Longitudinal stability of CSF tau levels in Alzheimer patients

Longitudinal Stability of CSF Tau Levels in Alzheimer Patients Trey Sunderland, Benjamin Wolozin, Douglas Galasko, James Levy, Ruth Dukoff, Marcel Bahro, Robert Lasser, Ruth Motter, Terho Lehtima¨ki, and Peter Seubert Background: Antemortem levels of tau in the cerebrospinal fluid (CSF) of Alzheimer’s disease (AD) patients have repeatedly been demonstrated to be elevated when compared to controls. Although CSF tau has been reported to be elevated even in very mild AD, it is unknown how tau levels change during the course of the disease. Methods: We have followed 29 mild-to-moderately affected AD subjects over 2 years with repeated CSF taps. Clinical measures of dementia severity (Clinical Dementia Rating Scale, Global Deterioration Scale and Mini-Mental Status Examination) were obtained at the start and conclusion of the observation period, and CSF tau was measured with a standard enzyme-linked immunoabsorbent assay (ELISA) using two monoclonal antibodies. Results: Despite significant changes in the clinical measures consistent with progression of the disease, no significant overall change in CSF tau levels (548 ⫾ 355 vs. 557⫾275 pg/mL, NS) was observed. None of the clinical variables was significantly correlated with either baseline measures of CSF tau or delta CSF tau (last-first). Similarly, CSF tau at baseline and changes over time were not significantly related to Apolipoprotein E (APO E) phenotype. Conclusions: These data suggest that CSF tau levels are stable over extended periods of time in a group of mild-to-moderately demented AD subjects and that CSF tau levels do not predict the severity or rate of progression of AD, at least not during the middle stages of the illness. Biol Psychiarty 1999;46:750 –755 Key Words: Cerebrospinal fluid, tau, dementia, surrogate marker

From the Geriatric Psychiatry Branch, National Institute of Mental Health, Bethesda, Maryland (TS, BW, JL, RD, MB, RL); Department of Neurosciences, University of California, San Diego, California (DG); Department of Clinical Chemistry, University of Tampere, Tampere, Finland (TL); and Athena Neurosciences, San Francisco, California (RM, PS). Address reprint requests to Trey Sunderland, MD, Bldg. 10, Rm. 3N228, NIMH, NIH, Clinical Center, 10 Center Dr. MSC 1275, Bethesda, MD 20892-1275. Received July 17, 1998; revised April 12, 1999; accepted June 4, 1999.

Introduction

R

esearch criteria for the diagnosis of Alzheimer’s disease (AD) have existed for over 15 years (McKhann et al 1984; American Psychiatric Association 1987), but the definitive diagnosis of AD is still relatively elusive. Although the clinical diagnosis of AD can be highly accurate, the only way to confirm the diagnosis is pathologically by autopsy or in rare circumstances with antemortem cortical biopsy (Mendez et al 1992; DeKosky et al 1992; Victoroffet al 1995). The recent discoveries of elevated cerebrospinal fluid (CSF) measures of the microtubule-associated protein tau in AD (Motter et al 1995; Arai et al 1995; Vigo-Pelfrey et al 1995; Vandermeeren et al 1993; Munroe et al 1995; Jensen et al 1995; Tato et al 1995) have generated considerable interest in the possibility of biologic markers as an antemortem diagnostic aid in dementia evaluations. Because abnormally phosphorylated tau is a major building block of the paired helical filaments found to accumulate in the neurofibrillary tangles (NFTs) of AD and there is a correlation between dementia severity and numbers of NFTs in AD brains (Terry 1994; Samuel et al 1994), it has been suggested that elevations of CSF tau might reflect the pathologic process in the brains of AD subjects (Galasko et al 1995). Although it might be appealing to attribute the rise in CSF tau to the dystrophic neurites and NFTs associated with AD, there have not yet been adequate studies of such clinicopathologic correlations. Furthermore, it does seem that the elevated CSF tau is not simply a measure of dying neurons, because CSF levels of the more general marker, neuron specific enolase, have not been found to be elevated in AD subjects as compared with controls (Parnetti et al 1995). Another unknown is whether CSF tau levels change with the progression of AD and thereby offer a biological index of the illness. In this study, we report on the course of CSF tau in a group of clinically diagnosed AD subjects who have been followed for several years.

Methods and Materials Patient Population Twenty-nine patients with probable AD (11 M/18 F) were included in this longitudinal cerebrospinal fluid study. These

0006-3223/99/$20.00 PII S0006-3223(99)00149-3

CSF Tau in Alzheimer’s Disease

subjects were diagnosed clinically based on DSM-III-R criteria (American Psychiatric Association 1987) and were followed at the NIH Clinical Center as part of a longitudinal study. No subjects met criteria for diffuse Lewy Body dementia (McKeith et al 1996). At the time of entry into the study, the subjects (mean age in years ⫽ 67.7 ⫾ 6.6 SD) were highly educated (mean years ⫽ 15.5 ⫾ 2.9 SD) and mild-to-moderately impaired, but they were able to give informed consent for the lumbar puncture procedure (LP) either individually or with the assistance of a previously assigned surrogate with durable power of attorney for research purposes. Informed consent was similarly obtained for the follow-up LPs 2 years later. The hospital ethics committee had approved all procedures. An additional group of 5 AD subjects were evaluated and tested in a similar fashion at the University of California, San Diego (UCSD). Repeat LPs in their cases were performed within a 12-week period to evaluate the short-term stability of the CSF measures only.

Clinical Rating Instruments Clinical rating instruments used to assess the global functioning of the subjects included the Clinical Dementia Rating (CDR) (Hughes et al 1982) Global Deterioration Scale (GDS) (Reisberg et al 1982) and the Mini-Mental State Examination (MMSE) (Folstein et al 1975). Trained inpatient staff administered these well-established rating instruments within 1 month of each LP for those subjects who participated in the NIH longitudinal study.

Medication Treatment The 29 AD subjects tested at the NIMH were part of a longitudinal study of a MAO-B inhibitor and a commonly available ergotamine. Each subject received both medications for 1 year in a random sequence during the course of the 2-year follow-up period. Previous analysis had determined that no clinical differences were noted between treatment groups at any point during the crossover trial (data on file). Other sporadic medications included occasional low-dose benzodiazepines, antihypertensives and intermittent analgesics. No subjects received chronic cholinesterase inhibitors or nonsteroidal anti-inflammatory agents during the 2-year trial.

Cerebrospinal Fluid Procedure All subjects in the longitudinal follow-up study were inpatients at the time of the LP and had been on a standardized low monoamine diet for at least 1 week. After an overnight fast, patients were placed on their side for administration of local anesthesia and insertion of a 20-gauge needle in the lumbar area (generally L4-L5). Fluid was removed in a standard fashion and frozen in aliquots on dry ice within 1 min of extraction and then placed in a ⫺70°C freezer until the time of assay. The specimens for CSF tau measurement for this study were taken from the first 5-mL aliquot of the CSF tap in all subjects.

CSF Assay CSF tau was measured with a two-site sandwich enzyme-linked immunoabsorbent assay (ELISA) using two monoclonal antibod-

751

BIOL PSYCHIATRY 1999;46:750 –755

Table 1. Follow-Up Study of 29 Alzheimer Patients Over Two Years

n Age (Range) Duration CDR GDS MMSE CSF Tau (pg/ml) (Range)

Baseline

Follow-up

29 67.7 ⫾ 6.6 (56 – 80) 4.1 ⫾ 2.0 1.6 ⫾ 0.5 4.4 ⫾ 0.7 18.6 ⫾ 6.1 548 ⫾ 355 (217–2083)

29 70.0 ⫾ 6.6 (58 – 81) 6.3 ⫾ 2.0 2.2 ⫾ 0.6 5.6 ⫾ 0.6 11.3 ⫾ 7.6 557 ⫾ 275 (327–1757)

F

p

37.8 69.2 33.3 0.15

.0001 .0001 .0001 NS

ies (16B5 and 16B7), previously determined to be gamma 1 kappa isotypes specific for tau. This assay has been shown to measure tau with a detection limit of approximately 25 pg/mL, independent of the tau phosphorylation state (Vigo-Pelfrey et al 1995). Each assay was run in duplicate with an intra assay variation less than 8%. The full details of the ELISA assay are described elsewhere (Vigo-Pelfrey et al 1995). CSF tau measures were performed on frozen samples that had been continuously stored at ⫺70°C for up to 10 years from the time of collection.

APO E Phenotyping APO E phenotypes were determined by isoelectric focusing of delipidated serum followed by immunoblotting with polyclonal anti-APO E antibodies. This method was described by Menzel and Utermann (1986) and then modified by Lehtima¨ki et al (1990, 1995). All serum samples had been frozen continually at ⫺70°C until the time of the assay.

Statistical Analysis Longitudinal CSF tau data was analyzed by one-way analysis of variance with repeated measures (ANOVAR) using the BMDP 2V program. Baseline and delta CSF tau levels were compared across APO E phenotypes by ANOVA; p-values for all ANOVA’s were adjusted for nonhomogeneity of variance with the Greenhouse-Geiser correction. Pearson’s product-moment correlations were performed to explore the relationship between baseline CSF tau and dementia severity measures, demographic data and change in CSF tau over time. Data are presented as mean ⫾ SD of the mean unless otherwise noted.

Results CSF tau measures in 29 clinically diagnosed AD subjects were remarkably stable (548 ⫾ 355 vs. 557 ⫾ 275 pg/mL) over a more than 2-year period of follow-up (Table 1). Over this same time interval, clinical measures of dementia severity, including the CDR, MMSE, and GDS all revealed significant changes consistent with a progression

752

T. Sunderland et al

BIOL PSYCHIATRY 1999;46:750 –755

of dementia (Table 1). Interestingly, neither the baseline level nor the change in CSF tau was significantly correlated with onset age, education level, and duration of illness or baseline measures of dementia severity (CDR, MMSE, and GDS). There was a significant positive correlation of CSF tau levels at baseline and the follow-up point over 2 years later (r2 ⫽ .872, y ⫽ .746⫻ ⫹ 125.172, p ⬍ .001) (Figure 1). Subjects were divided and compared according to whether their baseline CSF tau level was above or below the mean for the group (548 pg/mL). There were no significant differences between the “low tau” and “high tau” groups (n ⫽ 15 and 14, respectively) along their baseline measures of illness duration or dementia severity (GDS, CDR or MMSE) although the two groups did reveal some differences in CSF tau over time. Specifically, the “low tau” group demonstrated a significant increase in CSF tau levels over the 2-year period (364 ⫾ 85 vs. 427 ⫾ 81 pg/mL, F(1,12) ⫽ 12.33, p ⫽ .0027) whereas the “high tau” group tended to show a decrease in CSF tau over time (848 ⫾ 426 vs. 771 ⫾ 346 pg/mL, F(1,11) ⫽ 2.40, p ⫽ .15). As a consequence, there was a significant negative relationship between baseline level of CSF tau and the change in tau over the 2-year period of study in these 29 subjects (r ⫽ ⫺0.72, p ⬍ .009). Because the CSF samples had been stored for some time before the CSF tau assay was performed, we also measured the time lapse between CSF tap and assay for both time points and found a significant difference (7.5 ⫾ 2.0 vs. 5.1 ⫾ 2.0 years, F(1,28) ⫽ 256, p ⬍ .0001). Investigating whether the measurement of CSF tau might be influenced by this difference, we found no significant correlation between this time lapse and either the baseline (r ⫽ ⫺0.13, NS) or follow-up tau values (r ⫽ .08, NS). In a separate group of five subjects from UCSD, we also tested whether CSF tau would reveal any short-term changes over a period of 12 weeks or less. Once again, there were no significant differences when the two tau values were compared in this smaller group (475 ⫾ 98 [range ⫽ 339 – 610] vs. 499 ⫾ 154 [range ⫽ 348 –749] pg/mL, NS). To examine the possible relationship between the APO E4 allele and CSF findings, we first assessed the distribution of APO E phenotypes in our population. Consistent with the published literature (Motter et al 1995; Arai et al 1995; Strittmatter et al 1993), we found an overall E4 frequency of 50% for our 29 subjects [E4/E4 (n ⫽ 10), E3/E4 (n ⫽ 8), E2/E4 (n ⫽ 1), E3/E3 (n ⫽ 8), E2/E3 (n ⫽ 2)]. Next, we analyzed whether the absence (n ⫽ 10) or presence (n ⫽ 19) of E4 in individual subjects influenced either the initial level or the change in CSF tau over time, and the findings were remarkably similar (Delta tau (E4-negative) ⫽ 17 ⫾ 103 pg/mL vs. Delta tau (E4-

positive) ⫽ 6 ⫾ 150 pg/mL; F(1,87) ⫽ 0.05, NS). Even when the sample was partitioned according to the total number of E4 alleles present in individual subjects, the comparison of delta tau measures across groups was negative as well (E4 [none] ⫽ 17 ⫾ 103 pg/mL, E4 (1 copy) ⫽ ⫺13 ⫾ 169 pg/mL, E4 (2 copies) ⫽ 22 ⫾ 137 pg/mL, F(2,20) ⫽ 0.18, NS, for 10, 9 and 10 subjects, respectively.

Discussion Elevations in CSF measures of tau in AD patients have recently been reported by several groups (Motter et al 1995; Arai et al 1995; Vigo-Pelfrey et al 1995; Vandermeeren et al 1993; Jensen et al 1995; Tato et al 1995; Hock et al 1995; Mori et al 1995; Blomberg et al 1996; Arai et al 1997; Riemenschneider et al 1996; Galasko et al 1997). We have extended this investigation by asking whether total CSF tau changes over time in a group of 29 well-characterized AD patients. Interestingly, we found that CSF tau levels remain stable over a period in excess of 2 years despite a significant clinical progression of the illness. Furthermore, there were no correlations between baseline clinical measures and either initial CSF tau levels or the amount of CSF tau change over time. In trying to explain these data, several methodological questions come to mind. First, it is obvious that a sample of 29 represents a small cross-section of AD patients and may be insufficient to detect a subtle change in CSF change in a slowly progressive illness. Although the possibility of a type II statistical error cannot be ruled out, it appears clear that there is no striking or uniform shift in CSF tau levels over a 2-year period in this population. Other groups recently reported either no change (Andreasen et al 1998) or an increase in CSF tau (Kanai et al 1998) when AD subjects were followed longitudinally. Second, it is possible that testing mostly mild-to-moderately demented subjects might have serendipitously led to a study sample at a point in the AD progression where CSF tau levels are relatively stable. Perhaps CSF tau levels show more dramatic changes in the prodrome and very early stages of AD or, alternatively, in the later stages of the illness where CSF tau levels have been shown to have a modest correlation with postmortem markers of neurofibrillary tangles (Tapiola et al 1997). These questions will have to be addressed in future studies. Third, it is possible that the concurrent medications may have influenced the CSF tau levels. As noted in the methods section, however, all medications were either administered sporadically or in a counter-balanced fashion, and there were no group differences either clinically or biologically at any point during the follow-up period. Finally, we were concerned that time lag between the CSF procedure and the CSF tau

CSF Tau in Alzheimer’s Disease

Figure 1. Correlation of baseline and follow-up of CSF Tau levels in 29 Alzheimer patients over approximately 2 years time.

assay might have influenced our results, because the samples obtained at the second time point were more proximal to the time of the assay. There was no significant relationship between freezer shelf-life and measures of CSF tau in either this group or a large pool of previously studied subjects (P. Seubert, personal communication), suggesting that our time differences cannot explain the stability of CSF tau over 2 years. If CSF tau is indeed a stable measure over time, at least during the early and middle stages of AD, how can we explain this phenomena? Even casual review of the data reveals a large variance in means, both at the start and finish of our study. This observation led us to consider possible subgroups within our population, and the APO E isoform was an obvious first choice for subtyping. There were no suggestions of baseline or longitudinal CSF tau differences between those who were E4 (⫹) and E4 (⫺), and there was also no trend for differences amongst those who had zero, 1 or 2 E4 alleles, data that is supported by most (Arai et al 1997; Skoog et al 1996; Galasko et al 1998) but not all reports in the literature (Tapiola et al 1997; Golombowski et al 1997). In another published longitudinal study, Blomberg and colleagues report a small but significant increase in CSF levels of tau over a 14-month period in 18 AD patients (Blomberg 1996). A change of 5% was found to be significant despite an 11% intra assay variation and a greater than 50% standard deviation in the measurement of tau levels; however, these significant differences were not seen in the mean values but reported only when the CSF results were analyzed categorically as increased or decreased over time and stratified by APO E4 allele status. By contrast, when our CSF tau changes over time were dichotomized as in-

BIOL PSYCHIATRY 1999;46:750 –755

753

creased or decreased and compared by APO E subtype, there were no significant differences noted. Furthermore, the mean CSF tau changes in our report with a larger sample size are less than 2% with an intra assay variability of 8% and a similar 50% standard deviation in tau levels. Nonetheless, the relatively small size of both studies suggests that further investigation is needed to definitively answer this question. Other potential predictors of change such as duration, baseline severity, educational levels, and onset age were also not related to baseline tau levels or CSF tau change over time. The lack of baseline correlation between CSF tau levels and clinical measures is in contrast to some (Tato et al 1995; Hock et al 1995; Skoog et al 1996) but not all published reports (Motter et al 1995; Arai et al 1997), perhaps reflecting the relatively narrow window of severity seen in most of these reports. In the current study, only the baseline level of CSF tau itself was highly associated with change in CSF tau over time (Figure 1). Those patients with the lowest CSF tau levels at the beginning of the study showed a significant rise in the marker with time whereas the patients with the higher levels at the start of the study tended to decrease over the 2-year period, perhaps suggesting a regression toward the mean with repeat observations over time. Despite the recent plethora of papers describing an increase in CSF tau, the biologic explanation for this finding is not yet clear. Although it is known that brain tau increases in AD (Khatoon et al 1992; Ghanbari et al 1990; Wolozin et al 1987; Mandelkow and Mandelkow 1993; Baumann et al 1993), presumably as a result of the neuronal destruction and development of neurofibrillar tangles, there are reports of clinically diagnosed AD patients who go to autopsy without demonstrating abnormally elevated numbers of tangles or dystrophic neurites (Terry et al 1987). The CSF tau noted in our assay has recently been shown to be primarily a proteolytically derived fragment (Johnson et al 1998), thereby complicating studies of the mechanism of tau release into CSF and analysis of the phospho-isoforms that may initially give rise to the protein-bound tau. Furthermore, because our assay of CSF tau measures tau independent of its phosphorylation state, it is possible that phosphorylated tau alone might follow a different pattern over time than the tau fragments described in this report. It is also unknown whether the rate at which tau is released into the CSF is variable across different stages of the dementia process and if it is influenced by other factors such as glutamate-induced neurotoxicity (Pizzi et al 1995, 1994) or rate of tau fibrillization by sulfated glycosaminoglycans (Goedert et al 1996). Nonetheless, our data suggest relative stability in CSF tau levels over the middle stages of the illness, but even there, it appears that the rate

754

BIOL PSYCHIATRY 1999;46:750 –755

of change may be influenced by the CSF tau level at the start of the observation period more than by any global clinical variables. If CSF tau does, in fact, reflect an underlying neurodegenerative process in AD, then there is potential utility for this test as a surrogate marker in therapeutic trails aimed at slowing the course of the illness, but more longitudinal data are needed. The observation that CSF tau levels are fairly consistent within subjects for a period of time that is longer than most drug trials is encouraging in this regard. Finally, it is not known if the persistent elevation of CSF tau is related to the ongoing tangle-forming process or neuronal degeneration; this question will best be answered through careful clinicopathological correlation of antemortem CSF tau levels and post-mortem pathology, including subjects with diffuse Lewy Body dementia. These studies are ongoing.

The authors would like to thank Nina Peltonen for her skillful technical help and Wilma Davis for typing the manuscript. Dr. Lehtima¨ki’s contribution was supported by grants from the Paulo Foundation, the Finnish Foundation of Cardiovascular Research and the Emil Aeltoner Foundation.

References American Psychiatric Association (1987): Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised. Washington: American Psychiatric Association. Andreasen N, Vanmechelen E, Van de Voorde A (1998): Cerebrospinal fluid tau protein as a marker for Alzheimer’s disease: A community based follow up study. J Neurol Neurosurg Psychiatry 64:298 –305. Arai H, Higuchi S, Sasaki H (1997): Apolipoprotein E genotyping and cerebrospinal fluid tau protein: Implications for the clinical diagnosis of Alzheimer’s disease. Gerontology 43(suppl 1):2–10. Arai H, Terajima M, Miura M, et al (1995): Tau in cerebrospinal fluid: a potential diagnostic marker in Alzheimer’s disease. Ann Neurol 38:649 – 652. Baumann K, Mandelkow E-M, Biernat J, et al (1993): Abnormal Alzheimer-like phosphorylation of tau-protein by cyclindependent kinases cdk2 and cdk5. FEBS Lett 336:417– 424. Blomberg M, Jensen M, Basun H, et al (1996): Increasing cerebrospinal fluid tau levels in a subgroup of Alzheimer patients with apolipoprotein E allele E4 during 14 months follow-up. Neurosci Lett 214:163–166. DeKosky ST, Harbaugh RE, Schmitt FA, et al (1992): Cortical biopsy in Alzheimer’s disease: Diagnostic accuracy and neurochemical, neuropathological, and cognitive correlations. Ann Neurol 32:625– 632. Folstein MF, Folstein SE, McHugh PR (1975): “Mini-mental state”: A practical method for grading the cognitive state of patients for the clinician. J Psychiatry Res 12:189 –198. Galasko D, Chang L, Motter R, et al (1998): High cerebrospinal fluid tau and low amyloid Beta42 levels in the clinical

T. Sunderland et al

diagnosis of Alzheimer disease and relation to apolipoprotein E genotype. Arch Neurol 55:937–945. Galasko D, Clark C, Chang L, et al (1997): Assessment of CSF levels of tau protein in mildly demented patients with Alzheimer’s disease. Neurology 48:632– 635. Galasko D, Hansen L, Vigo-Pelfrey C, et al (1995): Antemortem CSF Tau is related to neuronal pathology at autopsy in Alzheimer’s disease. Soc Neurosci Abstr 21:581. Ghanbari HA, Miller BE, Haigler HJ, et al (1990): Biochemical assay of Alzheimer’s disease—associated protein(s) in human brain tissue. A clinical study. JAMA 263:2907–2910. Goedert M, Jakes R, Spillantini MG, et al (1996): Assembly of microtubule-associated protein tau into Alzheimer-like filaments induced by sulphated glycosaminoglycans. Nature 383:550 –553. Golombowski S, Muller-Spahn F, Romig H, et al (1997): Dependence of cerebrospinal fluid tau protein levels on apolipoprotein E4 allele frequency in patients with Alzheimer’s disease. Neurosci Lett 225:213–215. Hock C, Golombowski S, Naser W, et al (1995): Increased levels of tau protein in cerebrospinal fluid of patients with Alzheimer’s disease— correlation with degree of cognitive impairment. Ann Neurol 37:414 – 415. Hughes CP, Berg L, Danziger WL, et al (1982): A new clinical scale for the staging of dementia. Br J Psychiatry 140:566 – 572. Jensen M, Basun H, Lannfelt L (1995): Increased cerebrospinal fluid tau in patients with Alzheimer’s disease. Neurosci Lett 186:189 –191. Johnson GVW, Seubert P, Cox TM, et al (1998): The ␶ protein in human cerebrospinal fluid in Alzheimer’s disease consists of proteolytically derived fragments. Int J Neurochem 68: 430 – 433. Kanai M, Matsubara E, Isoe K, et al (1998): Longitudinal study of cerebrospinal fluid levels of tau, abeta1– 40, and abeta1– 42(3) in Alzheimer’s disease: A study in Japan. Ann Neurol 44:17–26. Khatoon S, Grundke-Iqbal I, Iqbal K (1992): Brain levels of microtubule-associated protein ␶ are elevated in Alzheimer’s disease: A radioimmuno-slot-blot assay for nanograms of the protein. J Neurochem 59:750 –753. Lehtima¨ki T, Moilanen T, Viikari J, et al (1990): Apolipoprotein E phenotypes in Finnish youths: a cross-sectional and 6-year follow-up study. J Lipid Res 31:487– 495. Lehtima¨ki T, Pirttila T, Mehta PD, et al (1995): Apolipoprotein E (apoE) polymorphism and its influence on ApoE concentrations in the cerebrospinal fluid in Finnish patients with Alzheimer’s disease. Hum Genet 95:39 – 42. Mandelkow E-M, Mandelkow E (1993): Tau as a marker for Alzheimer’s disease. Trends Biochem Sci 18:480 – 483. McKeith IG, Galasko D, Kosaka K, et al (1996): Consensus guidelines international workshop for the clinical and pathological diagnosis of dementia with Lewy bodies (DLB): Report of the consortium on DLB. Neurology 47:1113–1124. McKhann G, Drachman D, Folstein M, et al (1984): Clinical diagnosis of Alzheimer’s disease: report of the NINCDSADRDA work group under the auspices of Department of Health and Human Services task force on Alzheimer’s disease. Neurology 34:939 –944.

CSF Tau in Alzheimer’s Disease

Mendez MF, Mastri AR, Sung JH, et al (1992): Clinically diagnosed Alzheimer disease: Neuropathologic findings in 650 cases. Alzheimer Dis Assoc Disord 6:35– 43. Menzel H-J, Utermann G (1986): Apolipoprotein E phenotyping from serum by Western blotting. Electrophoresis 7:492– 495. Mori H, Hosoda K, Matsubara E, et al (1995): Tau in cerebrospinal fluids: Establishment of the sandwich ELISA with antibody specific to the repeat sequence in tau. Neurosci Lett 186:181–183. Motter R, Vigo-Pelfrey C, Kholodenko D, et al (1995): Reduction of ␤-amyloid peptide42 in the cerebrospinal fluid of patients with Alzheimer’s disease. Ann Neurol 38:643– 648. Munroe WA, Southwick PC, Chang L, et al (1995): Tau protein in cerebrospinal fluid as an aid in the diagnosis of Alzheimer’s disease. Ann Clin Lab Sci 25:207–217. Parnetti L, Palumbo B, Cardinali L, et al (1995): Cerebrospinal fluid neuron-specific enolase in Alzheimer’s disease and vascular dementia. Neurosci Lett 183:43– 45. Pizzi M, Valerio A, Arrighi V, et al (1994): Antisense strategy unravels tau proteins as molecular risk factors for glutamateinduced neurodegeneration. Cell Mol Neurobiol 14:569 –578. Pizzi M, Valerio A, Arrighi V, et al (1995): Inhibition of glutamate-induced neurotoxicity by a tau antisense oligonucleotide in primary culture of rat cerebellar granule cells. Eur J Neurosci 7:1603–1613. Reisberg B, Ferris SH, DeLeon MJ, et al (1982): The Global Deterioration Scale for Assessment of Primary Degenerative Dementia. Am J Psychiatry 139:1136 –1139. Riemenschneider M, Buch K, Schmolke M, Kurz A, Guder WG (1996): Cerebrospinal protein tau is elevated in early Alzheimer’s disease. Neurosci. Lett 212:209 –211. Samuel W, Terry RD, Maslich E (1994): Frontal cortex and nucleus basalis pathology in Alzheimer’s disease. Arch Neurol 51:772–778.

BIOL PSYCHIATRY 1999;46:750 –755

755

Skoog I, Vanmechelen E, Andreasson LA, et al (1996): A population-based study of tau protein and ubiquitin in cerebrospinal fluid in 85-year-olds: Relation to severity of dementia and cerebral atrophy, but not to the apolipoprotein allele E4 during 14 months follow-up. Neurosci Lett 214: 163–166. Strittmatter WJ, Saunders AM, Schmechel D, et al (1993): Apolipoprotein E: High-avidity binding to beta-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc Natl Acad Sci USA 90:1977–1981. Tapiola T, Overmyer M, Lehtovirta M, et al (1997): The level of cerebrospinal fluid tau correlates with neurofibrillary tangles in Alzheimer’s disease. Neuroreport 8:3961–3963. Tato RE, Frank A, Hernanz A (1995): Tau protein concentrations in cerebrospinal fluid of patients with dementia of the Alzheimer type. J Neurol Neurosurg Psychiatry 59:280 –283. Terry RD (1994): Neuropathological changes in Alzheimer disease. In: Svennerholm L, Asbury AK, Reisfeld RA, et al, editors. Progress in Brain Research. Amsterdam: Elsevier, pp 303–390. Terry RD, Hansen LA, DeTeresa R, et al (1987): Senile dementia of the Alzheimer type without neocortical neurofibrillarly tangles. J Neuropathol Exp Neurol 46:2262–2268. Vandermeeren M, Mercken M, Vanmechelen E, et al (1993): Detection of tau proteins in normal and Alzheimer’s disease cerebrospinal fluid with a sensitive sandwich enzyme-linked immunosorbent assay. J Neurochem 61:1828 –1834. Victoroff J, Mack WJ, Lyness SA, et al (1995): Multicenter clinicopathological correlation in dementia. Am J Psychiatry 152:1476 –1484. Vigo-Pelfrey C, Seubert P, Barbour R, et al (1995): Elevation of microtubule-associated protein tau in the cerebrospinal fluid of patients with Alzheimer’s disease. Neurology 45:788 –793. Wolozin B, Davies P (1987): Alzheimer-related neuronal protein A68: Specificity and distribution. Ann Neurol 22:521–526.