CSF Amyloid-beta 38 as a novel diagnostic marker for dementia with Lewy bodies

Abstract

The clinical distinction between Alzheimer's disease (AD) and dementia with Lewy bodies (DLB) is sometimes difficult, particularly in mild cases. Although CSF markers such as amyloid β42 (Aβ42) and P-tau can distinguish between AD and normal controls, their ability to distinguish between AD and DLB is not adequate.
This study aims to investigate whether CSF markers, in particular levels of Aβ38, can differentiate between mild AD and DLB.
85 individuals were included after standardised diagnostic procedures: 30 diagnosed as probable AD, 23 probable DLB, 20 probable Parkinson's disease dementia and 12 non-demented control subjects. CSF levels of Aβ38, Aβ40 and Aβ42 were determined using commercially available ultra-sensitive multi-array kit assay (MSD) for human Aβ peptides. Total tau (T-tau) and phosphorylated tau (P-tau) were analysed using ELISA (Innotest). In addition, combinations (Aβ42/Aβ38, Aβ42/Aβ40, Aβ42/P-tau and Aβ42/Aβ38/P-tau) were assessed.
Significant between group differences were found for all CSF measures, and all except Aβ40, Aβ42 and Aβ42/P-tau differed between AD and DLB. The Aβ42/Aβ38 ratio was the measure that best discriminated between AD and DLB (AUC 0.765; p<0.005), with a sensitivity of 78% and a specificity of 67%.
This study suggests that the level of Aβ38 can potentially contribute in the diagnostic distinction between AD and DLB when combined with Aβ42. Single measures had low diagnostic accuracy, suggesting that developing a panel of markers is the most promising strategy. Studies with independent and larger samples and a priori cut-offs are needed to test this hypothesis.

Full text

1
CSF Amyloid-beta 38 as a novel diagnostic marker for dementia with Lewy bodies.
Ezra Mulugeta1,2, Elisabet Londos3, Clive Ballard2, Guido Alves4, Henrik Zetterberg5, Kaj
Blennow5, Ragnhild Skogseth6, Lennart Minthon3, *Dag Aarsland1
1 Department of Old Age Psychiatry, Psychiatric Clinic, Stavanger University Hospital, Norway,
2Wolfson Centre for Age Related Diseases, King’s College London, UK
3Research Unit, Department of Clinical Sciences, Malmö, University of Lund, Sweden
4 The Norwegian Centre for Movement Disorders, Stavanger University Hospital, Norway
5Institute of Neuroscience and Physiology, Department of Psychiatry and Neurochemistry, the
Sahlgrenska Academy at the University of Gothenburg, Molndal, Sweden.
6Institute of Clinical Medicine, University of Bergen, Norway
Correspondence to:
* Dag Aarsland
Department of Old Age Psychiatry
Psychiatric Clinic, Stavanger University Hospital
PO Box 8100
N-4068 Stavanger
Norway
E-mail: daarsland@gmail.com Telephone: +47 51 51 50 62, Fax: +47 51 51 55 15
Word count: Abstract: 245, Full paper including References 4027
Keywords: Dementia with Lewy bodies, biomarker, cerebrospinal fluid, A
β
38, A
β
42,T-tau, P-
tau Alzheimer’s disease, Parkinson’s disease
peer-00589971, version 1 - 3 May 2011
Author manuscript, published in "Journal of Neurology, Neurosurgery & Psychiatry 82, 2 (2010) 160"
DOI : 10.1136/jnnp.2009.199398

2
Abstract
Background
The clinical distinction between Alzheimer’s disease (AD) and dementia with Lewy bodies
(DLB) is sometimes difficult, particularly in mild cases. Although cerebrospinal fluid (CSF)
markers such as A
β
42 and P-tau can distinguish between AD and normal controls, their ability to
distinguish between AD and DLB is not adequate.
Objective
This study aims at investigating whether CSF markers, in particular the level of A
β
38, can
differentiate between mild AD and DLB.
Methods
In total 85 individuals were included after standardized diagnostic procedures: 30 diagnosed as
probable AD, 23 probable DLB, 20 with probable Parkinson’s disease dementia (PDD), and
12 non-demented controls subjects. CSF levels of A
β
38, A
β
40 and A
β
42 were determined using
commercially available Ultra-Sensitive multi-array kit assay (MSD) for human A
β
peptides.
Total-tau (T-tau) and Phosphorylated tau (P-Tau) were analysed using ELISA (Innotest). In
addition, combinations (A
β
42/A
β
38, A
β
42/A
β
40, A
β
42/P-tau, and A
β
42/A
β
38/P-tau) were
assessed.
Results
Significant between-group differences were found for all CSF measures, and all except A
β
40,
A
β
42, and A
β
42/P-tau differed between AD and DLB. A
β
42/A
β
38 ratio was the measure that
best discriminated between AD and DLB (AUC 0.781; p<0.005), with sensitivity 74% and
specificity 77%.
Conclusion
This study suggests that the level of A
β
38 can potentially contribute in the diagnostic distinction
between AD and DLB when combined with A
β
42. Single measures had low diagnostic accuracy,
suggesting that developing a panel of markers is the most promising strategy. Studies with
independent and larger samples and a priori cut-offs are needed to test this hypothesis.
peer-00589971, version 1 - 3 May 2011

3
Introduction
Dementia with Lewy Bodies (DLB) is a common dementia with a complex clinical presentation,
reduced quality of life and higher costs [1, 2], higher mortality [3], poorer drug response [4], and
increased risk for nursing home admission [5] compared to Alzheimer’s disease (AD).
Identifying people with DLB is therefore crucial, but can be difficult, in particular in early cases
where the clinical features are less characteristic.[6] There is therefore a need for biomarkers to
aid in the distinction between DLB and AD. However, with the exception of dopamine
transporter SPECT, which is expensive and not available at all centres, there are yet no such
established biomarkers.[7]
Potential biomarkers should be based on the underlying pathology of the disease. In AD,
accumulation of amyloid peptides (A
β
-peptides) in plaques as well as species of microtubule-
associated axonal protein tau, which is deposited in tangles, are key pathological features. The
A
β
-peptides are derived from amyloid precursor protein (APP) by sequential cleavage involving
proteolytic enzymes
β
- and
γ
secretaces.[8] The A
β
42 peptide variant is prone to aggregation, is
predominantly deposited in senile plaques and is shown to be neuro-toxic.[9]
Hyper phosphorylated tau (P-tau) is one of the major components of neurofibrillary tangles.[10]
Lewy bodies, consisting mainly of alpha-synuclein deposits, are the characteristic feature of
DLB, although AD type changes are also common.[11]
Cerebrospinal fluid (CSF) A
β
42 together with hyper-phosphorylated tau (P-tau) have been
shown to identify incipient AD with good accuracy.[12, 13] However, these peptides
discriminate less well between AD and other dementia subtypes, including DLB.[12, 13] In a
study using Western blot and quantitative analyses, Bibl and co-workers showed that by
expanding the number of amyloid species by including carboxy-terminally truncated peptides
such as A
β
37, A
β
38, and A
β
39, such markers might provide useful biomarker to distinguish AD
from other dementias.[14, 15] In subsequent studies using ELISA, the same group reported that
A
β
38 and the A
β
42/A
β
38 ratio, have better discriminative power between AD and other
dementias than A
β
42 alone.[15, 16] However, in the latter study, only five patients had DLB.
The objective of this paper was therefore to explore whether A
β
38 and combinations including
peer-00589971, version 1 - 3 May 2011

4
A
β
38 could distinguish between AD and DLB, using a larger sample size with DLB, and
including normal controls and Parkinson’s disease dementia (PDD) as comparison groups.
Methods
Subjects
Patients were drawn from two cohorts: The Norwegian DemWest study (n=42) recruited
referrals to psychiatry, neurology, and geriatric medicine clinics in Rogaland and Hordaland
counties who were diagnosed with mild dementia (i.e Mini-Mental State Examination (MMSE)
>20) between 2005-2007.[17] For this study, only patients diagnosed as probable AD (n=30),
probable DLB (n=9) or PDD (n=3) who consented to lumbar puncture were included. In
addition, patients with PDD (n=17) and probable DLB (n=14) who were screened between
2006-2008 to participate in a clinical trial of memantine [17] and who consented to lumbar
puncture were recruited from the Department of Clinical Sciences, Malmö, University of Lund,
Sweden. Subjects without known brain disease who underwent lumbar puncture during
orthopaedic surgery or neurologic outpatient assessment with a minimum MMSE score of >24 at
the Stavanger University Hospital were recruited as non-demented controls.
Diagnostic procedures
Diagnostic procedures are described in detail elsewhere.[17, 18] In brief, patients were
diagnosed as DLB if they fulfilled clinical consensus criteria for probable DLB [19], probable
AD [20] or probable PDD [21] after a detailed clinical assessment by a registered specialist in
psychiatry, neurology or geriatric medicine, using standardised assessments for
parkinsonism[22] psychiatric symptoms including visual hallucinations (Neuropsychiatric
Inventory, NPI)[23], and fluctuating cognition.[24, 25] Blood samples and brain imaging (CT or
MRI) were taken from all patients, and a subset of DLB patients underwent dopamine transporter
SPECT scans.
Pre-analytical treatment of CSF
Lumbar puncture (LP) was performed in the L3-L4 or L4-L5 interspace and CSF sampling was
performed in all cases between 7-10 am in order to minimize diurnal variation of the level of
peer-00589971, version 1 - 3 May 2011

5
CSF A
β
[26]. The first 3-4mLs of the CSF were dedicated for routine analyses for assessment of
relevant CSF abnormalities. Thus, samples were immediately sent on ice to the routine
laboratory where routine assay for cell counts, levels of glucose and protein were performed.
Study samples were collected in separate polypropylene tubes, and centrifuged at 2000xg, 4ºC
for 10 min to get rid of cell debris and other insoluble materials. Following centrifugation,
samples were aliquoted and immediately frozen at -80ºC until analyses were performed. Samples
from the DemWest study, Stavanger, were originally stored in larger volumes, thus the portions
of samples used in this study were aliquots derived from samples frozen and thawed (on ice)
once.
Tri-plex human CSF A
β
38, A
β
40 and A
β
42Assay
All CSF analyses were performed randomized and in duplicates the same day by one of the
authors (EM), blinded to clinical information. CSF levels of A
β
42, A
β
40, and A
β
38 were
determined using the Aβ triplex assay (Human A
β
peptide Ultra-Sensitive Kits) developed by
Meso Scale Discovery, Gaithersburg, Maryland, USA. This assay uses C-terminus specific
antibodies to capture the different Aβ peptides and a SULFO-TAG ™-labeled anti-Aβ
antibody (4G8) for detection with electrochemiluminescence.
The tri-plex assay was performed on CSF samples from patients and control subjects as well as
standards of specific markers in duplicate and according to the manufacturer’s instructions.
Briefly, the assay technology is based on MULTI-ARRAY® technology combining
electrochemiluminescence detection and patterned arrays offering combination of sensitivity and
dynamic range. The triplex assay mentioned here utilizes peptide specific antibodies to capture
A
β
38, A
β
40 and A
β
42 peptides present in CSF. The CSF content of each peptide was then
detected by SULFO-TAG–labeled 4G8 detection antibodies. The standard ranges for A
β
38
and A
β
42 were 4-3000pg/mL respectively, and for A
β
40 27.4-20000pg/mL. The lower limit
of detection and limit of quantitation (LLOD/LOQ) for all three analytes were A
β
38
(8.69/<25pg/mL), A
β
40 (1.28/~50pg/ml) and A
β
42 (12.37/~35pg/mL) pg/mL respectively. To
determine inter and intra assay variations we included different CSF samples (1= low level,
2=medium level) in replicates as run controls. Within-assay precision for replicated
samples on same plate (intra-assay) variation for the “low level” sample and for individual
peer-00589971, version 1 - 3 May 2011

6
analytes; A
β
38, A
β
40 and A
β
42 were between 5-6%. The inter-assay variability, (same
sample analysed on different plates) for the “low level” sample and each of the analytes
A
β
38, A
β
40 and A
β
42 was 9, 13 and 8% respectively. Similarly the intra-assay variability
for the “medium level sample” and each of the analytes mentioned above were between 4-
7% and the inter-assay variability for “medium level sample” and each of the analytes
were 15, 14 and 12% respectively.
ELISA of total and phosphorylated tau
CSF T-tau & P-tau (181)
CSF T-tau was analyzed using a commercial sandwich ELISA (INNOTEST® hTAU-Ag,
Innogenetics, Gent, Belgium) specifically constructed to measure all tau isoforms irrespective of
phosphorylation status, as previously described [27]. In this assay, the monoclonal antibody
AT120 was used for capture, while the biotinylated monoclonal antibodies HT7 and BT2 were
used as detection antibodies. AT120 and HT7 both react equally well with both normal and
hyperphosphorylated, while BT2 preferentially recognizes normal tau protein. The
standard range of the T-tau assay was 75-1200 pg/ml, the LOD was 59 pg/ml, and intra-
and inter-assay CVs ranged from 1.2-5.9% and 1.7-6.0% respectively.
CSF P-tau was measured using a commercial sandwich ELISA method (INNOTEST®
PHOSPHO-TAU(181P), Innogenetics, Ghent, Belgium), as described previously [28]. In this
assay, the monoclonal antibody HT7 was used for capture and the biotinylated monoclonal
antibody AT720 as detector antibody. HT7 both react equally well with both normal and
hyperphosphorylated, while AT270 specifically reacts with tau phosphorylated at
threonine-181. The P-tau assay had a standard range of 15.6–500 pg/ml, a LOD of 15.6
pg/ml, and intra- and inter-assay CVs of <5% and <10% respectively.
Statistics
Values of CSF markers were expressed as absolute (pg/ml). In addition to the single markers,
pre-specified combinations were analysed: A
β
42/ A
β
40, A
β
42/P-tau, A
β
42/A
β
38, and A
β
42/A
β
38/P-tau. Since measures were not normally distributed, median and interquartile range (IQR)
were expressed, and Spearman correlations were performed. Between-group comparisons were
peer-00589971, version 1 - 3 May 2011

7
made using Kruskall-Wallis and chi square tests. Pre-planned post-hoc pair-wise comparisons
(Mann-Whitney) between AD and DLB were performed subsequently. The global diagnostic
accuracies were assessed by the received-operated characteristic curve (AUC). Since this was a
hypothesis-generating study, a p value of < 0.05 was considered significant, and no attempts to
adjust for multiple comparisons were made. Cut-off points and sensitivity and specificity were
determined based on the coordinate points of the curves. Positive (sensitivity/1-specificity) and
negative (specificity/1-sensitivity) likelihood ratios (LR) were calculated. Cut-off points and
sensitivity and specificity were determined based on the coordinate points of the curves.
Results
Characteristics
In total, 85 subjects participated: AD (n=30), DLB (n=23), PDD (n=20) and normal controls
(n=12). The characteristics of the subjects are shown in table 1. The groups did not differ in
terms of age, and the dementia groups did not differ regarding MMSE score. As expected, there
were more males in the DLB and PDD groups compared to the other groups, and the AD and
DLB groups differed significantly (chi square 4.6, p=.03). The disease duration differed
significantly, with a longer duration in DLB than AD (p=.042).
Table 1. Characteristics of the groups
NC AD DLB PDD P
N 12 30 23 20
Age 73.5(16.8) 75.5(11.3) 74(10.8) 73(11) .89
MMSE 28.5(1.8) 23.5(4.3) 23(7.8) 23(9) <.0005
Gender, M/F* 4/8 12/18 16/7 13/7 =.055
UPDRS motor ND 0(2) 30.5(32.5) 34.5(17.3) <.0005
Duration of
disease
NA 2(2) 3.5(3) 8(4) <.0005
Numbers represent median and inter-quartile range or *number of people and %
P values based on Kruskall-wallis test; ND: Not done NA: not applicable
NC=Normal control, AD=Alzheimer’s disease, DLB=Dementia with Lewy bodies,
PDD=Parkinson’s disease dementia
peer-00589971, version 1 - 3 May 2011

8
CSF measures and associations with diagnosis and other characteristics
The CSF values are shown in table 2 and Figure 1. There were significant between-group
differences on all markers. Compared to NC; the pattern in AD was as expected, with low A
β
42
and high T-tau and P-tau. Significant differences between DLB and AD were observed for all
measures except A
β
40 and A
β
42, and A
β
42/P-tau. Whereas the single markers were changed in
the same direction in AD and DLB, a different picture emerged for the combined markers
involving A
β
38, which were decreased in AD but increased in DLB and PDD compared to NC.
Mean CSF concentrations did not differ between the Norwegian and Swedish DLB/PDD
patients. The AUC analyses demonstrated that the A
β
42/A
β
38 ratio was the strongest marker
for differentiation between AD and DLB, with an AUC of 0.732 (95% CI 0.587-0.876) (p=.007),
with sensitivity 74% and specificity 67% at cut-off 0.50. Positive LR was 2.2 (95% CI 1.3 – 3.6
and negative LR 0.3 (0.25-0.8).
There was a trend towards association between A
β
38 and gender (p=0.061). Since there were
gender differences between AD and DLB, the analyses were therefore performed for each gender
separately. The overall analyses were confirmed in males, whereas no significant differences
were found in the smaller female group (data not shown). When all subjects were included, age
correlated with total tau (rho 0.29, p=.007), A
β
38 (0.24, p=.029), and A
β
40 (0.26, p=.015), and
all three A
β
species correlated significantly with MMSE score (rho 0.24-0.31, p 0.004 – 0.03).
Finally, A
β
38 correlated with duration of disease (0.25, p=.04). Including patients only,
significant correlations with age were found for total tau (rho =0.25, p<.05), A
β
42
correlated with UPDRS motor score (0.37, p=.04), and duration with A
β
38 (rho=-0.26,
p=.032), whereas a non-significant trend between MMSE and A
β
38 was found (rho=0.22,
p=.06).
peer-00589971, version 1 - 3 May 2011

9
Table 2. CSF concentration of protein markers in the diagnostic groups
Markers NC AD DLB PDD P1 P2 AUC
A
β
38 (pg/mL) 635(712) 440(305) 385(244) 404(299) .005 .013 0.70
A
β
40 (pg/mL) 8286(6652) 5461(2632) 5507(2109) 5036(3633) .048 .36 0.43
A
β
42 (pg/mL) 337(378) 192(74) 223(163) 287(107) .001 .84 0.52
T-tau (pg/mL) 250(166) 382(363) 303(116) 303(122) .023 .009 0.68
P-tau (pg/mL) 58.3(35.4) 86.5(61) 60.2(52.8) 57.1(41.3) .027 .011 0.69
A
β
42/A
β
38 0.62(0.14) 0.46(0.29) 0.66(0.34) 0.74(0.28) .0005
.001 0.77
A
β
42/A
β
40 0.058(0.02) 0.032(0.02) 0.043(0.02) 0.055(0.1) .002 .004 0.64
A
β
42x1000/A
β
38/P-tau 9.2(6.3) 4.8(9.5) 10.8(11.1) 11.4(15.2) .0005
.015 0.74
A
β
42/P-tau 9.0(2.9) 2.1(2.0) 3.9(5.0) 4.8(2.7) .0005
.076 0.71
NC = non-demented control subjects, AD = Alzheimer’s disease, DLB = Dementia with Lewy
Bodies and PDD = Parkinson’s Disease Dementia
Numbers represent median and interquartile rnage (pg/ml) or ratio
P1=All groups, Kruskall-wallis test; p2=AD vs DLB, Mann-Whitney test
AUC: Area under the curve.
Discussion
We investigated whether CSF markers could distinguish between mild DLB and AD. The main
finding was that the ratio between A
β
42 and A
β
38 was the CSF marker that best distinguished
between AD and DLB. However, the accuracy is still well below the 85% level which is
recommended by Consensus Report of the Working Group on: “Molecular and Biochemical
Markers of Alzheimer’s Disease” for a diagnostic test.[29] Given the increasingly recognised
need for an early dementia diagnosis [30], and the difficulties in making an accurate clinical
diagnosis of DLB, in particular early in course, these findings are nevertheless encouraging. Our
findings further suggest that the pattern from a panel of CSF markers is the best strategy forward,
although continued research is needed to develop an adequate biomarker for the differentiation
between AD and DLB.
peer-00589971, version 1 - 3 May 2011

10
There were associations between A
β
-peptide species with cognitive impairment, and A
β
38
correlated with disease duration. This is consistent with a recent study demonstrating that the
CSF level of alpha-synuclein was associated with dementia severity in DLB,[31] and suggests
that CSF markers can be useful markers of the progression of disease pathology, and thus might
serve as biological outcome measures in clinical trials of disease modifying therapies.
This study has limitations which need to be considered when interpreting the findings. First, the
diagnoses were based on clinical assessment and may thus not be 100% accurate. However, the
two research groups have long experience in diagnosing DLB, and standardised and
recommended diagnostic procedures and consensus criteria were used. These have shown good
clinico-pathologic correlation [32], suggesting that the diagnostic accuracy is relatively high. Of
note, there is also pathological overlap between AD and DLB, and the pathological classification
of the two diseases is still under discussion.[33]
There were gender differences in the expected direction. The overall findings were confirmed in
the male patients, supporting the validity of the overall findings. The lack of significant
differences in the female patients should be interpreted in light of the low statistical power due to
the low number of female DLB patients. Difference in duration of disease might also have
introduced a bias. Lumbar puncture was not standardised in relation to meals at one study
site (Malmø). This might have influenced the findings [26], although such an effect is most
likely to be minor as no differences in mean values were observed in the DLB/PDD group
between study centres.
Finally, as most previous studies, we derived cut-offs from the population under study. This
procedure has an inherent risk for overestimating diagnostic accuracy.[34] Further studies are
therefore need to test the accuracy of a priori determined cut-offs rather than establishing a cut-
off directly on the cohort under investigation.
The observation of a similar pattern of CSF markers in DLB and PDD, with largely shared
underlying brain changes, adds biological validity to the observations. However, we found
somewhat lower diagnostic accuracies for the A
β
species compared with a previous study.[16].
Furthermore, our findings of Tau and P-Tau, although significantly different between the groups,
peer-00589971, version 1 - 3 May 2011

11
show lower diagnostic accuracy compared to a recent study [35]. Possible explanations for these
differences include that we specifically explored the differentiation between AD and DLB,
whereas only 10% of the non-AD group in the previous study were DLB or PDD. Thus, since
AD-type brain changes are typically found in DLB [36], this differentiation may be more
challenging than distinguishing AD from frontotemporal and vascular dementias. Secondly, our
sample included older patients (age 73-75) with mild dementia, with a mean MMSE score of 23-
24, compared to younger patients (mean age below 70) with a lower mean MMSE score (below
20) in [16]. Whereas the previous study (35) was based on an autopsy-cohort, we used clinically
diagnosed cases with presumed lower diagnostic accuracy. In addition, there were differences
in the pre-analytic handling and assay methods between the different studies, possibly
leading to variation in CSF marker concentrations between different centres. This was
shown in a recent multi-centre study [12], in which some of the Stavanger cases
participated. Finally, although there are strong correlations between the different methods
and commercially available kits used to determine level of CSF markers, there is
considerable variability in the average values of results.
In conclusion, our findings suggest that A
β
38 should be included in future studies to identify
CSF biomarkers to differentiate between AD and DLB. Future studies should also explore
whether CSF markers can increase diagnostic accuracy over and above standard clinical
assessment, and whether they can predict disease progression. Finally, whether establishing CSF
panels combining specific A
β
-peptides and tau markers with alpha-synuclein species might
improve this differentiation should also be investigated.
peer-00589971, version 1 - 3 May 2011

12
Acknowledgements
We would like to thank all patients in Western Norway and Malmö, Sweden and control subjects
for their willingness to participate in the study. We also thank all personnel (in the different
clinics in Western Norway and Malmö, Sweden) for their effort to contribute to this study as
well as for collecting clinical data and CSF. We are especially grateful to Hilde Rydland
Marianayagam, and Ingrid Langeland Braut for their assistance in collection of CSF from control
patients. Special thanks to Sara Hulberg, Sahlgrenska Academy for help with preparation of CSF
samples and analyses. Karen Simonsen for excellent administrative support. This study was
funded by the Western Norway Regional Health Authority, HelseVest (grant# 911390).
Licence for Publication
The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf
of all authors, an exclusive licence (or non exclusive for government employees) on a worldwide
basis to the BMJ Publishing Group Ltd to permit this article (if accepted) to be published in
JNNP and any other BMJPGL products and sublicences such use and exploit all subsidiary
rights, as set out in our licence. (http://group.bmj.com/products/journals/instructions-
for-authors/licence-forms)
Competing Interests
None declared.
peer-00589971, version 1 - 3 May 2011

13
REFERENCES
1. Bostrom F, Jonsson L, Minthon L, et al. Patients with dementia with lewy bodies have
more impaired quality of life than patients with Alzheimer disease. Alzheimer Dis Assoc Disord.
2007;21:150-4.
2. Bostrom F, Jonsson L, Minthon L, et al. Patients with Lewy body dementia use more
resources than those with Alzheimer's disease. Int J Geriatr Psychiatry. 2007;22:713-9.
3. Williams MM, Xiong C, Morris JC, et al. Survival and mortality differences between
dementia with Lewy bodies vs Alzheimer disease. Neurology. 2006;67:1935-41.
4. Aarsland D, Perry R, Larsen JP, et al. Neuroleptic sensitivity in Parkinson's disease and
parkinsonian dementias. J Clin Psychiatry. 2005;66:633-7.
5. Rongve A, Skogseth R, Aarsland D. Risk of nursing home placement in Dementia with
Lewy Bodies. ICAD 2009; 2009; Vienna, Austria: The Journal of The Alzheimer's Association;
2009.
6. Tiraboschi P, Salmon DP, Hansen LA, et al. What best differentiates Lewy body from
Alzheimer's disease in early-stage dementia? Brain. 2006;129:729-35.
7. Aarsland D, Kurz M, Beyer M, et al. Early discriminatory diagnosis of dementia with
Lewy bodies. The emerging role of CSF and imaging biomarkers. Dement Geriatr Cogn Disord.
2008;25:195-205.
8. Mattson MP. Pathways towards and away from Alzheimer's disease. Nature.
2004;430:631-9.
9. Iwatsubo T, Odaka A, Suzuki N, et al. Visualization of A beta 42(43) and A beta 40 in
senile plaques with end-specific A beta monoclonals: evidence that an initially deposited species
is A beta 42(43). Neuron. 1994;13:45-53.
10. Blennow K, Vanmechelen E, Hampel H. CSF total tau, Abeta42 and phosphorylated tau
protein as biomarkers for Alzheimer's disease. Mol Neurobiol. 2001;24:87-97.
11. Lippa CF, Duda JE, Grossman M, et al. DLB and PDD boundary issues: diagnosis,
treatment, molecular pathology, and biomarkers. Neurology. 2007;68:812-9.
12. Mattsson N, Zetterberg H, Hansson O, et al. CSF biomarkers and incipient Alzheimer
disease in patients with mild cognitive impairment. Jama. 2009;302:385-93.
13. Blennow K. CSF biomarkers for mild cognitive impairment. J Intern Med. 2004;256:224-
34.
peer-00589971, version 1 - 3 May 2011

14
14. Bibl M, Mollenhauer B, Esselmann H, et al. CSF amyloid-beta-peptides in Alzheimer's
disease, dementia with Lewy bodies and Parkinson's disease dementia. Brain. 2006;129:1177-87.
15. Bibl M, Mollenhauer B, Lewczuk P, et al. Validation of amyloid-beta peptides in CSF
diagnosis of neurodegenerative dementias. Mol Psychiatry. 2007;12:671-80.
16. Welge V, Fiege O, Lewczuk P, et al. Combined CSF tau, p-tau181 and amyloid-beta
38/40/42 for diagnosing Alzheimer's disease. J Neural Transm. 2009;116:203-12.
17. Aarsland D, Londos E, Ballard C. Parkinson's disease dementia and dementia with Lewy
bodies: different aspects of one entity. Int Psychogeriatr. 2009;21:216-9.
18. Aarsland D, Rongve A, Nore SP, et al. Frequency and case identification of dementia
with Lewy bodies using the revised consensus criteria. Dement Geriatr Cogn Disord.
2008;26:445-52.
19. McKeith IG, Dickson DW, Lowe J, et al. Diagnosis and management of dementia with
Lewy bodies: third report of the DLB Consortium. Neurology. 2005;65:1863-72.
20. McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of Alzheimer's disease:
report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and
Human Services Task Force on Alzheimer's Disease. Neurology. 1984;34:939-44.
21. Emre M. Dementia associated with Parkinson's disease. Lancet Neurol. 2003;2:229-37.
22. Fahn S ER, & Committee., M. o. t. U. D. Unified Parkinson's Disease Rating Scale. NJ:
Florham Park, NJ: MacMillan Health Care Information; 1987.
23. Cummings JL, Mega M, Gray K, et al. The Neuropsychiatric Inventory: comprehensive
assessment of psychopathology in dementia. Neurology. 1994;44:2308-14.
24. Ferman TJ, Smith GE, Boeve BF, et al. DLB fluctuations: specific features that reliably
differentiate DLB from AD and normal aging. Neurology. 2004;62:181-7.
25. Walker MP, Ayre GA, Cummings JL, et al. Quantifying fluctuation in dementia with
Lewy bodies, Alzheimer's disease, and vascular dementia. Neurology. 2000;54:1616-25.
26. Bateman RJ, Wen G, Morris JC, et al. Fluctuations of CSF amyloid-beta levels:
implications for a diagnostic and therapeutic biomarker. Neurology. 2007;68:666-9.
27. Blennow K, Wallin A, Agren H, et al. Tau protein in cerebrospinal fluid: a biochemical
marker for axonal degeneration in Alzheimer disease? Mol Chem Neuropathol. 1995;26:231-45.
peer-00589971, version 1 - 3 May 2011

15
28. Vanderstichele H, De Vreese K, Blennow K, et al. Analytical performance and clinical
utility of the INNOTEST PHOSPHO-TAU181P assay for discrimination between Alzheimer's
disease and dementia with Lewy bodies. Clin Chem Lab Med. 2006;44:1472-80.
29. Consensus report of the Working Group on: "Molecular and Biochemical Markers of
Alzheimer's Disease". The Ronald and Nancy Reagan Research Institute of the Alzheimer's
Association and the National Institute on Aging Working Group. Neurobiol Aging. 1998;19:109-
16.
30. Dubois B, Feldman HH, Jacova C, et al. Research criteria for the diagnosis of
Alzheimer's disease: revising the NINCDS-ADRDA criteria. Lancet Neurol. 2007;6:734-46.
31. Ballard C, Jones EL, Londos E, et al. alpha-synuclein antibodies recognize a protein
present at lower levels in the CSF of patients with dementia with Lewy bodies. Int Psychogeriatr.
2009:1-7.
32. Fujishiro H, Ferman TJ, Boeve BF, et al. Validation of the neuropathologic criteria of the
third consortium for dementia with Lewy bodies for prospectively diagnosed cases. J
Neuropathol Exp Neurol. 2008;67:649-56.
33. Dickson DW, Fujishiro H. In dementia with Lewy bodies, Braak stage determines
phenotype, not Lewy body distribution. Neurology. 2008;70:2087-8; author reply 8-9.
34. Bossuyt PM, Reitsma JB, Bruns DE, et al. Towards complete and accurate reporting of
studies of diagnostic accuracy: the STARD initiative. Bmj. 2003;326:41-4.
35. Koopman K, Le Bastard N, Martin JJ, et al. Improved discrimination of autopsy-
confirmed Alzheimer's disease (AD) from non-AD dementias using CSF P-tau(181P).
Neurochem Int. 2009;55:214-8.
36. Mori H. Pathological substrate of dementia in Parkinson's disease--its relation to DLB
and DLBD. Parkinsonism Relat Disord. 2005;11 Suppl 1:S41-5.
peer-00589971, version 1 - 3 May 2011

Figure 1. Diagnostic accuracy for DLB versus AD for A
β
42/A
β
38 ratio and A
β
42 as
measured using the area under the curve
Note. A
β
42 A
β
38ratio is the ratio between CSF A
β
42 and A
β
38.
A
β
42_MSD_pgml is the CSF level of A
β
42 in pg/ml.
peer-00589971, version 1 - 3 May 2011