Cerebrospinal fluid tau and β-amyloid in Alzheimer patients, disease controls and an age-matched random sample

Neurobiology of Aging 27 (2006) 1202–1211

Cerebrospinal fluid tau and ␤-amyloid in Alzheimer patients, disease controls and an age-matched random sample Bernd Ibach a,∗ , Harald Binder a , Margarethe Dragon a , Stefan Poljansky a , Ekkehard Haen a , Eberhard Schmitz b , Horst Koch a , Albert Putzhammer a , Hans Kluenemann a , Wolf Wieland c , Goeran Hajak a b

a Department of Psychiatry, Psychosomatics and Psychotherapy, Geriatric Psychiatry Research Group, University of Regensburg at the Bezirksklinikum, Universit¨atsstraße 84, D-93053 Regensburg, Germany Department of Anaesthesiology, Cartias Hospital St. Joseph, Landshuter Straße 65, 93053 Regensburg, Germany c Department of Urology, Cartias Hospital St. Joseph, Landshuter Straße 65, 93053 Regensburg, Germany

Received 22 December 2004; received in revised form 15 June 2005; accepted 16 June 2005 Available online 8 August 2005

Abstract We prospectively evaluated the diagnostic accuracy of cerebrospinal fluid (CSF)-␤-amyloid1–42 (A␤42 ), -total-tau (tau) and -phosphorylatedtau181 (p-tau181 ) as measured by sandwich ELISAs in the clinical routine of a community state hospital to discriminate between patients with Alzheimer’s disease (AD), healthy controls (HC), non-AD-dementias, a group composed of various psychiatric disorders (non-AD-dementias, mental diseases) and an age-matched random sample (RS) (total N = 219). By comparing patients with AD to HC as reference, tau revealed sensitivity (sens)/specificity (spec) of 88%/80%, p-tau181 88%/80%, tau/A␤42 -ratio 81%/85% and phospho-tau181 /A␤42 -ratio 81%/78%. Discriminative power between HC and all dementias under investigation was estimated lower for tau (78%/77%) and p-tau181 (73%/79%). Relative to patients with AD, ROC analysis for the RS revealed highest sens/spec for p-tau181 (79%/77%) and p-tau181 /A␤42 ratio (78%/75%). Differentiation between AD versus a group made of patients with various psychiatric disorders was optimised by using CSF-p-tau181 (80%/77%). Under clinical routine conditions current CSF-biomarkers show a substantial capacity to discriminate between AD and HC as reference and to mark off AD patients from RS and heterogeneous diagnostic groups composed of non-AD dementias and other psychiatric conditions. Despite a residual substantial overlap between the groups, we conclude that current CSF markers are well suited to support AD-related diagnostic procedures in every-day clinics. © 2005 Elsevier Inc. All rights reserved. Keywords: Alzheimer’s disease; Non-Alzheimer dementias; Dementia; Psychiatric disorders; Depression; Schizophrenia; Cerebrospinal fluid; Tau; Phosphorylated tau; ␤-Amyloid42 ; Random sample

1. Introduction Alzheimer’s disease (AD) is diagnosed extensively by clinical criteria [19,32]. In searching for optimized diagnoses [5,25], many studies reported elevated concentrations of tauproteins and decreased levels of A␤42 peptides in the cerebrospinal fluid (CSF) of patients with AD [1,2,6,8,10,27,30]. Only a few studies report no significant change or even ∗

Corresponding author. Tel.: +49 941 941 2063; fax: +49 941 941 2079. E-mail addresses: [email protected], [email protected] (B. Ibach). 0197-4580/$ – see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.neurobiolaging.2005.06.005

an increase of CSF-A␤42 levels in AD [14,21]. Previous research emphasizes CSF tau, p-tau181,199,231 and A␤42 or their ratios as biological markers to differentiate clinical AD from healthy controls and from other dementias [4,9,13,17]. Besides CSF-banks, the AD-consensus report [7] emphasizes studies on brain banks, because neuropathological confirmation is defined as the gold standard for diagnosing AD. The validity of this approach has been challenged by a community-based clinico-neuropathological study, which showed that a substantial number of demented and non-demented elderly people exhibited Alzheimer’s-, Lewy body- and vascular pathology simultaneously [23]. Thus, a

B. Ibach et al. / Neurobiology of Aging 27 (2006) 1202–1211

concept of pure clinical Alzheimer’s disease that is based on pure Alzheimer’s pathology seemingly does not reflect the high prevalence of cerebral multimorbidity in the elderly and therefore would be a limiting factor for the use of the current biomarkers. Furthermore, results from CSF-studies that derive from neuropathologically validated dementias do not differ significantly from those, that are exclusively based on the clinical diagnosis of dementing illness [6,30]. The majority of the CSF-studies in neurodegenerative diseases are derived from highly specialized dementia centers and these studies suffer from a lack of participants [30]. More than 200 overall probands are exceptional [2,3,8,10,13,15,24,28]. A lot of normative data is derived from small numbers of self-selected controls or does not always report the clinical characterization of control persons precisely [2,9,27,28,30]. Very little information exists about the distribution of CSF-tau and -A␤42 in elderly populations [29]. Investigations of random samples to compare the natural distribution of putative biological markers directly to the target populations are lacking. In the present comparative study, we prospectively assessed CSF-A␤42 , -tau and -p-tau181 concentrations in a large community hospital-based cohort of patients with AD and other psychiatric diseases. By using neuropsychological and biological methods, we investigated an age-matched random sample, from which we extracted a clinically healthy control subset. Above all, our intention was to determine the diagnostic accuracy of current CSF markers in clinical routine in AD versus non-AD dementias and other conditions as present in a psychiatric state hospital.

2. Methods The investigation was performed in a State Hospital for Psychiatry and Psychotherapy (Bezirksklinikum Regens-

1203

burg, Germany). The predominantly rural catchment area of the hospital covers about 750,000 inhabitants. 2.1. Participants and clinical evaluation Two hundred and nineteen patients underwent a diagnostic procedure for suspected cognitive decline or dementia in the in-patient service and/or memory disorders outpatient clinic (Table 1). The comprehensive diagnostic work was comprised of a thorough clinical examination, including medical and family history; neurological, internal, and psychiatric examinations; routine laboratory testing; neuropsychological testing (Consortium to establish a registry for Alzheimer’s disease battery, CERAD [20], structured interview for the diagnosis of Alzheimer’s disease, multi-infarct or vascular dementia and dementias with other etiologies according to DSM-III-R, DSM-IV and ICD-10 = SIDAM [33]), selected neuropsychological tests; computed tomographic or magnetic resonance imaging of the brain and, if available, fluordeoxy-glucose-18 positron emission tomography, for which excellent group discrimination has been shown recently [11]. A random sample of 69 elderly in-patients with urological disease was examined in a general hospital (Caritas Krankenhaus, Regensburg) from which we extracted 39 psychiatric and neurological healthy controls. All probands received spinal anesthesia before minor surgical intervention (prostate hyperplasia, vesical condition). Work-up included medical history, evaluation of anesthesia and urologic information, cognitive assessment (SIDAM). Written informed consent was obtained from the patients, control subjects and, if appropriate, from caregivers. The local ethics committee of the University of Regensburg approved the study. The diagnosis was made by consensus of at least two experienced neuropsychiatrists who were not aware of the

Table 1 Cohort demographic data Diagnostic group

Patients/probands no. n = 219 Main groups

1. Random sample 2. Healthy controls 3. AD

69

4. Non-AD-dementias (a) FTD (b) PPA (c) LBD (d) Dementia not specified (e) Rare dementias

48

5. Mental disorders (a) Schizophrenia (b) Depression

26

Sex (M)

Age, mean (S.D.)

MMSE score, mean (S.D.)

0.87 0.97 0.4

67.8 (10.1) 65.9 (10.1) 67.1 (9.4)

27.8 (2.4) 29.0 (1.0) 20.8 (5.2)

17 4 6 13 8

0.25 0.47 0.5 0.5 0.38 0.62

62.9 (11.4) 58.7 (7.8) 70.0 (9.1) 70.2 (3.5) 65.2 (14.5) 59.1 (13.6)

21.0 (7.0) 23.6 (5.9) 13.7 (14.4) 19.2 (4.2) 22.2 (4.7) 18.0 (5.0)

11 15

0.54 0.73 0.4

49.3 (14.9) 45.7 (13.5) 51.9 (15.8)

n.a. n.a. n.a.

Subgroups 39

76

Basic clinical characteristics of the diagnostic groups. Abbreviations: MMSE indicates mini-mental state examination; AD, Alzheimer’s disease; FTD, frontotemporal degeneration (including frontotemporal dementia, semantic dementia, corticobasal degeneration); PPA, primary progressive aphasia; LBD, Lewy-body dementia. Diagnostic groups 4 and 5 = various psychiatric disorders.

1204

B. Ibach et al. / Neurobiology of Aging 27 (2006) 1202–1211

CSF results. One hundred and twenty-four patients fulfilled the diagnostic criteria for dementia (76 AD, 48 non-AD dementias cf. Table 1). The diagnosis of Alzheimer’s disease satisfied NINCDS-ADRDA for probable and the ICD10 criteria for AD [19,32]. Lewy-body dementia was diagnosed according to the Newcastle criteria [18], while dementias with frontotemporal lobar degeneration were grouped as proposed recently [22]. The rare dementias are comprised of polycystic lipomembranous osteodysplasia (NasuHakola-disease) (n = 1), Herpes-simplex encephalitis (n = 2), radiation-induced frontal lobe dementia (n = 1), Huntington’s disease (n = 1), Marchiafava–Bignami syndrome (n = 1), and alcohol-related dementia (n = 2). Dementias that could not be clinically specified further additionally underwent an extensive battery of various investigations that included immunologic parameters as well as such indicative for paraneoplastic conditions. Patients with schizophrenia and depression were diagnosed in accordance with the ICD-10 WHOclassification of disease. 2.2. Genetic mutations We used established methods (PCR, automated sequencing) to identify disease-causing mutations in the amyloid precursor protein and presenilin 1.

On the identical basis of the above-mentioned ELISA, a monoclonal antibody AT270, specific for tau proteins phosphorylated at threonine 181, was used (prototype version, INNOTEST Phospho-tau 181) [31]. The micro titer plates were coated with antibody HT7 (which recognizes aminoacids of epitopes 159–163 in normal tau and p-tau). 2.5. Statistical analysis Diagnostic accuracy was assessed by using receiver operating characteristic (ROC) curves. This method allows for the evaluation of specificity for different levels of sensitivity and vice versa. We determined cutoffs for tau, p-tau181 and their appropriate ratios with A␤42 , which resulted in a reasonable combination of sensitivity and specificity. As a global measure that is independent of a specific value for sensitivity/specificity the area under curve (AUC) will be given. Differences between mean tau and p-tau181 levels and the corresponding ratios with A␤42 were evaluated with a onefactor analysis of variance. The factor is ‘diagnosis group’ and p-values were obtained from pre-specified contrasts for the comparison of the healthy controls and/or the Alzheimer’s disease group to other diagnoses groups. All analyses were computed using SPSS 12.0.1.

2.3. Lumbar puncture

3. Results

Cerebrospinal fluid was obtained by means of a lumbar puncture in all patients. Subsequently, all probes were taken in polypropylene tubes to avoid absorption of A␤ molecules into the container surface and were centrifuged (2000 × g/10 min) prior to aliquotation, freezing and storage at −80 ◦ C within two hours, pending biochemical analysis.

3.1. Study cohort

2.4. Cerebrospinal fluid assays All assays for CSF-tau, -p-tau181 and -A␤42 (Innogenetics NV, Gent, Belgium) were run in the hospitals’ neurochemical laboratory. For visualization of the antibody binding, the biotin-streptavidin principle with tetramethylbenzidine chromogene was used. The CSF samples (25 ␮l) and the standards were assayed in duplicates. Absorbance was read at 450 nm on a microplate reader (TECAN Biophotometer). The intraand interassay coefficients of variations (CV) were less than 10%. A sandwich A␤42 ELISA (INNOTEST ␤-amyloid1–42 ) with a monoclonal antibody 21F12, specific for the free Cterminal end of A␤42 was used as capturing antibody. The monoclonal antibody 3D6 specific for the N-terminal end of A␤42 (␤-amyloid peptide1–5 ) was used as the detector. CSF tau was determined using an ELISA constructed to measure normal human tau concentrations (INNOTEST hTau-Ag). The phosphorylation independent AT120 anti-tau capture antibody and monoclonal antibodies HT7 and BT2 as reporter antibodies were used.

We included all probands (n = 219) in our analysis. Demographic data is outlined in Table 1. Decreased CSF-A␤42 values correlated with age (R = −0.273, P < 0.001) and with low MMSE-scores (R = 0.332, P < 0.001). An inverse correlation has been observed between CSF-tau and MMSE scores (R = −0.318, P < 0.001). The current literature on CSF markers of dementia reports on a modest correlation between clinical measures of age and CSF-total tau [30] or -A␤42 [6], which is assigned to the increasing AD prevalence with advancing age [29], while others have reported no significant relationship [1,2]. The AD group included three patients with autosomal dominant inheritance pattern, two of them belonging to a family with a mutation in the APP gene and one of them with mutation in the PS 1 gene, all with a mean age of disease onset at 47 years. The age of patients with AD matched perfectly with the probands included in the random sample and the healthy controls and consequently revealed no statistically significant differences. Patients grouped together as psychiatric disorders (non-AD dementias and mental diseases) were younger (Table 1). Our random sample and healthy control group was comprised predominantly of male subjects. However, this factor was not significantly associated with CSF-A␤42 , -tau or -ptau181 concentrations and in this respect confirms the results of previous studies [9,10,30].

B. Ibach et al. / Neurobiology of Aging 27 (2006) 1202–1211

3.2. CSF-Aβ42-, -tau and -p-tau181 levels All cerebrospinal fluid level calculations of A␤42 , tau, ptau181 and their appropriate ratios with A␤42 are illustrated in detail in the box plots in Fig. 1 and Table 2a and b. The mean values of A␤42 were lowest in patients with AD. Inversely, CSF-tau and -p-tau181 proteins were higher in AD as compared to the other groups. Table 3 contains CSF-values for diagnostic subgroups. 3.3. Contrast-matrix-analysis (including age and MMSE as covariates) All markers under investigation showed a high discriminating capacity between patients with AD and the HC group as reference (P < 0.001) (Table 2a). This capacity diminished by comparison to all dementias and the non-AD-dementia group, the latter reaching negligible statistical significance for p-tau181 and p-tau/A␤42 ratio (P = 0.15/0.054). Relative to the heterogeneous psychiatric disorders group p-values indicated no significant differences. Comparing the AD group as reference (Table 2b) to the random sample, non-AD-dementias and the group with various psychiatric disorders revealed highly significant differences for all markers under investigation (P < 0.001) but for A␤42 , which were less significant (PRandomSample = 0.02; Pnon-AD-dementias = 0.01; Pvarious psychiatric disorders = 0.003). 3.4. Sensitivity and specificity of the CSF-markers and ratios Receiver operating characteristic curves (ROC) for paired comparisons of A␤42 , tau, p-tau181 , tau/A␤42 -ratio, and ptau181 /A␤42 -ratio levels between HC and the groups with AD, all dementias, non-AD-dementias, as well as the various psychiatric diseases are shown in Fig. 2a. The ROC-curves for paired comparisons of all biomarkers between the AD group as reference and the random sample, patients with non-AD-dementias and the various psychiatric disorders group are exhibited in Fig. 2b. The group differences of the ratios (tau/A␤42 and p-tau181 /A␤42 ) added significant information when comparing patients with AD versus patients with non-AD-dementias (Table 2b). The corresponding area under curve values (AUC) for each group compared are also displayed in Table 2a and b and were estimated between HC ≥AD to ≥0.85 for tau, p-tau181 as well as tau/A␤42 - and p-tau181 /A␤42 -ratio. All other comparisons revealed AUC values < 0.85 (ranging from 0.53 to 0.84).

4. Discussion In the present data set, the measured biomarkers tau, ptau181 and their ratios with A␤42 satisfy the widely accepted recommendations of the consensus guidelines for AD markers [7]. Optimized sensitivity (sens) and specificity (spec) for

1205

AD versus healthy controls reached 80–88% for tau, p181 tau, tau/A␤42 - and p-tau181 /A␤42 -ratio. For A␤42 as a single marker maximal sens and spec did not exceed 69%/69%. These major findings agree with previous investigations, reporting sens/spec of 85%/84% for tau [6], sens/spec of 85%/86% for tau/A␤42 -ratio [10] and sens/spec of 85%/81% for p-tau181 [9]. Differences to other studies with lower or higher discriminating CSF-values may be attributable to selection criteria of the cohorts [1,13,17,27,28,30]. An advantage of our study is the inclusion of patients derived from clinical routine services. The patients represent a diagnostic compilation of consecutively admitted patients with suspected dementia and other psychiatric diseases, as they appear in the health care of community hospitals. According to our current information, data was usually obtained from admission populations of specialized centers, which are characterized by a large AD prevalence [1,2,6,8,10,27,30]. Comparing the AD group with the random sample revealed highest sens 79%/spec 77% for p181 -tau, which were estimated lower for all other biomarkers as measured in our study (63–78%/62–75%). On the one hand, such a minor diagnostic accuracy reflects a more realistic approach to the possibilities and limitations of a diagnostic marker for AD than the value of data that results from artificial contrasts, i.e. from self-selected controls. On the other hand, it is desirable to keep the background noise as low as possible for tau and A␤42 , which is achievable by defined controls. At least the discriminating power of p181 -tau even between the random sample and patients with AD still can be considered as sufficient for a diagnostic marker. To prevent useless testing of small samples we divided patients up into a non-AD dementia group characterized by a broad diagnostic spectrum, which included dementias with FTLD, DLB, Herpes simplex virus-encephalitis (HSV-E), Huntington disease (HD), Marchiafava–Bignami syndrome (MBS) and one with alcohol-related dementia (ARD), all of which exhibited elevated tau and p-tau181 levels. For small samples of patients with FTD [26] (moderate) and LBD [9] elevations of CSF tau-proteins have been described. The 13 cases of dementias that were not further specified presented features somewhere between AD, DLB, FTD or questionable cerebral amyloid-angiopathy. Because of this diagnostic uncertainty, we preferred not to categorize them as possible AD. As expected from previous evidence [6,9,13,17], comparison of the AD-group with non-AD dementias also resulted in a lower discriminative capacity of all three biomarkers, reaching highest sens of 74% and spec of 74% for p181 -tau and 78%/74% for p181 -tau/A␤42 -ratio. The marked heterogeneity of this non-AD dementia group would explain the relatively low diagnostic accuracy, but was not biased by specific research purposes (Table 3). Any diagnostic approaches using CSF analysis should consider that comparing isolated subgroups of dementia and other diseases do not represent realistic clinical settings. Therefore, we performed a more valid test and assembled patients with various psychiatric disorders (non-AD

1206

B. Ibach et al. / Neurobiology of Aging 27 (2006) 1202–1211

Fig. 1. Boxes represent the 25th, 50th (median), and 75th percentiles of the data. The length of the box is the interquartile range. The lower and upper horizontal line represent the smallest value above/the largest value below the 25th and 75th percentile plus or minus 1.5 times the interquartile range, respectively.

B. Ibach et al. / Neurobiology of Aging 27 (2006) 1202–1211

1207

Table 2 Summary of diagnostic measures Concentrations (pg/mL)

Ratios

A␤42

Tau

Phospho-tau181

Tau/A␤42

Phospho-tau181 /A␤42

731 (278)

249 (118)

51.8 (18.6)

0.41 (0.29)

0.09 (0.06)

Alzheimer’s disease (n = 76), mean (S.D.) AUC P-value Cut off Sensitivity Specificity

463 (218) 0.77 <0.001 530 69 69

615 (362) 0.90 <0.001 348 88 80

87.1 (28.9) 0.85 <0.001 65 88 80

1.60 (1.16) 0.90 <0.001 0.67 81 85

0.23 (0,13) 0.86 <0.001 0.132 81 78

All dementias (n = 124) (AD, non-AD-dementias), mean (S.D.) AUC P-value Cut off Sensitivity Specificity

524 (267)

546 (389)

76.7 (33.8)

1.34 (1.15)

0.193 (0.136)

0.70 <0.001 564 66 63

0.81 <0.001 263 78 77

0.73 <0.001 55 72 79

0.82 <0.001 0.47 75 73

0.76 <0.001 0.091 66 71

Non-AD-dementias (n = 48), mean (S.D.) AUC P-value Cut off Sensitivity Specificity

633 (307) 0.59 0.009 672 53 55

422 (404) 0.68 0.003 244 66 62

60.3 (34.9) 0.53 0.15 53 56 45

0.93 (1.00) 0.70 0.005 0.39 66 62

0.138 (0.123) 0.60 0.054 0.083 63 50

Psychiatric disorders except AD (n = 74) (non-AD-dementias, mental disorders), mean (S.D.) AUC P-value Cut off Sensitivity Specificity

660 (300)

343 (331)

54.3 (30.7)

0.7 (0.87)

0.105 (0.100)

0.54 0.05 692 50 50

0.59 0.26 238 53 50

0.46 0.67 50 50 48

0.59 0.09 0.32 63 59

0.51 0.29 0.065 53 53

(b) Diagnostic groups compared to Alzheimer’s disease Random sample (n = 69), mean (S.D.) AUC P-value Cut off Sensitivity Specificity

627 (276) 0.69 0.02 494 63 62

287 (149) 0.84 <0.001 370 78 73

53.6 (21.3) 0.83 <0.001 65 79 77

0.59 (0.50) 0.83 <0.001 0.81 75 75

0.11 (0.08) 0.82 <0.001 0.132 78 75

Non-AD-Dementias (n = 48), mean (S.D.) AUC P-value Cut off Sensitivity Specificity

633 (307) 0.66 0.01 540 71 57

422 (404) 0.74 <0.001 400 72 71

60.3 (34.9) 0.78 <0.001 69 74 74

1.34 (1.15) 0.75 <0.001 0.78 75 74

0.193 (0.136) 0.78 <0.001 0.130 78 74

660 (300)

343 (331)

54.3 (30.7)

0.7 (0.87)

0.105 (0.100)

0.72 0.003 550 70 64

0.81 <0.001 350 79 76

0.82 <0.001 64 80 77

0.82 <0.001 0.66 79 77

0.84 <0.001 0.116 79 76

(a) Diagnostic groups compared to healthy controls Healthy controls (n = 39), mean (S.D.)

Psychiatric disorders except AD (n = 74) (non-AD-dementias, mental disorders), mean (S.D.) AUC P-value Cut off Sensitivity (%) Specificity (%)

Diagnostic value of potential cerebrospinal fluid biomarkers. Abbreviations: A␤42 indicates ␤-amyloid protein at amino acid 42; AD, Alzheimer’s disease; S.D., standard deviation. The p-values are from an ANOVA with factor ‘diagnosis group’ and contrasts for comparison of all groups to the ‘healthy control’ (a) and ‘Alzheimer’s disease’ (b) group.

dementias, mental diseases) into one group. Our findings about lowered CSF-A␤42 values versus controls are attributable to the fact that we included non-AD dementias into this group [6,10,17]. The normal p181 -tau values

in patients with various psychiatric diseases, however, support the idea of a higher specificity for patients with AD [9]. Contrasting them to patients with AD resulted in maximal sens/spec of 80%/77% for p181 -tau. The discriminating

1208

B. Ibach et al. / Neurobiology of Aging 27 (2006) 1202–1211

Fig. 2. Area under the receiver operating characteristic curve indicating the discriminating ability of CSF-measures of amyloid-␤42 (grey points), tau (continuous grey line), phosphorylated tau181 (black points), the ratio of tau/amyloid-␤42 (continuous black line) and the ratio of phosphorylated tau181 /amyloid-␤42 (interrupted black line) in (a) healthy controls vs. patients with AD, all patients with dementia, non-AD-dementias, various psychiatric disorders (non-AD dementias, mental diseases) and (b) patients with AD vs. the random sample, non-AD dementias, various psychiatric disorders (non-AD dementias, mental diseases).

B. Ibach et al. / Neurobiology of Aging 27 (2006) 1202–1211

1209

Fig. 2. (Continued ).

capacity of p181 -tau/A␤42 -ratio (79%/76%), tau/A␤42 ratio (79%/77%) and tau (79%/76%) was marginally lower, but still close to the recommended guideline of 80% sens/spec for a biomarker in AD [7]. These results may have important

impact on clinical routine by supporting the clinical differential diagnosis of early onset AD facing other dementias with presenile onset and the manifestation of mental diseases as schizophrenic or depression in mid-adult age.

1210

B. Ibach et al. / Neurobiology of Aging 27 (2006) 1202–1211

Table 3 Mean values for diagnostic subgroups Diagnostic group

n

CSF-A␤42 mean (S.D.) (pg/mL)

CSF- tau mean (S.D.) (pg/mL)

CSF-phospho-tau181 mean (S.D.) (pg/mL)

FTD PPA LBD Schizophrenia Depression

17 4 6 11 15

734.8 (326.2) 511.8 (360.4) 468.8 (200.9) 732.2 (304.6) 749.0 (257.9)

314.1 (117.6) 925.5 (850.2) 575.3 (562.1) 206.1 (116.3) 191.9 (69.3)

49.0 (12.9) 97.8 (55.2) 76.6 (51.1) 48.2 (22.8) 40.5 (12.2)

Cerebrospinal fluid (CSF) levels of A␤42 , tau and p-tau181 as measured in the diagnostic subgroups. Abbreviations: A␤42 indicates ␤-amyloid protein at amino acid 42; AD, Alzheimer’s disease; FTD, frontotemporal degeneration (including Frontotemporal dementia, semantic dementia, corticobasal degeneration); PPA, primary progressive aphasia; LBD, Lewy-body dementia.

However, both our findings and previous ones support the assumption of a substantial overlap between CSF biochemical markers in AD and non-AD conditions. This observation may be explained by broad clinical and neuropathological similarities between dementia diseases as well as highly prevalent common neuropathological features found in both demented and non-demented elderly [23]. These overlapping features implicate a cerebral multimorbidity rather than the presence of pure dementia entities with increasing age. In summary, the advantages of doing this work in a clinical setting are large numbers, lack of selection bias and diverse clinical populations. Possible weakness may arise from diagnostic accuracy that may be blurred by the unrestricted inclusion of patients with heterogenous clinical dementia syndromes, which have not been neuropathologically confirmed. Nevertheless, our data supports the use of tau- and A␤42 proteins as a tool providing additional valuable information for the diagnosis of AD in clinical routine. Due to the complementary character of these biomarkers, it appears neither realistic nor necessary to chase sensitivities and specificities of 100% for current AD-biomarkers at the expense of an artificially narrowed concept of Alzheimer’s disease in highly selected probands [12,16]. Independent of their usefulness in supporting clinical diagnosis, our results furthermore point out the potential of biomarkers to track treatment effects in longitudinal trials.

Acknowledgements We thank Maria Blab, Heidi Kollmer-Liebl, Helene Niebling and Michaela Nowicki for their excellent clinical study support; Tatjana Nonenmacher and Doris Melchner for excellent technical support; Stephan Grubwinkler for carrying out many lumbar punctures; Matthias van Darl for providing the phosphorylated tau protein kits.

References [1] Andreasen N, Minthon L, Clarberg A, Davidsson P, Gottfries J, Vanmechelen E, et al. Sensitivity, specificity, and stability of CSF-tau in AD in a community-based patient sample. Neurology 1999;53:1488–94.

[2] Andreasen N, Minthon L, Davidsson P, Vanmechelen E, Vanderstichele H, Winblad B, et al. Evaluation of CSF-tau and CSF-A␤42 as diagnostic markers for Alzheimer disease in clinical practice. Arch Neurol 2001;58:373–9. [3] Andreasen N, Vanmechelen E, Vanderstichele H, Davidsson P, Blennow K. Cerebrospinal fluid levels of total-tau, phospho-tau and A beta 42 predicts development of Alzheimer’s disease in patients with mild cognitive impairment. Acta Neurol Scan (Suppl) 2003;179:47–51. [4] Buerger K, Teipel SJ, Zinkowski R, Blennow K, Arai H, Engel R, et al. CSF tau protein phosphorylated at threonine 231 correlates with cognitive decline in MCI subjects. Neurology 2002;59:627–9. [5] Chapel HM, Esiri MM, Wilcock GK. Immunoglobulin and other proteins in the cerebrospinal fluid of patients with Alzheimer’s disease. J Clin Pathol 1984;37:697–9. [6] Clark CM, Xie S, Chittams J, Ewbank D, Peskind E, Galasko D, et al. Cerebrospinal fluid tau and beta-amyloid: how well do these biomarkers reflect autopsy-confirmed dementia diagnoses? Arch Neurol 2003;60(12):1696–702. [7] 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. [8] Galasko D, Chang L, Motter R, Clark CM, Kaye J, Knopman D, et al. High cerebrospinal fluid tau and low amyloid beta42 levels in the clinical diagnosis of Alzheimer disease and relation to apolipoprotein E genotype. Arch Neurol 1998;55:937–45. [9] Hampel H, Buerger K, Zinkowski R, Teipel SJ, Goernitz A, Andreasen N, et al. Measurement of phosphorylated tau epitopes in the differential diagnosis of Alzheimer disease: a comparative cerebrospinal fluid study. Arch Gen Psychiatry 2004;61(1):95– 102. [10] Hulstaert F, Blennow K, Ivanoiu A, Schoonderwaldt HC, Riemenschneider M, De Deyn PP, et al. Improved discrimination of AD patients using beta-amyloid(1–42) and tau levels in CSF. Neurology 1999;52(8):1555–62. [11] Ibach B, Poljansky S, Marienhagen J, Sommer M, Manner P, Hajak G. Contrasting metabolic impairment in frontotemporal degeneration and early onset Alzheimer’s disease. Neuroimage 2004;23(2):739–43. [12] Ibach B, Haen E. Acetylcholinesterase inhibition in Alzheimer’s Disease. Curr Pharm Des 2004;10(3):231–51. [13] Itoh N, Arai H, Urakami K, Ishiguro K, Ohno H, Hampel H, et al. Large-scale, multicenter study of cerebrospinal fluid tau protein phosphorylated at serine 199 for the antemortem diagnosis of Alzheimer’s disease. Ann Neurol 2001;50(2):150–6. [14] Jensen M, Schroder J, Blomberg M, Engvall B, Pantel J, Ida N, et al. Cerebrospinal fluid A beta42 is increased early in sporadic Alzheimer’s disease and declines with disease progression. Ann Neurol 1999;45(4):504–11. [15] Kanai M, Matsubara E, Isoe K, Urakami K, Nakashima K, Arai H, et al. Longitudinal study of cerebrospinal fluid levels of tau, A beta1-

B. Ibach et al. / Neurobiology of Aging 27 (2006) 1202–1211

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

40, and A beta1-42(43) in Alzheimer’s disease: a study in Japan. Ann Neurol 1998;44(1):17–26. Klunemann HH, Fronhofer W, Wurster H, Fischer W, Ibach B, Klein HE. Alzheimer’s second patient: Johann F. and His family. Ann Neurol 2002;52(4):520–3. Maddalena A, Papassotiropoulos A, Muller-Tillmanns B, Jung HH, Hegi T, Nitsch RM, et al. Biochemical diagnosis of Alzheimer disease by measuring the cerebrospinal fluid ratio of phosphorylated tau protein to beta-amyloid peptide42. Arch Neurol 2003;60:1202–6. McKeith IG, Galasko D, Kosaka K, Perry EK, Dickson DW, Hansen LA, et al. Consensus guidelines for the clinical and pathologic diagnosis of dementia with Lewy bodies (DLB): report of the consortium on DLB international workshop. Neurology 1996;47(5):1113–24. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. 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. Morris JC, Heyman A, Mohs RC, Hughes JP, van Belle G, Fillenbaum G, et al. The consortium to establish a registry for Alzheimer’s disease (CERAD) Part I. Clinical and neuropsychological assessment of Alzheimer’s disease. Neurology 1989;39:1159–65. Nakamura T, Shoji M, Harigaya Y, Watanabe M, Hosoda K, Cheung TT, et al. Amyloid beta protein levels in cerebrospinal fluid are elevated in early-onset Alzheimer’s disease. Ann Neurol 1994;36(6):903–11. Neary D, Snowden JS, Gustafson L, Passant U, Stuss D, Black S, et al. Frontotemporal lobar degeneration: a consensus on clinical diagnostic criteria. Neurology 1998;51(6):1546–54. Neuropathology Group. Medical Research Council Cognitive Function and Aging S: Pathological correlates of late-onset dementia in a multicentre, community-based population in England and Wales. Neuropathology Group of the Medical Research Council Cognitive Function and Ageing Study (MRC CFAS). Lancet 2001;357:169–75. Nishimura T, Takeda M, Nakamura Y, Yosbida Y, Arai H, Sasaki H, et al. Basic and clinical studies on the measurement of tau protein in cerebrospinal fluid as a biological marker for Alzheimer’s disease and related disorders: multicenter study in Japan. Methods Find Exp Clin Pharmacol 1998;20(3):227–35.

1211

[25] Pardridge WM, Vinters HV, Miller BL, Tourtellotte WW, Eisenberg JB, Yang J. High molecular weight Alzheimer’s disease amyloid peptide immunoreactivity in human serum and CSF is an immunoglobulin G. Biochem Biophys Res Commun 1987;145: 241–8. [26] Riemenschneider M, Wagenpfeil S, Diehl J, Lautenschlager N, Theml T, Heldmann B, et al. Tau and Abeta42 protein in CSF of patients with frontotemporal degeneration. Neurology 2002;58 (11):1622–8. [27] Shoji M, Matsubara E, Kanai M, Watanabe M, Nakamura T, Tomidokoro Y, et al. Combination assay of CSF tau, A beta 1-40 and A beta 1-42(43) as a biochemical marker of Alzheimer’s disease. J Neurol Sci 1998;158(2):134–40. [28] Shoji M, Matsubara E, Murakami T, Manabe Y, Abe K, Kanai M, et al. Cerebrospinal fluid tau in dementia disorders: a large scale multicenter study by a Japanese study group. Neurobiol Aging 2002;23(3):363–70. [29] Sjogren M, Vanderstichele H, Agren H, Zachrisson O, Edsbagge M, Wikkelso C, et al. Tau and Abeta42 in cerebrospinal fluid from healthy adults 21–93 years of age: establishment of reference values. Clin Chem 2001;47(10):1776–81. [30] Sunderland T, Linker G, Mirza N, Putnam KT, Friedman DL, Kimmel LH, et al. Decreased beta-amyloid1-42 and increased tau levels in cerebrospinal fluid of patients with Alzheimer disease. JAMA 2003;289(16):2094–103. [31] Vanmechelen E, Vanderstichele H, Davidsson P, Van Kerschaver E, Van Der Perre B, Sjogren M, et al. Quantification of tau phosphorylated at threonine 181 in human cerebrospinal fluid: a sandwich ELISA with a synthetic phosphopeptide for standardization. Neurosci Lett 2000;285(1):49–52. [32] World Health Organization: Tenth Revision of the International Classification of Diseases, Chapter V (F): Mental and Behavioural Disorders (including disorders of psychological development). Clinical Descriptions and Diagnostical Guidelines. 2002. [33] Zaudig M, Mittelhammer J, Hiller W, Pauls A, Thora C, Morinigo A, et al. SIDAM—a structured interview for the diagnosis of dementia of the Alzheimer type, multi-infarct dementia and dementias of other aetiology according to ICD-10 and DSM-III-R. Psychol Med 1991;21:225–36.