression in AD and MS, it is not surprising that they are DMT (119) may also influence the interpretation of CSF
considered as interesting targets for future treatments. One inflammatory biomarkers with regards to pathophysiological
possible approach could be to change the microglia state in one mechanisms such as relation to disease progression. We generally
that clears debris and drives regeneration. However, as already observed a lack of longitudinal studies with appropriate time of
mentioned, microglia are highly versatile depending on the specific observation and periodic clinical evaluation, as well as diagnostic
context and it is likely that the environment, not the state, is the confirmation after extended follow-up in both AD and MS papers.
determining factor in deciding how microglia act in disease (20) Moreover, the majority of included studies did not report on the
Nonetheless, attempts are being made to alter the innate immune analytical quality, including the intra- and inter-assay coefficient of
response in AD and MS through microglia. Bruton’s tyrosine kinase variance, upper and lower limit of quantification and detectability
(BTK) is a non-receptor cytosolic signal transducing enzyme rate. Similarly, statistical analyses generally lacked reports of model
predominantly expressed by B-lymphocytes and myeloid cells. Its assumptions (e.g., visual inspection of QQ-plots and residuals for
downstream signaling pathways are involved in multiple processes, linear regression models) and approaches for regression analyses
including cellular growth, metabolism, proliferation and were highly heterogenous. Some biomarkers were based on
differentiation (106). Increased activity patterns have been computations that showed statistically significant relevance
observed post-mortem in AD brain and focal lesions of patients concerning disease progression but might be too complex for the
with progressive MS (107, 108). BTK inhibition has been found to routine clinical setting. It will be important to make a distinction
reduce phagocytic activity of microglia in preclinical AD models but between statistical and clinical significance when looking for
did not have a major impact on cytokine release (107). Further clinically relevant biomarkers and to translate outcomes into
research is required to truly determine its potential within the AD interpretable results for clinical use. The abovementioned aspects
continuum. On the other hand, BTK inhibition is a booming contributed to the limited quality we observed in a substantial
business in the MS field with several oral agents currently tested number of papers based on our quality assessment tool (see Table
in clinical trials (109). For example, a recent phase IIb study has S3). Additionally, there is also an important caveat with regard to
demonstrated a significantly decreased lesional accumulation on our own approach for this systematic review. Disease progression
conventional brain MRI after 12 weeks of treatment with was defined only by validated clinical scores, not by other measures
tolebrutinib, a selective BTK inhibitor for which evidence of BBB such as imaging parameters (e.g., MRI brain volume changes)
penetration exists in animals and humans, as compared with which will have influenced the selection of papers. However, we
placebo (110). Multiple phase III trials with this agent are wanted to ensure the findings from this manuscript were relevant
ongoing for the moment in relapsing, non-active SP and PP MS, for the clinical practice, therefore we only included those measures
and their results will be eagerly anticipated by the scientific that were examined in this setting. The EDSS score was confirmed
community. Interestingly, glatiramere acetate and dimethyl as the most frequently used measure of clinical disability (appearing
fumarate, which are commonly used as DMT for MS, have shown in 96.6% of the MS papers) but is also known to be a rather
therapeutic promise for AD and other models of neurocognitive insensitive outcome, especially in older patients with a higher grade
decline, at least partially attributed to effects on CNS microglia of disability (e.g., on average merely showing an increase of 1 point
(111–114). over 10 years in the natural history of MS) (120), which may be a

Frontiers in Immunology 22 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

reason why most of the biomarkers failed to demonstrate project. All authors contributed to the article and approved the
clinical relevance. submitted version.

5 Conclusion Acknowledgments
YKL-40 was identified as the only CSF-based biomarker related
We would like to thank Odile Chevalley, Carl Verduyn and
to clinical disease progression in both AD and MS. MCP-1 and
Melissa Wittens for assisting us throughout this review.
OPN showed similar relevance in patients with AD, while the same
accounts for IL-1b, TNF-a, CXCL13, GFAP and OCB in subjects
affected by MS. Our findings support the increasingly recognized
role of the innate immune system, including the microglia, in the
Conflict of interest
pathology of both disorders. However, the retrieved biomarker
The authors declare that the research was conducted in the
studies for AD and MS were generally heterogeneous in nature,
absence of any commercial or financial relationships that could be
challenging the interpretational reliability of their results. We have
construed as a potential conflict of interest.
framed some suggestions to tackle this problem in future research
(see Table 5), focusing on better cohort characterization (e.g., the
use of highly qualitative cohorts with biomarker confirmation for
AD studies, particularly for the MCI stage, and clear separation of
Publisher’s note
MS subtypes, as well as a distinct grouping of remitting and
All claims expressed in this article are solely those of the authors
relapsing MS patients), transparency in neurochemical quality
and do not necessarily represent those of their affiliated
measures, interpretable statistical analyses and longitudinal
organizations, or those of the publisher, the editors and the
studies with repeated clinical assessments and diagnostic
reviewers. Any product that may be evaluated in this article, or
confirmation during follow-up.
claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher.

Author contributions
JT, SE, MB and MD contributed to conception and design of the Supplementary material
study. JT performed the systematic search and organized the
database. JT and MB performed the quality assessment and The Supplementary Material for this article can be found online at:
analyses. JT wrote the first draft of the manuscript. MB and MD https://www.frontiersin.org/articles/10.3389/fimmu.2023.1162340/
wrote sections of the manuscript. MB, SE and MD supervised the full#supplementary-material

References
1. Lane CA, Hardy J, Schott JM. Alzheimer’s disease. Eur J Neurol (2018) 25(1):59– 10. Luczynski P, Laule C, Hsiung GYR, Moore GRW, Tremlett H. Coexistence of
70. doi: 10.1111/ene.13439 multiple sclerosis and alzheimer’s disease: a review. Mult Scler Relat Disord (2019)
2. Dobson R, Giovannoni G. Multiple sclerosis - a review. Eur J Neurol (2019) 26 27:232–8. doi: 10.1016/j.msard.2018.10.109
(1):27–40. doi: 10.1111/ene.13819 11. Huang J, Su B, Karhunen V, Gill D, Zuber V, Ahola-Olli A, et al. Inflammatory
3. Chiaravalloti ND, Deluca J. Cognitive impairment in multiple sclerosis. Lancet diseases, inflammatory biomarkers, and Alzheimer disease: an observational analysis
Neurology (2008) 7:1139–51. doi: 10.1016/S1474-4422(08)70259-X and mendelian randomization. Neurology (2023) 100(6):568–81. doi: 10.1212/
WNL.0000000000201489
4. Deloire MSA, Salort E, Bonnet M, Arimone Y, Boudineau M, Amieva H, et al.
Cognitive impairment as marker of diffuse brain abnormalities in early relapsing 12. Mahmoudi E, Sadaghiyani S, Lin P, Kamdar N, Norcott A, Peterson MD, et al. Diagnosis of
remitting multiple sclerosis. J Neurol Neurosurg Psychiatry (2005) 76:519–26. doi: alzheimer’s disease and related dementia among people with multiple sclerosis: Large cohort study,
10.1136/jnnp.2004.045872 USA. Mult Scler Relat Disord (2022) 57:103351. doi: 10.1016/j.msard.2021.103351

5. Compston A, Coles A. Multiple sclerosis. Lancet (2008) 372(9648):1502–17. doi: 13. Goldacre MJ, Wotton CJ. Associations between specific autoimmune diseases
10.1016/S0140-6736(08)61620-7 and subsequent dementia: retrospective record-linkage cohort study, UK. J Epidemiol
Community Health (2017) 71(6):576–83. doi: 10.1136/jech-2016-207809
6. Timmers M, Tesseur I, Bogert J, Zetterberg H, Blennow K, Börjesson-Hanson A, et al.
Relevance of the interplay between amyloid and tau for cognitive impairment in early alzheimer’s 14. Califf RM. Biomarker definitions and their applications. Exp Biol Med (2018)
disease. Neurobiol Aging (2019) 79:131–41. doi: 10.1016/j.neurobiolaging.2019.03.016 243(3):213–21. doi: 10.1177/1535370217750088
7. Lassmann H. Mechanisms of neurodegeneration shared between multiple 15. Ziemssen T, Akgün K, Brück W. Molecular biomarkers in multiple sclerosis. J
sclerosis and alzheimer’s disease. J Neural Transm (2011) 118:747–52. doi: 10.1007/ Neuroinflamm (2019) 16(1):1–11. doi: 10.1186/s12974-019-1674-2
s00702-011-0607-8 16. Niemantsverdriet E, Ottoy J, Somers C, De Roeck E, Struyfs H, Soetewey F, et al.
8. Nichols E, Steinmetz JD, Vollset SE, Fukutaki K, Chalek J, Abd-Allah F, et al. The cerebrospinal fluid Ab1-42/Ab1-40 ratio improves concordance with amyloid-
Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in PET for diagnosing alzheimer’s disease in a clinical setting. J Alzheimers Dis (2017) 60
2050: an analysis for the global burden of disease study 2019. Lancet Public Health (2):561–76. doi: 10.3233/JAD-170327
(2022) 7(2):e105–25. doi: 10.1016/S2468-2667(21)00249-8 17. Jack CR, Knopman DS, Jagust WJ, Petersen RC, Weiner MW, Aisen PS, et al.
9. Walton C, King R, Rechtman L, Kaye W, Leray E, Marrie RA, et al. Rising Tracking pathophysiological processes in alzheimer’s disease: an updated hypothetical
prevalence of multiple sclerosis worldwide: insights from the atlas of MS, third edition. model of dynamic biomarkers. Lancet Neurol (2013) 12(2):207–16. doi: 10.1016/S1474-
Mult Scler (2020) 26(14):1816–21. doi: 10.1177/1352458520970841 4422(12)70291-0

Frontiers in Immunology 23 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

18. Sperling RA, Aisen PS, Beckett LA, Bennett DA, Craft S, Fagan AM, et al. 44. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded
Toward defining the preclinical stages of alzheimer’s disease: recommendations from disability status scale (EDSS). Neurology (1983) 33(11):1444–52. doi: 10.1212/
the national institute on aging-alzheimer’s association workgroups on diagnostic WNL.33.11.1444
guidelines for alzheimer’s disease. Alzheimers Dement (2011) 7(3):280–92. doi: 45. Berg L. Clinical dementia rating. Br J Psychiatry (1984) 145(3):339–9. doi:
10.1016/j.jalz.2011.03.003 10.1192/S0007125000118082
19. Kreisl WC, Kim MJ, Coughlin JM, Henter ID, Owen DR, Innis RB. PET imaging 46. Rosen WG, Mohs RC, Davis KL. A new rating scale for alzheimer’s disease. Am J
of neuroinflammation in neurological disorders. Lancet Neurol (2020) 19(11):940–50. Psychiatry (1984) 141(11):1356–64. doi: 10.1176/ajp.141.11.1356
doi: 10.1016/S1474-4422(20)30346-X
47. Griffiths S, Sherman EMS, Strauss E. Dementia rating scale-2. In: Encyclopedia of
20. Paolicelli RC, Sierra A, Stevens B, Tremblay ME, Aguzzi A, Ajami B, et al. clinical neuropsychology (2011) (New York: Springer New York). p. 810–1.
Microglia states and nomenclature: a field at its crossroads. Neuron (2022) 110
(21):3458–83. doi: 10.1016/j.neuron.2022.10.020 48. Dubois B, Slachevsky A, Litvan I, Pillon B. The FAB: a frontal assessment battery
at bedside. Neurology (2000) 55(11):1621–6. doi: 10.1212/WNL.55.11.1621
21. Kinney JW, Bemiller SM, Murtishaw AS, Leisgang AM, Salazar AM, Lamb BT.
Inflammation as a central mechanism in alzheimer’s disease. Alzheimer’s Dementia: 49. Morris JC, Heyman A, Mohs RC, Hughes JP, van Belle G, Fillenbaum G, et al.
Trans Res Clin Interventions (2018) 4(1):575–90. doi: 10.1016/j.trci.2018.06.014 The consortium to establish a registry for alzheimer’s disease (CERAD). part i. clinical
and neuropsychological assesment of alzheimer’s disease. Neurology (1989) 39
22. Gao C, Shen X, Tan Y, Chen S. Pathogenesis, therapeutic strategies and (9):1159–9. doi: 10.1212/wnl.39.9.1159
biomarker development based on “omics” analysis related to microglia in alzheimer’s
disease. J Neuroinflamm (2022) 19(1):1–23. doi: 10.1186/s12974-022-02580-1 50. Nasreddine ZS, Phillips NA, Bé dirian V, Charbonneau S, Whitehead V, Collin I,
et al. The Montreal cognitive assessment, MoCA: a brief screening tool for mild
23. Frost GR, Jonas LA, Li YM. Friend, foe or both? immune activity in alzheimer’s cognitive impairment. J Am Geriatr Soc (2005) 53(4):695–9. doi: 10.1111/j.1532-
disease. Front Aging Neurosci (2019) 11:337. doi: 10.3389/fnagi.2019.00337 5415.2005.53221.x
24. Prokop S, Lee VMY, Trojanowski JQ. Neuroimmune interactions in alzheimer’s 51. Yu H, Liu X, Zhong Y. The effect of osteopontin on microglia. BioMed Res Int
disease–new frontier with old challenges? Prog Mol Biol Transl Sci (2019) 168:183–201. (2017) 2017:1879437. doi: 10.1155/2017/1879437
doi: 10.1016/bs.pmbts.2019.10.002
52. Comi C, Carecchio M, Chiocchetti A, Galimberti D. Osteopontin is increased in
25. Bellaver B, Ferrari-Souza JP, da Ros LU, Carter SF, Rodriguez-Vieitez E, Nordberg A, the cerebrospinal fluid of patients with alzheimer’s disease and its levels correlate with
et al. Astrocyte biomarkers in Alzheimer disease: a systematic review and meta-analysis. cognitive decline. J Alzheimer’s Dis (2010) 19:1143–8. doi: 10.3233/JAD-2010-1309
Neurology (2021) 96(24):E2944–55. doi: 10.1212/WNL.0000000000012109
53. Sun Y, Yin XS, Guo H, Han RK, Dong He R, Chi LJ. Elevated osteopontin levels
26. Absinta M, Maric D, Gharagozloo M, Garton T, Smith MD, Jin J, et al. A in mild cognitive impairment and alzheimer’s disease. Mediators Inflamm (2013)
lymphocyte-microglia-astrocyte axis in chronic active multiple sclerosis. Nature (2021) 2013:615745. doi: 10.1155/2013/615745
597(7878):709–14. doi: 10.1038/s41586-021-03892-7
54. Roxburgh RHSR, Seaman SR, Masterman T, Hensiek AE, Sawcer SJ, Vukusic S,
27. Dendrou CA, Fugger L, Friese MA. Immunopathology of multiple sclerosis. Nat et al. Multiple sclerosis severity score. Neurology (2005) 64(7):1144–51. doi: 10.1212/
Rev Immunol (2015) 15:545–58. doi: 10.1038/nri3871 01.WNL.0000156155.19270.F8
28. Correale J, Gaitá n MI, Ysrraelit MC, Fiol MP. Progressive multiple sclerosis: 55. Manouchehrinia A, Westerlind H, Kingwell E, Zhu F, Carruthers R, Ramanujam
from pathogenic mechanisms to treatment. Brain (2017) 140(3):527–46. doi: 10.1093/ R, et al. Age related multiple sclerosis severity score: disability ranked by age. Mult Scler
brain/aww258 (2017) 23(14):1938–46. doi: 10.1177/1352458517690618
29. Faissner S, Plemel JR, Gold R, Yong VW. Progressive multiple sclerosis: from 56. Poser S, Bauer HJ, Poser W. Prognosis of multiple sclerosis. results from an
pathophysiology to therapeutic strategies. Nat Rev Drug Discov (2019) 18(12):905–22. epidemiological area in Germany. Acta Neurol Scand (1982) 65(4):347–54. doi:
doi: 10.1038/s41573-019-0035-2 10.1111/j.1600-0404.1982.tb03091.x
30. Sucksdorff M, Matilainen M, Tuisku J, Polvinen E, Vuorimaa A, Rokka J, et al. 57. Tackenberg B, Ramanujam R, Institute K, Trygve Holmøy S, Marie Weideman
Brain TSPO-PET predicts later disease progression independent of relapses in multiple A, Barbour C, et al. New multiple sclerosis disease severity scale predicts future
sclerosis. Brain (2020) 143(11):3318–30. doi: 10.1093/brain/awaa275 accumulation of disability. Front Neurol (2017) 8:598. doi: 10.3389/fneur.2017.00598
31. Patsopoulos NA, Baranzini SE, Santaniello A, Shoostari P, Cotsapas C, Wong G, 58. Bergamaschi R, Quaglini S, Trojano M, Amato MP, Tavazzi E, Paolicelli D, et al.
et al. Multiple sclerosis genomic map implicates peripheral immune cells and microglia Early prediction of the long term evolution of multiple sclerosis: the Bayesian risk
in susceptibility. Science (2019) 365(6460):eaav7188. doi: 10.1126/science.aav7188 estimate for multiple sclerosis (BREMS) score. J Neurol Neurosurg Psychiatry (2007) 78
32. McNamara NB, Munro DAD, Bestard-Cuche N, Uyeda A, Bogie JFJ, Hoffmann (7):757–9. doi: 10.1136/jnnp.2006.107052
A, et al. Microglia regulate central nervous system myelin growth and integrity. Nature 59. Krupp LB, Larocca NG, Muir Nash J, Steinberg AD. The fatigue severity scale:
(2022) 17:1–10. doi: 10.1038/s41586-022-05534-y application to patients with multiple sclerosis and systemic lupus erythematosus. Arch
33. Allen CM, Mowry EM, Tintoré M, Dphil NE. Prognostication and Neurol (1989) 46(10):1121–3. doi: 10.1001/archneur.1989.00520460115022
contemporary management of clinically isolated syndrome. Neurol Neurosurg 60. Hughes CP, Berg L, Danziger WL, Coben LA, Martin RL. A new clinical scale for
Psychiatry (2021) 92:391–7. doi: 10.1136/jnnp-2020-323087 the staging of dementia. Br J Psychiatry (1982) 140(6):566–72. doi: 10.1192/
34. Makhani N, Tremlett H. The multiple sclerosis prodrome. Nat Rev Neurol bjp.140.6.566
(2021) 17(8):515–21. doi: 10.1038/s41582-021-00519-3 61. Sokolova A, Hill MD, Rahimi F, Warden LA, Halliday GM, Shepherd CE.
35. Disanto G, Adiutori R, Dobson R, Martinelli V, Costa GD, Runia T, et al. Serum Monocyte chemoattractant protein-1 plays a dominant role in the chronic
neurofilament light chain levels are increased in patients with a clinically isolated syndrome. J inflammation observed in alzheimer’s disease. Brain Pathol (2009) 19(3):392–8. doi:
Neurol Neurosurg Psychiatry (2016) 87(2):126–9. doi: 10.1136/jnnp-2014-309690 10.1111/j.1750-3639.2008.00188.x
36. Bjornevik K, Munger KL, Cortese M, Barro C, Healy BC, Niebuhr DW, et al. 62. Pillai JA, Bena J, Bebek G, Bekris LM, Bonner-Jackson A, Kou L, et al.
Serum neurofilament light chain levels in patients with presymptomatic multiple Inflammatory pathway analytes predicting rapid cognitive decline in MCI stage of
sclerosis. JAMA Neurol (2020) 77(1):58–64. doi: 10.1001/jamaneurol.2019.3238 alzheimer’s disease. Ann Clin Transl Neurol (2020) 7(7):1225–39. doi: 10.1002/
37. Hosseiny M, Newsome SD, Yousem DM. Radiologically isolated syndrome: a acn3.51109
review for neuroradiologists. Am J Neuroradiol (2020) 41(9):1542–9. doi: 10.3174/ 63. Westin K, Buchhave P, Nielsen H, Minthon L, Janciauskiene S, Hansson O.
ajnr.A6649 CCL2 is associated with a faster rate of cognitive decline during early stages of
38. Thompson AJ, Banwell BL, Barkhof F, Carroll WM, Coetzee T, Comi G, et al. alzheimer’s disease. PloS One (2012) 7(1):e30525. doi: 10.1371/journal.pone.0030525
Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria. Lancet Neurol 64. Fischer JS, Rudick RA, Cutter GR, Reingold SC. The multiple sclerosis functional
(2018) 17(2):162–73. doi: 10.1016/S1474-4422(17)30470-2 composite measure (MSFC): an integrated approach to MS clinical outcome assessment.
39. van Bokhoven P, de Wilde A, Vermunt L, Leferink PS, Heetveld S, Cummings J, Multiple Sclerosis J (1999) 5(4):244–50. doi: 10.1177/135245859900500409
et al. The alzheimer’s disease drug development landscape. Alzheimers Res Ther (2021) 65. Bever CT, Grattan L, Panitch HS, Johnson KP. The brief repeatable battery of
13(1):1–9. doi: 10.1186/s13195-021-00927-z neuropsychological tests for multiple sclerosis: a preliminary serial study. Multiple
40. Piehl F. Current and emerging disease-modulatory therapies and treatment targets for Sclerosis (1995) 1(3):165–9. doi: 10.1177/135245859500100306
multiple sclerosis. J Intern Med (2021) 289(6):771–91. doi: 10.1111/joim.13215 66. Parmenter BA, Weinstock-Guttman B, Garg N, Munschauer F, Benedict RHB.
41. Samson D, Schoelles KM. Chapter 2: medical tests guidance (2) developing the Screening for cognitive impairment in multiple sclerosis using the symbol digit
topic and structuring systematic reviews of medical tests: utility of PICOTS, analytic modalities test. Mult Scler (2007) 13(1):52–7. doi: 10.1177/1352458506070750
frameworks, decision trees, and other frameworks. J Gen Intern Med (2012) 27 67. Dichev V, Kazakova M, Sarafian V. YKL-40 and neuron-specific enolase in
(SUPPL.1):11–9. doi: 10.1007/s11606-012-2007-7 neurodegeneration and neuroinflammation. Rev Neurosci (2020) 31(5):539–53. doi:
42. Page MJ, Mckenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. 10.1515/revneuro-2019-0100
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. 68. Mendiola AS, Cardona AE. The IL-1b phenomena in neuroinflammatory
BMJ (2021) 372:n71. doi: 10.1136/bmj.n71 diseases. J Neural Transm (2018) 125(5):781–95. doi: 10.1007/s00702-017-1732-9
43. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”: a practical method 69. Swanson A, Willette AA, Initia ADN. Neuronal pentraxin 2 predicts medial
for grading the cognitive state of patients for the clinician. J Psychiatr Res (1975) 12 temporal atrophy and memory decline across the alzheimer’s disease spectrum. Brain
(3):189–98. doi: 10.1016/0022-3956(75)90026-6 Behav Immun (2016) 58:201–8. doi: 10.1016/j.bbi.2016.07.148

Frontiers in Immunology 24 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

70. Kester MI, Teunissen CE, Sutphen C, Herries EM, Ladenson JH, Xiong C, et al. 96. Zrzavy T, Hametner S, Wimmer I, Butovsky O, Weiner HL, Lassmann H. Loss
Cerebrospinal fluid VILIP-1 and YKL-40, candidate biomarkers to diagnose, predict of “homeostatic” microglia and patterns of their activation in active multiple sclerosis.
and monitor alzheimer’s disease in a memory clinic cohort. Alzheimers Res Ther (2015) Brain (2017) 140(7):1900–13. doi: 10.1093/brain/awx113
7(1):59. doi: 10.1186/s13195-015-0142-1 97. Jäckle K, Zeis T, Schaeren-Wiemers N, Junker A, van der Meer F, Kramann N,
71. Huber AK, Irani DN. Targeting CXCL13 during neuroinflammation. Adv et al. Molecular signature of slowly expanding lesions in progressive multiple sclerosis.
Neuroimmune Biol (2015) 6(1):1. doi: 10.3233/NIB-150101 Brain (2020) 143(7):2073–88. doi: 10.1093/brain/awaa158
72. Lehnardt S. Innate immunity and neuroinflammation in the CNS: the role of 98. Giovannoni G, Popescu V, Wuerfel J, Hellwig K, Iacobeus E, Jensen MB, et al.
microglia in toll-like receptor-mediated neuronal injury. Glia (2010) 58(3):253–63. Smouldering multiple sclerosis: the “real MS. Ther Adv Neurol Disord (2022)
doi: 10.1002/glia.20928 15:17562864211066751. doi: 10.1177/17562864211066751
73. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al. 99. Cree BAC, Hollenbach JA, Bove R, Kirkish G, Sacco S, Caverzasi E, et al. Silent
Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD progression in disease activity-free relapsing multiple sclerosis. Ann Neurol (2019) 85
initiative. BMJ (2003) 326(7379):41–4. doi: 10.1136/bmj.326.7379.41 (5):653–66. doi: 10.1002/ana.25463
74. Colonna M, Butovsky O. Microglia function in the central nervous system 100. Kappos L, Butzkueven H, Wiendl H, Spelman T, Pellegrini F, Chen Y, et al.
during health and neurodegeneration. Annu Rev Immunol (2017) 35:441–68. doi: Greater sensitivity to multiple sclerosis disability worsening and progression events
10.1146/annurev-immunol-051116-052358 using a roving versus a fixed reference value in a prospective cohort study. Mult Scler
75. Fan Z, Brooks DJ, Okello A, Edison P. An early and late peak in microglial activation (2018) 24(7):963–73. doi: 10.1177/1352458517709619
in alzheimer’s disease trajectory. Brain (2017) 140(3):792–803. doi: 10.1093/brain/aww34 101. Kappos L, Wolinsky JS, Giovannoni G, Arnold DL, Wang Q, Bernasconi C,
76. Rosé n C, Hansson O, Blennow K, Zetterberg H. Fluid biomarkers in alzheimer’s et al. Contribution of relapse-independent progression vs relapse-associated worsening
disease - current concepts. Mol Neurodegener (2013) 8(1):1–11. doi: 10.1186/1750- to overall confirmed disability accumulation in typical relapsing multiple sclerosis in a
1326-8-20 pooled analysis of 2 randomized clinical trials. JAMA Neurol (2020) 77(9):1132–40. doi:
10.1001/jamaneurol.2020.1568
77. Bose S, Cho J. Role of chemokine CCL2 and its receptor CCR2 in
neurodegenerative diseases. Arch Pharm Res (2013) 36(9):1039–50. doi: 10.1007/ 102. Baecher-Allan C, Kaskow BJ, Weiner HL. Multiple sclerosis: mechanisms and
s12272-013-0161-z immunotherapy. Neuron (2018) 97(4):742–68. doi: 10.1016/j.neuron.2018.01.021

78. Molinuevo JL, Ayton S, Batrla R, Bednar MM, Bittner T, Cummings J, et al. 103. Rissanen E, Tuisku J, Vahlberg T, Sucksdorff M, Paavilainen T, Parkkola R, et al.
Current state of alzheimer’s fluid biomarkers. Acta Neuropathol (2018) 136(6):23. doi: Microglial activation, white matter tract damage, and disability in MS. Neurol
10.1007/s00401-018-1932-x Neuroimmunol Neuroinflamm (2018) 5(3):e443. doi: 10.1212/NXI.0000000000000443

79. McGrowder DA, Miller F, Vaz K, Nwokocha C, Wilson-Clarke C, Anderson- 104. Poirion E, Tonietto M, Lejeune FX, Ricigliano VAG, Boudot de la Motte M,
Cross M, et al. Cerebrospinal fluid biomarkers of alzheimer’s disease: current evidence Benoit C, et al. Structural and clinical correlates of a periventricular gradient of
and future perspectives. Brain Sci (2021) 11(2):1–56. doi: 10.3390/brainsci11020215 neuroinflammation in multiple sclerosis. Neurology (2021) 96(14):e1865–75. doi:
10.1212/WNL.0000000000011700
80. Rosmus DD, Lange C, Ludwig F, Ajami B, Wieghofer P, Cappellano G, et al. The
role of osteopontin in microglia biology: current concepts and future perspectives. 105. Misin O, Matilainen M, Nylund M, Honkonen E, Rissanen E, Sucksdorff M,
Biomedicines (2022) 10(4):840. doi: 10.3390/biomedicines10040840 et al. Innate immune cell–related pathology in the thalamus signals a risk for disability
progression in multiple sclerosis. Neurol Neuroimmunol Neuroinflamm (2022) 9(4):
81. Wojcieszak J, Kuczyń ska K, Zawilska JB. Role of chemokines in the development e1182. doi: 10.1212/NXI.0000000000001182
and progression of alzheimer’s disease. J Mol Neurosci (2022) 72(9):1929–51. doi:
10.1007/s12031-022-02047-1 106. Carnero Contentti E, Correale J. Bruton’s tyrosine kinase inhibitors: a
promising emerging treatment option for multiple sclerosis. Expert Opin Emerg
82. De Schepper S, Ge JZ, Crowley G, Ferreira LSS, Garceau D, Toomey CE, et al. Drugs (2020) 25(4):377–81. doi: 10.1080/14728214.2020.1822817
Perivascular cells induce microglial phagocytic states and synaptic engulfment via SPP1
in mouse models of alzheimer’s disease. Nat Neurosci (2023) 26(3):406–15. 107. Keaney J, Gasser J, Gillet G, Scholz D, Kadiu I. Inhibition of bruton’s tyrosine
doi: 10.1038/s41593-023-01257-z kinase modulates microglial phagocytosis: therapeutic implications for alzheimer’s
disease. J Neuroimmune Pharmacol (2019) 14(3):448–61. doi: 10.1007/s11481-019-
83. Mcfarland KN, Chakrabarty P. Microglia in alzheimer’s disease: a key player in 09839-0
the transition between homeostasis and pathogenesis. Neurotherapeutics (2022)
19:186–208. doi: 10.1007/s13311-021-01179-3 108. Gruber RC, Dufault MR, Chretien N, Proto JD, Zhang M, Havari E, et al.
Establishing a role for the bruton’s tyrosine kinase inhibitor tolebrutinib in modulating
84. Goldman MD, Ward M, Goldman MD. Epidemiology and pathophysiology of neuroinflammation and disease progression in MS (1974). Neurology (2021) 96(15
multiple sclerosis. CONTINUUM Lifelong Learn Neurol (2022) 28(4):988–1005. Supplement).
doi: 10.1212/CON.0000000000001136
109. Oh J, Bar-Or A. Emerging therapies to target CNS pathophysiology in multiple
85. Hauser SL, Cree BAC. Treatment of multiple sclerosis: a review. Am J Med sclerosis. Nat Rev Neurol (2022) 18(8):466–75. doi: 10.1038/s41582-022-00675-0
(2020) 133:1380–90.e2. doi: 10.1016/j.amjmed.2020.05.049
110. Reich DS, Arnold DL, Vermersch P, Bar-Or A, Fox RJ, Matta A, et al. Safety
86. Lassmann H, Van Horssen J, Mahad D. Progressive multiple sclerosis: pathology and efficacy of tolebrutinib, an oral brain-penetrant BTK inhibitor, in relapsing
and pathogenesis. Nat Rev Neurol (2012) 8(11):647–56. doi: 10.1038/nrneurol.2012.168 multiple sclerosis: a phase 2b, randomised, double-blind, placebo-controlled trial.
87. Preziosa P, Pagani E, Meani A, Moiola L, Rodegher M, Filippi M, et al. Slowly Lancet Neurol (2021) 20(9):729–38. doi: 10.1016/S1474-4422(21)00237-4
expanding lesions predict 9-year multiple sclerosis disease progression. Neurol 111. Butovsky O, Koronyo-Hamaoui M, Kunis G, Ophir E, Landa G, Cohen H, et al.
Neuroimmunol Neuroinflamm (2022) 9:e1139. doi: 10.1212/NXI.0000000000001139 Glatiramer acetate fights against alzheimer’s disease by inducing dendritic-like
88. Calvi A, Carrasco FP, Tur C, Chard DT, Stutters J, De Angelis F, et al. microglia expressing insulin-like growth factor 1. Proc Natl Acad Sci USA (2006) 103
Association of slowly expanding lesions on MRI with disability in people with (31):11784–9. doi: 10.1073/pnas.0604681103
secondary progressive multiple sclerosis. Neurology (2022) 98(17):E1783–93. doi: 112. Paraiso HC, Kuo PC, Curfman ET, Moon HJ, Sweazey RD, Yen JH, et al.
10.1212/WNL.0000000000200144 Dimethyl fumarate attenuates reactive microglia and long-term memory deficits
89. Klistorner S, Barnett MH, Graham SL, Wang C, Klistorner A. The expansion following systemic immune challenge. J Neuroinflamm (2018) 15(1):1–14. doi:
and severity of chronic MS lesions follows a periventricular gradient. Multiple Sclerosis J 10.1186/s12974-018-1125-5
(2022) 28(10):1504–14. doi: 10.1177/13524585221080667 113. Rosito M, Testi C, Parisi G, Cortese B, Baiocco P, Di Angelantonio S. Exploring
90. Absinta M, Sati P, Masuzzo F, Nair G, Sethi V, Kolb H, et al. Association of the use of dimethyl fumarate as microglia modulator for neurodegenerative diseases
chronic active multiple sclerosis lesions with disability in vivo. JAMA Neurol (2019) 76 treatment. Antioxidants (2020) 9(8):700. doi: 10.3390/antiox9080700
(12):1474–83. doi: 10.1001/jamaneurol.2019.2399 114. Sun X, Suo X, Xia X, Yu C, Dou Y. Dimethyl fumarate is a potential therapeutic
91. Eisele P, Kraemer M, Dabringhaus A, Weber CE, Ebert A, Platten M, et al. option for alzheimer’s disease. J Alzheimers Dis (2022) 85(1):443–56. doi: 10.3233/JAD-
Characterization of chronic active multiple sclerosis lesions with sodium ( 23 Na) 215074
magnetic resonance imaging-preliminary observations. Eur J Neurol (2021) 28:2392–5. 115. McGhee DJM, Ritchie CW, Thompson PA, Wright DE, Zajicek JP, Counsell
doi: 10.1111/ene.14873 CE. A systematic review of biomarkers for disease progression in alzheimer’s disease.
92. Maggi P, Kuhle J, Schädelin S, van der Meer F, Weigel M, Galbusera R, et al. PloS One (2014) 9(2):e88854. doi: 10.1371/JOURNAL.PONE.0088854
Chronic white matter inflammation and serum neurofilament levels in multiple 116. Brosseron F, Krauthausen M, Kummer M, Heneka MT. Body fluid cytokine
sclerosis. Neurology (2021) 97(6):e543–53. doi: 10.1212/WNL.0000000000012326 levels in mild cognitive impairment and alzheimer’s disease: a comparative overview.
93. Van Schependom J, Guldolf K, D’Hooghe MB, Nagels G, D’Haeseleer M. Mol Neurobiol (2014) 50(2):534–44. doi: 10.1007/s12035-014-8657-1
Detecting neurodegenerative pathology in multiple sclerosis before irreversible brain 117. Park JE, Choi KY, Kim BC, Choi SM, Song MK, Lee JJ, et al. Cerebrospinal fluid
tissue loss sets in. Transl Neurodegener (2019) 8(1):1–17. doi: 10.1186/s40035-019-0178-4 biomarkers for the diagnosis of prodromal alzheimer’s disease in amnestic mild cognitive
94. Lassmann H. Pathogenic mechanisms associated with different clinical courses of impairment. Dement Geriatr Cognit Dis Extra (2019) 9(1):100–13. doi: 10.1159/000496920
multiple sclerosis. Front Immunol (2019) 9:3116. doi: 10.3389/FIMMU.2018.03116/PDF 118. Sun MJ, Liu N, Xie QF, Li X, Sun J, Wang H, et al. A candidate biomarker of
95. Voet S, Prinz M, van Loo G. Microglia in central nervous system inflammation glial fibrillary acidic protein in CSF and blood in differentiating multiple sclerosis and
and multiple sclerosis pathology. Trends Mol Med (2019) 25(2):112–23. doi: 10.1016/ its subtypes: a systematic review and meta-analysis. Multiple Sclerosis Related Disord
j.molmed.2018.11.005 (2021) 51:102870. doi: 10.1016/j.msard.2021.102870

Frontiers in Immunology 25 frontiersin.org

Temmerman et al. 10.3389/fimmu.2023.1162340

119. Bross M, Hackett M, Bernitsas E. Approved and emerging disease modifying 120. Lorscheider J, Buzzard K, Jokubaitis V, Spelman T, Havrdova E, Horakova D,
therapies on neurodegeneration in multiple sclerosis. Int J Mol Sci (2020) 21(12):4312. et al. Defining secondary progressive multiple sclerosis. Brain (2016) 139(Pt 9):2395–
doi: 10.3390/ijms21124312 405. doi: 10.1093/brain/aww173

Frontiers in Immunology 26 frontiersin.org