Association of Cerebrospinal Fluid S100B Protein with Core Biomarkers and Cognitive Deficits in Prodromal and Mild Alzheimer’s Disease

Abstract

Background:

Increased expression of the astroglial Ca²⁺-binding protein S100B has been observed in various neurodegenerative diseases and also seems to play a role in the unfolding of pathophysiological events at early stages of Alzheimer’s disease (AD).

Objective:

To examine the association of cerebrospinal fluid (CSF) levels of S100B with 1) established CSF core biomarkers total tau (tau), hyperphosphorylated tau (p-tau), and amyloid β_1–42 (Aβ_1–42) as well as neuron-specific enolase (NSE) CSF levels and 2) cognition in early AD and mild cognitive impairment (MCI) due to AD (MCI-AD).

Methods:

Retrospective study assessing 49 pooled charts of Memory Clinic and inpatients diagnosed with AD (N = 26) and MCI-AD (N = 23) according to the National Institute of Aging and Alzheimer’s Disease Association (NIA-AA) criteria. Neuropsychological testing was performed with the Consortium to Establish a Registry for AD (CERAD)-Plus battery.

Results:

CSF levels of S100B correlated with NSE, but not the other CSF parameters. Stepwise multiple linear regression, adjusted for age, sex, and educational level, revealed that only increased CSF S100B was independently associated with lower CERAD-Plus total and Mini-Mental Status Examination scores together with poorer performance in wordlist learning (delayed recall and overall performance). We found no independent associations with other CSF biomarkers or cognitive domains.

Conclusion:

Our data suggest that CSF S100B may have a diagnostic value particularly at early stages of AD reflecting the significance of neuroinflammatory/astroglial processes. Thus, CSF S100B may complement the established array of available AD biomarkers to improve early stage diagnosis.

Keywords

Alzheimer’s disease CERAD-Plus cerebrospinal fluid biomarkers mild cognitive impairment S100B protein

INTRODUCTION

The diagnostic process of early stages of Alzheimer’s disease (AD) is increasingly relying on biomarkers such as radiologically quantitated hippocampal volume loss and cerebrospinal fluid (CSF) levels of amyloid-β_1–42 protein (Aβ_1–42) as well as total protein tau and tau phosphorylated at threonine 181 (“hyperphosphorylated tau”, p-tau) as neurodegenerative parameters [1 –3]. These biomarkers are directly related to the main pathophysiological pathways and reflect different stages of the disease. Amyloid-β_1–42 deposition as neuritic plaques is believed to occur already at presymptomatic stages and presumably many years before the onset of cognitive decline [4]. This process precedes and supposedly triggers a neurodegenerative cascade with the occurrence of intracellular neurofibrillary tangles that ultimately leads to neuronal injury and loss as the key determinants of cognitive impairment [5, 6]. Decreased levels of Aβ_1–42 and increased tau and p-tau values in the CSF may thus mirror the temporal order of pathological changes in the brain of AD patients. However, in order to facilitate early recognition of the disease, sensitive biomarkers to monitor additional molecular mechanisms of the disease are constantly sought.

It is assumed that Aβ_1–42 and tau pathology is triggered and exacerbated by the activation of microglia and astrocytes thus assigning neuroinflammation a prominent role in the pathophysiology of AD (e. g., [7]; for review see [8, 9]). Potent inductors of proinflammatory cascades that drive the response of the inflammasome via cytokine expression are certain endogenous signal molecules referred to as danger- or damage-associated molecular patterns (DAMPs) [10]. This highly diverse family of molecules comprising bacterial lipids/peptides, endogenous proteins, and nucleic acids is released from compromised cells into the extracellular space under conditions of cellular stress or death and tissue injury and seems to be strongly implicated in early AD pathophysiology [11].

One member of this family is the predominantly astrocytic Ca²⁺-binding protein S100B which is known to be elevated in the CSF or serum of patients in conditions such diverse as traumatic brain injury [12], stroke [13], schizophrenia [14], preeclampsia [15], and a variety of neurodegenerative diseases such as AD [11] and Parkinson’s disease [16]. In AD, neurotrophic activity of S100B is long known to be increased in relevant brain areas such as the temporal lobe [17]. In an earlier magnetic resonance imaging (MRI) study, CSF S100B correlated negatively with normalized brain volume in individuals with AD, but not frontotemporal lobe dementia [18]. Chaves et al. demonstrated an association of serum S100B levels and global functional deficits [19]. To our knowledge, however, neither associations between S100B levels and different cognitive subdomains nor between S100B levels and the core diagnostic CSF biomarkers of AD have yet been investigated.

The present study aims to provide a step toward an answer to this important question. Given that neuroinflammation and microglial activation are mandatory and early manifestations of AD pathophysiology, we hypothesized a relationship between CSF S100B levels and typical early stage cognitive deficits such as verbal memory. It is well conceivable that this general marker of neuroinflammation is particularly suited for the early diagnosis of AD at early clinical or even preclinical stages of the disease. Next to amyloid and neurodegenerative biomarkers such as tau and p-tau, CSF S100B may reasonably complement the established array of available AD biomarkers to improve early stage diagnosis.

MATERIALS AND METHODS

Patients

In this retrospective study, data were collected from charts of out- and inpatients of the Department of Psychiatry and Psychotherapy and the Department of Neurology, Heinrich Heine University of Düsseldorf, Germany. Patients had undergone a thorough clinical and neuropsychological examination, basic chemical/hematological laboratory assessment, a lumbar puncture with measurement of the CSF core biomarkers Aβ_1–42, tau, p-tau, and in addition NSE and S100B and an MRI scan. Biomarker supported diagnosis of both probable AD and mild cognitive impairment (MCI)-AD was established according to the revised National Institute of Aging and Alzheimer’s Association (NIA-AA) criteria [1, 2]. Charts of patients with a history of cerebrovascular incidents, epilepsy, traumatic brain injury, and other neurodegenerative (i.e., Parkinson’s disease/Lewy body disease) or major psychiatric diseases (i.e., schizophrenia, depression) apart from AD as well as with documented significant focal neurological signs upon admission/presentation were excluded. Documented MRI abnormalities such as acute or past cerebrovascular lesions (including significant periventricular or deep white matter lesions), brain tumor, inflammatory signs, or hydrocephalus were also exclusion criteria. The study was approved by the Ethics Committee of the Medical Faculty of the Heinrich Heine University Düsseldorf, Germany (study ID: #6039R).

Neuropsychological assessment

Neuropsychological function was assessed using the Consortium to Establish a Registry for Alzheimer’s Disease – Neuropsychological Battery (CERAD); specifically, the extended and validated German version CERAD-Plus commonly used in memory clinics across German-speaking Europe was applied [20, 21]. This battery includes subtests for verbal episodic learning and memory (Wordlist–Encoding, – Delayed recall and – Recognition), constructional praxis (Figures–Copy), visual episodic memory (Figures–Delayed recall), language (15-item Boston Naming Test, BNT), and executive functions (phonemic and semantic verbal fluency, Trail Making Tests (TMT) A and B). The Mini-Mental Status Examination (MMSE) is also an integrated part of the CERAD-Plus. For all subtests, raw scores and z-scores demographically adjusted for age, gender, and education were derived [22]. Moreover, a composite (“total”) score of the CERAD-plus was used as described previously [23].

Sampling and CSF analysis

CSF samples were obtained by lumbar puncture after informed consent and processed as part of the clinical routine. Analysis of all parameters was performed by a commercial laboratory partner (MVZ Synlab Leverkusen, Germany). Standardized sandwich ELISA methods were used for measurement of the core biomarkers, namely the INNOTEST^® β-AMYLOID (1–42), INNOTEST^® hTAU Ag, INNOTEST^® PHOSPHO TAU (181P). Neuron-specific enolase and S100B were measured using the fully automated commercially available chemiluminescence immunoassays LIAISON^® S100 and LIAISON^® NSE (DiaSorin, Italy). Standardized cut-off values for these parameters were provided by the laboratory.

Statistical analysis

Between-group differences (MCI-AD versus AD) were assessed using the Mann-Whitney-U-test for nonparametric continuous and χ²-test for categorical variables. For interrelations between the different CSF parameters, Spearman’s rank sum correlation coefficients with Bonferroni post-hoc correction for multiple comparisons were determined.

For evaluation of the individual contribution of each CSF biomarker on the prediction of cognitive deficits, we conducted a stepwise multiple linear regression with raw values of the CERAD-Plus subtests as dependent variables and the CSF parameters as independent factors. Models were adjusted for age, sex, and educational level (years of education). Statistical analyses were performed with SPSS (Version 22, IBM). The significance level was set at p < 0.05.

RESULTS

Sample characteristics

Sociodemographic, clinical, and neuropsychological characteristics of the sample are presented in Table 1. Charts of n = 49 individuals were included (n = 23 with MCI-AD and n = 26 with AD). The mean age of the whole sample was 69.3±7.4 years and the sex ratio was balanced (24/25). Subjects had a mean of 13.1±3.1 years of education. MMSE scores ranged from 20 to 30 within the whole sample. The AD group thus by definition consisted of mild cases. There were no significant differences in basic demographic variables, such as age, sex, and education, between the two groups. Table 1 also includes the cognitive global scores MMSE and CERAD-Plus total. Expectedly, the two groups differed significantly with lower overall scores for the AD group. Regarding CSF biomarkers, all assessed parameters of the AD group were pathological, i.e., the core biomarkers tau, p-tau (increased), and Aβ_1–42 (decreased) as well as the additional parameters NSE and S100B (both increased). Patients with MCI-AD also showed pathological levels for tau, Aβ_1–42, and NSE, but not for p-tau and S100B. Levels for tau and p-tau were significantly different between the groups. To further characterize the MCI-AD group, we used the A/T/N biomarker classification proposed by Jack et al. [24]. Herein, “A(+)” or “A(-)“ refers to the presence or absence of a Aβ biomarker (low CSF Aβ_1–42 or positive amyloid PET), “T(+)” or “T(–)“ to a tau biomarker (CSF p-tau or tau PET), and “N(+)” or “N(–)“ to biomarkers of neuronal degeneration or injury ([(¹⁸)F]-fluorodeoxy-glucose-PET, structural MRI, or CSF total tau). All 23 individuals with MCI-AD were N(+) as evidenced by medial temporal lobe atrophy of grade ≥2 (Scheltens score) and/or elevated CSF total tau. However, as depicted in Fig. 1, the group can be subdivided in A(+)T(+) cases with the highest risk of conversion to AD (N = 5), A(+)T(-) cases (N = 13), and A(-)T(-) cases (N = 5).

Table 1

Demographic data, global levels of cognition, and CSF marker levels

	MCI-AD (n = 23)	AD (n = 26)	Difference (p)
Demographic data
Age	67.8±6.8	70.6±7.8	0.182
Sex [male/female]	13/10	11/15	0.886
Education [y]	13.5±2.6	12.8±3.6	0.187
Cognitive level (global)
MMSE (raw)	27.9±1.7	23.9±2.3	<0.001
MMSE (z-score)	–1.0±1.5	–2.8±2.1	<0.001
CERAD total	72.3±9.0	57.1±11.5	<0.001
CSF marker level^*
Aβ_1–42 (>550 pg/mL)	488.0 (201.9)	446.0 (177.1)	0.596
tau (<500 pg/mL)	599.8 (278.7)	837.6 (389.6)	0.025
p-tau (<61 pg/mL)	57.5 (17.5)	81.1 (29.5)	0.003
NSE (<13.0 ng/mL)	19.6 (6.2)	23.2 (7.8)	0.098
S100B (<2700 pg/mL)	2410.1 (1200.2)	2925.8 (1703.5)	0.441

Mann-Whitney-U-Test (sex: χ²-Test); statistically significant p-values in bold Data are mean±standard deviation for Demographic data and Cognitive level and median (interquartile range) for CSF marker level. ^*Cut-off values in brackets. See text for abbreviations.

Fig.1

CSF Aβ_1–42 versus p-tau levels of the MCI-AD group. Subgroups according to the A/T/N biomarker classification are defined by circles (see text for details). Dotted lines indicate the reference range (Aβ_1–42) or value (p-tau).

With regard to neuropsychological testing, AD patients performed worse in all cognitive subtests of the CERAD-Plus compared to MCI subjects with statistical significance for semantic verbal fluency (p < 0.001), BNT (p = 0.008), immediate (p = 0.003) and delayed (p = 0.006) verbal recall, word list recognition (p = 0.043), constructional praxis (p = 0.013), and savings of constructional praxis (p = 0.02).

Interrelationships of CSF parameters

The strongest correlation between the CSF parameters could be demonstrated for the neuronal destruction marker NSE which was significantly correlated with tau, p-tau, and S100B (Table 2). CSF tau and p-tau showed moderate correlations with one another as well as with NSE. CSF S100B concentrations were correlated with NSE levels. Notably, Aβ_1–42 did not correlate with any of the other CSF parameters.

Table 2

Interrelationship of CSF parameters across all subjects

	Aβ_1–42	tau	p-tau	NSE	S100B
S100B	0.07	0.27	0.17	0.39^*	1
NSE	0.27	0.69^***	0.51^***	1
p-tau	0.22	0.55^***	1
tau	0.14	1
Aβ_1–42	1

Spearman’s rho; significant p-values in bold with ^*p < 0.05, ^**p < 0.01, ^***p < 0.001. See text for abbreviations.

Stepwise regression of sociodemographic, cognitive, and CSF parameters

To follow and establish the possible individual contribution of each biomarker on the prediction of deficits in CERAD-Plus subtests, we conducted a stepwise multiple regression defining raw values of the CERAD-Plus subtests as dependent variables and the CSF parameters as independent factors with age, sex, and educational level as covariates (Table 3). Two regression models were calculated on the pooled sample of AD and MCI-AD individuals:

Table 3

Stepwise linear regression analyses for cognitive measures versus CSF parameters

Dependent variables	Independent factors	B	SE B	β	p	R²
Model 1
CERAD total		S100B	–0.004	0.001	–0.47	0.002	0.22
MMSE	step 1	S100B	–0.001	0.000	–0.37	0.013	0.14
	step 2	S100B	–0.001	0.000	–0.41	0.006
		sex	–1.601	0.790	–0.28	0.049	0.22
Word list learning (delay)	step 1	S100B	–0.001	0.000	–0.47	0.001	0.22
	step 2	S100B	–0.001	0.000	–0.47	0.001
		Aβ_1–42	0.003	0.002	0.28	0.042	0.29
Word list learning (sum)		S100B	–0.002	0.000	–0.48	0.001	0.23
Model 2
CERAD total	step1	S100B	–0.004	0.001	–0.454	0.005	0.207
	step2	S100B	–0.003	0.001	–0.428	0.005
		education	1.162	0.534	0.313	0.037	0.304
MMSE		S100B	–0.001	0.000	–0.388	0.015	0.15

Models are adjusted for age, sex, and education. Only statistically significant associations are listed (p < 0.05). B, regression coefficient; SE, standard error; β, standardized regression coefficient. Model 1 includes all MCI-AD and AD patients. Model 2 excludes A(-) subjects. See text for details.

Model 1 included the complete MCI-AD sample. In this model, the regression was significant for both the CERAD-Plus total and the MMSE score with S100B as the only CSF parameter. This indicates that increased values of S100B are more likely to explain overall cognitive deficits than the other biomarkers in our sample of patients with AD and MCI-AD. If the CERAD-Plus total score was chosen as the dependent variable, the regression coefficient of S100B was B =–0.004±0.001, indicating that a decrease of one unit (score point) of CERAD-Plus total is associated with an increase of 250±62.5 pg/mL of S100B concentration (β= 0.47, p = 0.002), independent of age, sex, and educational level. For the MMSE score as the dependent variable, the regression coefficient of S100B was B =–0.001±0.000, indicating that a decrease of one unit (score point) of the MMSE corresponded to an increase of 1000 pg/mL of S100B concentration (β= 0.37, p = 0.013), independent of age, sex, and educational level. Of note, we found a separate (albeit weak) influence of sex on the MMSE score (β= –0.28, p = 0.049) in this sample (Table 3).

Subsequently, the regression analysis was rerun after exclusion of the prognostically less relevant A(-) subgroup of MCI-AD individuals (see Fig. 1; model 2). However, S100B still proved the only (and highly) significant CSF biomarker to explain global cognitive deficits as measured by CERAD-Plus total and MMSE scores (accordingly, Spearman’s correlations of S100B with these parameters revealed significant associations as displayed in Fig. 2). Interestingly, in a second step, educational level showed a significant influence (Table 3).

Fig.2

Spearman correlations of CSF S100B with the (A) CERAD-Plus total and (B) MMSE scores of the pooled sample of patients with MCI-AD (prodromal AD) and AD.

Table 3 also displays associations of the various CERAD-plus subtests (raw values) with CSF biomarkers. CSF biomarker changes were suitable to predict deficits regarding verbal memory but not other CERAD-plus subtests. Specifically, both a decreased total verbal learning score (β= 0.48, p = 0.001) and the recall of verbal material (β= 0.47, p = 0.001) were associated S100B level increase. However, to a lesser extent, step 2 of the linear regression model also revealed an association of verbal recall and CSF Aβ_1–42 (β= 0.28, p = 0.042) as the only other biomarker. Of note, in model 2 (i.e., after exclusion of the A(–) subgroup) no association with distinct CERAD-Plus subtests was found.

DISCUSSION

The present study was conducted to clarify the role of CSF S100B as a well-known nonspecific marker of astroglial activation compared to the established “core” biomarkers for the diagnosis of prodromal and mild AD. Therefore, we were interested in possible associations of S100B and the other CSF biomarker levels with cognitive deficits (as assessed by subtests of the CERAD-Plus battery) and interrelationships between these CSF biomarkers and in a routine diagnostic sample.

The main finding of our study was that S100B was the only CSF biomarker with a significant association with measures of global cognition (i.e., MMST, CERAD-Plus total score) and verbal subtests (i.e., verbal memory) in a stepwise multiple linear regression model adjusted for age, sex, and educational level. We also found a single association of Aβ_1–42 with delayed recall which, however, was marginal. Other CSF parameters showed no association with any other cognitive subtest. Thus, our data point towards a superiority of the rather unspecific biomarker S100B over the other CSF parameters in predicting reduced CERAD-Plus subtest scores (especially those testing early occurring deficits such as verbal memory loss) in this pooled sample of prodromal and mild AD. This is in agreement with results of Peskind et al. who also showed elevated CSF S100B levels in early stage AD [25]. These authors found that S100B was suitable to distinguish subjects with mild/moderate AD from patients in advanced disease stages on the one hand and matched healthy controls on the other hand. They also disclosed a significant correlation between S100B and MMSE scores of AD patients. Similar findings were also reported in the serum of AD patients [19]. Our data corroborate and add to these findings in that S100B was independently associated with score reduction in CERAD-Plus total and verbal memory subtests in detail as early neuropsychological markers of AD. Elevation of CSF S100B strongly reflects neuroinflammatory microglia and astrocytes activation and appears to play a direct role in the pathogenesis of AD, particularly regarding the formation of amyloid plaques [26] (for review see [11]).

In addition to tau, p-tau, Aβ_1–42, and S100B, we also included NSE in our analysis. For this consistent marker of neuronal destruction, we could demonstrate significant correlations with all other biomarkers except Aβ_1–42. Indeed, preclinical and clinical evidence point towards cross-relationships between NSE and the amyloid pathway [27] as well as tau and p-tau [28, 29] under physiological and pathological conditions. Total tau levels correlated with all other parameters except Aβ_1–42 and S100B. Interestingly, p-tau as a known predictor for disease progression [30, 31] as well as S100B with the best diagnostic value in our sample correlated with NSE.

When compared between the two subgroups, tau, p-tau, and NSE, but not S100B or Aβ_1–42, differed significantly and thus contributed to the distinction of manifest and prodromal AD. However, S100B and NSE levels were elevated in both groups (It has to be considered, though, that cut-off levels are somewhat arbitrary since they are not yet evaluated for AD diagnostics). The Aβ_1–42 level in this sample as a whole was pathological but mildly decreased in both groups which was expected since Aβ_1–42 is regarded an early disease marker often preceding even first clinical signs of AD [4]. Similarly, Haldenwanger et al. showed a correlation only in patients with amnestic MCI but not in manifest AD [32]. One could argue that Aβ_1–42 accumulation and subsequent concentration decline in CSF seems to be not yet in a steady state during preclinical disease stages, but may level off with the progression of neuropsychological deficits. In fact, as opposed to neuronal destruction/neurodegeneration markers, Aβ_1–42 has been shown to have fully changed many years before onset of AD [33, 34].

We found no associations for tau/p-tau or NSE with the CERAD-Plus subtest results. Previous studies have already established close relationships of cognitive function, particularly memory, with CSF tau/p-tau [30 , 36] as well as with the underlying neurofibrillary pathology [37] in samples with AD and MCI. Current guidelines therefore recommend their assessment in the advanced diagnostic process of AD and MCI-AD [1, 2]. Correlations of CSF and serum NSE with MMSE scores have been reported in the literature; however, these samples to a large proportion contained severe stages [19, 28]. As a matter of fact, the intraneuronal NSE is secreted into the extracellular space only after substantial neuronal damage which increases as the neurodegenerative disease advances [19].

Of note, when analyzing the interrelationship of CSF S100B levels with those of the other biomarkers, S100B only correlated (moderately) with the unspecific neuronal destruction marker NSE, but not with the established AD biomarkers, underscoring its unspecific (but possibly sensitive) character. To our knowledge, the present study is the first to investigate correlations of CSF S100B with the “diagnostic” biomarkers tau, p-tau, and/or Aβ_1–42.

Two limitations of the present study that need to be mentioned originate in its retrospective nature. The first is its lack of a control group which would of course strengthen our conclusions. The existing data show a strong association of S100B levels with measures of global cognition but do not provide enough evidence to assess its diagnostic value regarding in comparison to established CSF biomarkers. The second limitation is the relatively small sample size (which is due to a recent reorganization/harmonization of neuropsychological testing procedures and the fact that, for lumbar puncture, hospital admission is needed which is often not favored by patients and relatives).

To summarize, in a sample of individuals with prodromal and/or early stages of AD, S100B proved a powerful predictor for deficits of global cognition (assessed by the CERAD-Plus total and MMSE scores) and superior to other (including established) AD biomarkers. This may reflect the prominent pathophysiological role of neuroinflammation and astrocytic/microglial activation at these particular disease stages. Further studies should clarify if S100B might be sensitive to detect prodromal and early stages of AD. If confirmed, S100B could represent an additional molecular candidate biomarker to be integrated in an expanding biomarker array used to identify early disease stages in clinical routine diagnostic settings. It may also seem reasonable from a general point of view to add a typical inflammatory marker to the array of established amyloid and neurodegenerative CSF biomarkers to cover a broader spectrum of molecular events ensuing in the pathophysiological process of AD.

Footnotes

ACKNOWLEDGMENTS

This study was funded in part by Stiftung für Altersforschung (Aging Research Foundation) of the Heinrich Heine University Düsseldorf, Germany (Grant# 10-16-53/2018).

Authors’ disclosures available online ().

References

McKhann

, Knopman

, Chertkow

, Hyman

, Jack CR

, Kawas

, Klunk

, Koroshetz

, Manly

, Mayeux

, Mohs

, Morris

, Rossor

, Scheltens

, Carrillo

, Thies

, Weintraub

, Phelps

(2011) The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 263–269.

Albert

, DeKosky

, Dickson

, Dubois

, Feldman

, Fox

, Gamst

, Holtzman

, Jagust

, Petersen

, Snyder

, Carrillo

, Thies

, Phelps

(2011) The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 270–279.

Blennow

, Biscetti

, Eusebi

, Parnetti (2016) Cerebrospinal fluid biomarkers in Alzheimer’s and Parkinson’s diseases-From pathophysiology to clinical practice. Mov Disord 31, 836–847.

Ingelsson

, Fukumoto

, Newell

, Growdon

, Hedley-Whyte

, Frosch

, Albert

, Hyman

, Irizarry

(2004) Early Abeta accumulation and progressive synaptic loss, gliosis, and tangle formation in AD brain. Neurology 62, 925–931.

Jack

Jr , Knopman

, Jagust

, Shaw

, Aisen

, Weiner

, Petersen

, Trojanowski

(2010) Hypothetical model of dynamic biomarkers of the Alzheimer’s pathological cascade. Lancet Neurol 9, 119–128.

J Jack

Jr , Knopman

, Jagust

, Petersen

, Weiner

, Aisen

, Shaw

, Vemuri

, Wiste

, Weigand

, Lesnick

, Pankratz

, Donohue

, Trojanowski

(2013) Tracking pathophysiological processes in Alzheimer’s disease: An updated hypothetical model of dynamic biomarkers. Lancet Neurol 12, 207–216.

Rossner

, Lange-Dohna

, Zeitschel

, Perez-Polo

(2005) Alzheimer’s disease beta-secretase BACE1 is not a neuron-specific enzyme. J Neurochem 92, 226–234.

Nichols

, St-Pierre

, Wendeln

, Makoni

, Gouwens

, Garrad

, Sohrabi

, Neher

, Tremblay

, Combs

(2019) Inflammatory mechanisms in neurodegeneration. J Neurochem 149, 562–581.

Maphis

, Xu

, Kokiko-Cochran

, Jiang

, Cardona

, Ransohoff

, Lamb

, Bhaskar

(2015) Reactive microglia drive tau pathology and contribute to the spreading of pathological tau in the brain. Brain 138, 1738–1755.

10.

Roh

, Sohn

(2018) Damage-associated molecular patterns in inflammatory diseases. Immune Netw 18, e27.

11.

Venegas

, Heneka

(2017) Danger-associated molecular patterns in Alzheimer’s disease. J Leukoc Biol 101, 87–98.

12.

Kellermann

, Kleindienst

, Hore

, Buchfelder

, Brandner

(2016) Early CSF and serum S100B concentrations for outcome prediction in traumatic brain injury and subarachnoid hemorrhage. Clin Neurol Neurosurg 145, 79–83.

13.

Glushakova

, Glushakov

, Miller

, Valadka

, Hayes

(2016) Biomarkers for acute diagnosis and management of stroke in neurointensive care units. Brain Circ 2, 28–47.

14.

Rothermundt

, Falkai

, Ponath

, Abel

, Bürkle

, Diedrich

, Hetzel

, Peters

, Siegmund

, Pedersen

, Maier

, Schramm

, Suslow

, Ohrmann

, Arolt

(2004) Glial cell dysfunction in schizophrenia indicated by increased S100B in the CSF. Mol Psychiatry 9, 897–899.

15.

Bergman

, Zetterberg

, Kaihola

, Hagberg

, Blennow

, Akerud

(2018) Blood-based cerebral biomarkers in preeclampsia: Plasma concentrations of NfL, tau, S100B and NSE during pregnancy in women who later develop preeclampsia - A nested case control study. PloS One 13, e0196025.

16.

Gmitterova

, Gawinecka

, Llorens

, Varges

, Valkovic

, Zerr

(2018) Cerebrospinal fluid markers analysis in the differential diagnosis of dementia with Lewy bodies and Parkinson’s disease dementia. Eur Arch Psychiatry Clin Neurosci, doi: 10.1007/s00406-018-0928-9

17.

Marshak

, Pesce

, Stanley

, Griffin

(1992) Increased S100 beta neurotrophic activity in Alzheimer’s disease temporal lobe. Neurobiol Aging 13, 1–7.

18.

Petzold

, Jenkins

, Watt

, Green

, Thompson

, Keir

, Fox

, Rossor

(2003) Cerebrospinal fluid S100B correlates with brain atrophy in Alzheimer’s disease. Neurosci Lett 336, 167–170.

19.

Chaves

, Camozzato

, Ferreira

, Piazenski

, Kochhann

, Dall’Igna

, Mazzini

, Souza

, Portela

(2010) Serum levels of S100B and NSE proteins in Alzheimer’s disease patients. J Neuroinflammation 7, 6.

20.

Welsh

, Butters

, Hughes

, Mohs

, Heyman

(1992) Detection and staging of dementia in Alzheimer’s disease. Use of the neuropsychological measures developed for the Consortium to Establish a Registry for Alzheimer’s Disease. Arch Neurol 49, 448–452.

21.

Verhülsdonk

, Quack

, Höft

, Lange-Asschenfeldt

, Supprian

(2013) Anosognosia and depression in patients with Alzheimer’s dementia. Arch Gerontol Geriatr 57, 282–287.

22.

Berres

, Zehnder

, Blasi

, Monsch

(2008) Evaluation of diagnostic scores with adjustment for covariates. Stat Med 27, 1777–1790.

23.

Wolfsgruber

, Jessen

, Wiese

, Stein

, Bickel

, Mösch

, Weyerer

, Werle

, Pentzek

, Fuchs

, Köhler

, Bachmann

, Riedel-Heller

, Scherer

, Maier

, Wagner

; AgeCoDe Study Group (2014) The CERAD neuropsychological assessment battery total score detects and predicts Alzheimer disease dementia with high diagnostic accuracy. Am J Geriatr Psychiatry 22, 1017–1028.

24.

J Jack

Jr , Bennett

, Blennow

, Carrillo

, Feldman

, Frisoni

, Hampel

, Jagust

, Johnson

, Knopman

, Petersen

, Scheltens

, Sperling

, Dubois

(2016) A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology 87, 539–547.

25.

Peskind

, Griffin

, Akama

, Raskind

, Van Eldik

(2001) Cerebrospinal fluid S100B is elevated in the earlier stages of Alzheimer’s disease. Neurochem Int 39, 409–413.

26.

Mori

, Koyama

, Arendash

, Horikoshi-Sakuraba

, Tan

, Town

(2010) Overexpression of human S100B exacerbates cerebral amyloidosis and gliosis in the Tg2576 mouse model of Alzheimer’s disease. Glia 58, 300–314.

27.

Hwang

, Cho

, Lee

, Chae

, Lim

, Min

, Seo

, Song

, Paik

, Sheen

, Kim

(2004) Aberrant expressions of pathogenic phenotype in Alzheimer’s diseased transgenic mice carrying NSE-controlled APPsw. Exp Neurol 186, 20–32.

28.

Schmidt

, Mergl

, Stach

, Jahn

, Gertz

, Schönknecht

(2014) Elevated levels of cerebrospinal fluid neuron-specific enolase (NSE) in Alzheimer’s disease. Neurosci Lett 570, 81–85.

29.

Palumbo

, Siepi

, Sabalich

, Tranfaglia

, Parnetti

(2008) Cerebrospinal fluid neuron-specific enolase: A further marker of Alzheimer’s disease? Functional Neurol 23, 93–96.

30.

Reijs

BLR

, Ramakers

IHGB

, Köhler

, Teunissen

, Koel-Simmelink

, Nathan

, Tsolaki

, Wahlund

, Waldemar

, Hausner

, Vandenberghe

, Johannsen

, Blackwell

, Vanderstichele

, Verhey

, Visser

(2017) Memory correlates of Alzheimer’s disease cerebrospinal fluid markers: A longitudinal cohort study. J Alzheimers Dis 60, 1119–1128.

31.

Henneman

, Vrenken

, Barnes

, Sluimer

, Verwey

, Blankenstein

, Klein

, Fox

, Scheltens

, Barkhof

, van der

Flier WM

(2009) Baseline CSF-tau levels independently predict progression of hippocampal atrophy in Alzheimer disease. Neurology 73, 935–940.

32.

Haldenwanger

, Eling

, Kastrup

, Hildebrandt

(2010) Correlation between cognitive impairment and CSF biomarkers in amnesic MCI, non-amnesic MCI, and Alzheimer’s disease. J Alzheimers Dis 22, 971–980.

33.

Buchhave

, Minthon

, Zetterberg

, Wallin

, Blennow

, Hansson

(2012) Cerebrospinal fluid levels of beta-amyloid 1-42, but not of tau, are fully changed already 5 to 10 years before the onset of Alzheimer dementia. Arch Gen Psychiatry 69, 98–106.

34.

Jansen

, Ossenkoppele

, Knol

, Tijms

, Scheltens

, Verhey

, Visser

; Amyloid Biomarker Study Group, Aalten

, Aarsland

, Alcolea

, Alexander

, Almdahl

, Arnold

, Baldeiras

, Barthel

, van Berckel

, Bibeau

, Blennow

, Brooks

, van Buchem

, Camus

, Cavedo

, Chen

, Chetelat

, Cohen

, Drzezga

, Engelborghs

, Fagan

, Fladby

, Fleisher

, van der Flier

, Ford

, Förster

, Fortea

, Foskett

, Frederiksen

, Freund-Levi

, Frisoni

, Froelich

, Gabryelewicz

, Gill

, Gkatzima

, Gómez-Tortosa

, Gordon

, Grimmer

, Hampel

, Hausner

, Hellwig

, Herukka

, Hildebrandt

, Ishihara

, Ivanoiu

, Jagust

, Johannsen

, Kandimalla

, Kapaki

, Klimkowicz-Mrowiec

, Klunk

, Köhler

, Koglin

, Kornhuber

, Kramberger

, Van Laere

, Landau

, Lee

, de Leon

, Lisetti

, Lleó

, Madsen

, Maier

, Marcusson

, Mattsson

, de Mendonça

, Meulenbroek

, Meyer

, Mintun

, Mok

, Molinuevo

, Møllergård

, Morris

, Mroczko

, Van der Mussele

, Na

, Newberg

, Nordberg

, Nordlund

, Novak

, Paraskevas

, Parnetti

, Perera

, Peters

, Popp

, Prabhakar

, Rabinovici

, Ramakers

, Rami

, Resende de Oliveira

, Rinne

, Rodrigue

, Rodríguez-Rodríguez

, Roe

, Rot

, Rowe

, Rüther

, Sabri

, Sanchez-Juan

, Santana

, Sarazin

, Schröder

, Schütte

, Seo

, Soetewey

, Soininen

, Spiru

, Struyfs

, Teunissen

, Tsolaki

, Vandenberghe

, Verbeek

, Villemagne

, Vos

, van Waalwijk van Doorn

, Waldemar

, Wallin

ÅK

, Wiltfang

, Wolk

, Zboch

, Zetterberg

(2015) Prevalence of cerebral amyloid pathology in persons without dementia: A meta-analysis. JAMA 313, 1924–1938.

35.

Wallin

, Blennow

, Andreasen

, Minthon

(2006) CSF biomarkers for Alzheimer’s disease: Levels of beta-amyloid, tau, phosphorylated tau relate to clinical symptoms and survival. Dement Geriatr Cogn Disord 21, 131–138.

36.

Nathan

, Lim

, Abbott

, Galluzzi

, Marizzoni

, Babiloni

, Albani

, Bartres-Faz

, Didic

, Farotti

, Parnetti

, Salvadori

, Müller

, Forloni

, Girtler

, Hensch

, Jovicich

, Leeuwis

, Marra

, Molinuevo

, Nobili

, Pariente

, Payoux

, Ranjeva

, Rolandi

, Rossini

, Schönknecht

, Soricelli

, Tsolaki

, Visser

, Wiltfang

, Richardson

, Bordet

, Blin

, Frisoni

; PharmaCog Consortium (2017) Association between CSF biomarkers, hippocampal volume and cognitive function in patients with amnestic mild cognitive impairment (MCI). Neurobiol Aging 53, 1–10.

37.

Bennett

, Schneider

, Wilson

, Bienias

, Arnold

(2004) Neurofibrillary tangles mediate the association of amyloid load with clinical Alzheimer disease and level of cognitive function. Arch Neurol 61, 378–384.