CSF biomarkers in frontotemporal lobar degeneration with known pathology

/Home/Alzheimer's Diagnosis/Non-Alzheimer's Dementia Diagnosis/Frontotemporal Dementia (FTD) Diagnosis/Frontotemporal Dementia (FTD) Biomarkers

OBJECTIVE: To evaluate the diagnostic value of CSF biomarkers in patients with known pathology due to frontotemporal lobar degeneration (FTLD). BACKGROUND: It is important to distinguish FTLD from other neurodegenerative diseases like Alzheimer disease (AD), but this may be difficult clinically because of atypical presentations. METHODS: Patients with FTLD (n = 30) and AD (n = 19) were identified at autopsy or on the basis of genetic testing at University of Pennsylvania and Erasmus University Medical Center. CSF was obtained during a diagnostic lumbar puncture and was analyzed using assays for total tau and amyloid-beta 1-42 (A beta(42)). Patients also were assessed with a brief neuropsychological battery. RESULTS: CSF total tau level and the ratio of CSF total tau to A beta(42) (tau/A beta(42)) were significantly lower in FTLD than in AD. Receiver operating characteristic curve analyses confirmed that the CSF tau/A beta(42) ratio is sensitive and specific at discriminating between FTLD and AD, and is more successful at this than CSF total tau alone. Although some neuropsychological measures are significantly different in autopsy-proven FTLD and AD, combining these neuropsychological measures with CSF biomarkers did not improve the ability to distinguish FTLD from AD. CONCLUSIONS: The ratio of CSF tau/A beta(42) is a sensitive and specific biomarker at discriminating frontotemporal lobar degeneration from Alzheimer disease in patients with known pathology.


Bian H, Van Swieten JC, Leight S, Massimo L, Wood E, Forman M, Moore P, de Koning I, Clark CM, Rosso S, Trojanowski J, Lee VM, Grossman M. CSF biomarkers in frontotemporal lobar degeneration with known pathology. Neurology. 2008 May 6;70(19 Pt 2):1827-35.


FULL PAPER

It is important to distinguish between progressive neurodegenerative conditions such as frontotemporal lobar degeneration (FTLD) and Alzheimer disease (AD) for many reasons. They appear to have distinct underlying causes that may require different etiologically based treatments, for example, and patient counseling would benefit since clinical features such as survival may differ in these conditions. There are two definite ways to confirm the diagnosis of FTLD. One depends on brain pathology at autopsy; the other identifies a genetic mutation known to cause FTLD. In 1998, mutations in the microtubule-associated protein tau (MAPT) gene were first discovered in some FTDP-17 families,1-3 and patients with MAPT mutations consistently have tau-positive inclusion pathology at autopsy.4,5 More recently, mutations of the progranulin (PGRN) gene were confirmed as the cause of autosomal dominant tau-negative FTLD with ubiquitin-positive but tau-negative inclusions (FTLD-U) linked to chromosome 17.6,7 In this study, we examine the usefulness of CSF biomarkers obtained during life in distinguishing patients with autopsy evidence or with a genetic mutation causing known FTLD pathology from AD.

Among currently available CSF biochemical markers of neurodegenerative diseases, the protein tau appears most promising for differentiating FTLD from AD. Several studies of CSF tau levels have been reported in patients clinically diagnosed with FTLD. The results have been inconsistent. Some studies found statistically increased CSF tau levels,8-10 while others reported normal CSF tau concentrations.11 In our previous clinical study, we found that CSF tau levels are significantly lower in FTLD than AD. Moreover, a receiver operating characteristic (ROC) curve analysis showed that tau is more sensitive than amyloid or isoprostane at distinguishing between FTLD and AD.12 Differences between reported studies may be due in part to the fact that clinical diagnosis does not always correspond to pathologic findings. Diagnosis may be difficult because of atypical presentations, such as frontal variant AD13,14 or FTLD presenting with memory difficulty.15 The current study examines CSF biomarker levels in cases with known pathology found at autopsy or genetic ascertainment. We hypothesized that FTLD would have a lower CSF tau level and lower tau/Aβ42 ratio than AD.

Attempts to distinguish FTLD and AD during life also have focused on cognitive biomarkers. AD is characterized by memory decline, for example, while changes in personality/behavior and progressive aphasia are the most prominent features of FTLD.16 However, these broad descriptions have proven to be only modestly reliable in clinical-pathologic correlation studies.17,18 Neuropsychological measures such as category naming fluency, confrontation naming, and episodic memory appear to differ in autopsy-proven FTLD and AD, so we also assessed whether neuropsychological results can improve the diagnostic value of CSF biomarkers.


METHODS

Subjects

From among 113 patients with CSF samples, we identified 36 in whom we also had known pathologic evidence for a specific diagnosis. This included patients with FTLD determined at autopsy (n = 15), FTLD with a known genetic mutation (n = 2), and autopsy-proven AD (n = 19) recruited from the Department of Neurology at the University of Pennsylvania. CSF samples from another 13 patients with autopsy-proven FTLD (n = 4) or a genetically determined mutation (n = 9) causing FTLD were obtained from the Department of Neurology at Erasmus University Medical Center of the Netherlands. The clinical diagnosis of an FTLD spectrum disorder or AD was established through a consensus mechanism following a clinical evaluation by an experienced neurologist and was consistent with the results of serum studies (e.g., normal levels on studies such as B12, thyroid functioning, and sedimentation rate), the absence of clinically significant MRI or CT changes suggesting vascular pathology or hydrocephalus, normal clinical studies of CSF (when available), and functional neuroimaging studies such as SPECT or PET (when available). The demographic characteristics of these subjects are summarized in table 1. In the FTLD series, histopathologic and genetic mutation diagnoses were as follows: 13 tau-positive patients, including corticobasal degeneration (CBD, n = 2), dementia with Pick bodies (PiD, n = 1), tangle predominant senile dementia (TPSD, n = 1), FTDP-17 (n = 1), G272V mutation (n = 2), and P301L mutation (n = 6); and 17 tau-negative patients, including FTLD-U (n = 9), dementia lacking distinctive histopathology (DLDH, n = 3), FTLD with motor neuron disease (FTLD-MND, n = 2), and PGRN mutation (n = 3).



________
Table 1
Demographic features and CSF biomarker levels of patients with known pathology due to FTLD or AD
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2707002&rendertype=table&id=T1
________



CSF from 13 cognitively normal seniors established the range of normal CSF tau and Aβ42. All patients and their legally authorized caregivers participated in an informed consent procedure approved by the Institutional Review Board at the University of Pennsylvania or Erasmus University Medical Centre.


Materials and procedures

CSF was obtained during routine diagnostic lumbar puncture. After the collection of 10 mL CSF for clinical diagnostic purposes, an additional 2 mL was collected in a red top tube, separated into small aliquots of fluid, and immediately frozen at -80°C. As described in detail elsewhere,12 CSF was analyzed in duplicate using a sandwich ELISA for total tau level with a specific kit used at both sites (Innotest hTAU-Antigen kit; Innogenetics, Ghent, Belgium). CSF levels of Aβ42 were also assayed in duplicate with a sandwich ELISA using specific monoclonal antibodies, as described previously.19

Neuropsychological assessment was based on a brief battery of standardized tests administered in the same manner at both institutions. We assessed working memory (forward and backwards digit span)20; letter category naming fluency (number of words beginning with a specified letter F, A, S produced in 1 minute)21; confrontation naming (15-item version of the Boston naming test)22; animal category naming fluency (as many names of animals as possible in 60 s)23; geometric figure copy (copy four geometric designs)23; Verbal Serial List Learning Test (10-word list administered over three trials, delayed free recall, and delayed recognition).24

To establish a neuropathologic diagnosis, board-certified neuropathologists (M.S.F., J.Q.T.) reviewed all cases, and we established diagnoses according to consensus criteria including the recent Workgroup on Frontotemporal Dementia and Pick’s Disease, as described previously.16,25 Immunohistochemistry was performed using standard and previously published protocols with antibodies that detect phosphorylated tau (PHF126, provided by Dr P. Davies); β-amyloid (i.e., 4G8; Senetek, Maryland Heights, MO); α-synuclein (Syn30327); ubiquitin (Chemicon International, Temecula, CA, and Dako Cytomation, Glostrup, Denmark); phosphorylated NF subunits (RMO2428); and α-internexin (Zymed Laboratories, San Francisco, CA). In instances where multiple distinct pathologies were present (n = 1), a primary diagnosis was assigned based on the density and distribution of the observed pathology in relation to the clinical phenotype of the patient.


Statistical analysis

All statistical analyses were performed using SPSS version 12.0 (SPSS, Chicago, IL). Descriptive statistics were used to characterize the entire cohort, as well as each pathologic subgroup. Student t tests were used to investigate differences in demographics and levels of CSF tau, Aβ42, and ratio of tau/Aβ42 between the two pathologic groups. χ2 statistics were used to compare gender ratios and to compare ratios across different CSF tau level and tau/Aβ42 ratio groups. Performance on each measure of cognitive function was reported as a z-score relative to 25 age- and education-matched healthy seniors. Significance was set at -2.32, equivalent to p < 0.01 (two-tailed), unless otherwise noted. The Spearman coefficient was used for correlation. ROC curves were used to evaluate the utility of various measures at distinguishing between groups.


RESULTS

Demographic features and CSF levels of total tau, Aβ42, and tau/Aβ42 are summarized in table 1. Figure 1Figure 1 illustrates CSF total tau, Aβ42, and tau/Aβ42 ratio in FTLD, AD, and control groups. Relative to patients with AD, CSF total tau level [t(47) = 4.05, p < 0.01] and tau/Aβ42 ratio [t(46) = 6.93, p < 0.001] were reduced in patients with FTLD with known pathology. The CSF Aβ42 level was higher in FTLD than AD [t(46) = 4.13; p < 0.001]. Similar findings were evident in an assessment of the subset of autopsy-confirmed patients, that is, lower CSF tau [t(36) = 3.33, p < 0.01] and lower tau/Aβ42 ratio [t(35) = 5.33, p < 0.01] in FTLD than AD, and elevated CSF Aβ42 in FTLD relative to AD [t(35) = 3.48, p < 0.01]. Both CSF tau level [t(34) = 3.52, p < 0.01] and tau/Aβ42 ratio [t(22) = 5.59, p < 0.01] were lower in tau-negative patients than patients with AD. Similarly, CSF tau level [t(26) = 2.94, p < 0.01] and tau/Aβ42 ratio [t(19) = 5.66, p < 0.01] were lower in tau-positive patients than patients with AD. CSF total tau, Aβ42 and tau/Aβ42 ratio were lower in tau-negative than tau-positive patients, but this difference was not significant. Patients with AD had elevated CSF tau level [t(30) = 3.12, p < 0.01] and tau/Aβ42 ratio [t(30) = 3.87, p < 0.01], and lower CSF Aβ42 level[t(30) = 3.79, p < 0.01] compared to the control group, but FTLD and controls did not differ in these measures.



________
Figure 1
CSF total tau, Aβ42, and tau/Aβ42 ratio in frontotemporal lobar degeneration (FTLD), Alzheimer disease (AD), and control groups
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2707002&rendertype=figure&id=F1
________



Figure 2A shows the scatter plot of the raw values of the CSF tau level and tau/Aβ42 ratio in FTLD and AD. Figure 2B illustrates the ROC curves for CSF total tau, Aβ42, and tau/Aβ42 ratio in FTLD and AD based on these raw values. The area under the curve (AUC) for CSF tau was 0.79 (p = 0.001). For tau/Aβ42, the AUC was 0.93 (p < 0.001). Using a cutoff value of 1.06, CSF tau/Aβ42 ratio has a sensitivity of 78.9% and a specificity of 96.6% at distinguishing FTLD from AD. Using a cutoff value of 403.05 pg/mL, total CSF tau has a sensitivity of 68.4% and a specificity of 89.7% at distinguishing FTLD from AD. In evaluating FTLD and controls, the AUC for CSF tau was 0.52 (p = 0.817), and the AUC for tau/Aβ42 ratio was 0.42 (p = 0.422), showing modest ability to distinguish FTLD from controls based on CSF biomarkers.



________
Figure 2
Scatterplot and receiver operating characteristic curves of CSF tau, Aβ42, and tau/Aβ42 ratio in patients with frontotemporal lobar degeneration (FTLD) and Alzheimer disease (AD)
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2707002&rendertype=figure&id=F2
________



An analysis of total CSF tau levels was performed in individual patients using z-scores derived from the control sample obtained using the same CSF analyses. This was necessary to combine CSF and cognitive results in the diagnostic evaluation of FTLD and AD (see below). The range of normal CSF total tau is 81-445 pg/mL (z < -1.96, equivalent to a p value of < 0.05, two-tailed), based on this control group. Using these z-score values, we found that 4 (13.3%) of the 30 patients with FTLD have a significantly reduced CSF total tau level. Of the four cases with significantly reduced CSF tau levels, all were tau-negative patients: three had FTLD-U, and the remaining case had DLDH. All four patients with a low CSF tau level presented with a social/executive impairment. Three FTLD cases with significantly reduced tau were so low that they could not be detected by the kit we used. A significantly reduced level of CSF tau was never seen in individual patients with AD, according to these z-score analyses. However, a much higher proportion of patients with AD (10 of 19, or 52.6%) had an elevated CSF total tau level compared to FTLD (3 of 30, or 10.0%) [χ2(2) = 12.03; p = 0.002]. Of the three patients with FTLD with an elevated tau level, two were tau-positive (one with Pick disease presenting with a social/executive impairment and one with a P301L mutation presenting with mixed social impairment and primary progressive aphasia), and the other had FTLD-U presenting with a social/executive impairment. An analysis of individual patient tau/Aβ42 ratios, using z-scores derived from the control sample, revealed that none of the patients with FTLD have an elevated tau/Aβ42 ratio, while more than half (10 of 19, or 52.6%) of the patients with AD have an elevated tau/Aβ42 ratio [χ2(1) = 19.28, p < 0.001].

Table 2 shows the z-score of cognitive tests in patients with known pathology due to FTLD or AD. Using a z-score criterion of -2.32 (p < 0.01) based on 25 healthy seniors, the mean z-score of confrontation naming is significantly impaired in FTLD but not in AD. In contrast, the z-score of visual constructions is impaired in AD but not FTLD. Animal category naming fluency and delayed free recall memory were significantly impaired in both FTLD and AD.



________
Table 2
Mean z-scores of cognitive function, and relationship between CSF biomarker and neuropsychological measures in patients with FTLD and AD
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2707002&rendertype=table&id=T2
________



There were several correlations between CSF biomarkers and neuropsychological tests in FTLD and AD. As summarized in table 2, the ratio of tau/Aβ42 correlated with confrontation naming and delayed recall memory in FTLD. In AD, tau and the tau/Aβ42 ratio correlated with animal naming fluency output, and the ratio of tau/Aβ42 also correlated with confrontation naming accuracy.

ROC curves for the Boston Naming Test and animal naming fluency in FTLD and AD showed that the AUC for Boston Naming test was 0.62 (p = 0.22) and the AUC for animal fluency was 0.48 (p = 0.85). Thus, performance on these neuropsychological measures alone does not distinguish effectively between FTLD and AD.

Because of the correlations of CSF tau and tau/Aβ42 ratio with neuropsychological tests, and because of the differences between FTLD and AD in autopsy-proven cases, we tried to combine CSF tau and tau/Aβ42 with neuropsychological measures such as animal category naming fluency and confrontation naming in an attempt to improve sensitivity and specificity at distinguishing FTLD from AD. The results showed that only the combination of CSF total tau with confrontation naming improved the ROC marginally (AUC 0.81, p = 0.002) compared to CSF tau alone. The combinations of tau/Aβ42 ratio with confrontation naming (AUC = 0.81, p = 0.002) and with animal category naming fluency (AUC = 0.76, p = 0.009) were statistically robust, but these combined measures showed no improvement compared to the tau/Aβ42 ratio alone.


DISCUSSION

CSF tau is a potentially valuable biomarker for differentiating FTLD from AD. Based on analyses of clinically diagnosed cohorts, reports of CSF total tau levels in patients with FTLD have been highly variable, ranging from normal11 to significantly low12 and significantly high8-10 compared with AD and normal senior controls. In this study, we examined a large group of patients with known pathology confirmed by autopsy or genetic testing to evaluate the usefulness of CSF biomarkers in diagnosing FTLD during life. We found that CSF tau and the ratio of tau/Aβ42 are significantly lower in FTLD than in AD. The ratio of tau/Aβ42 appears to be a particularly sensitive and specific biomarker at discriminating FTLD from AD.

The impression that CSF tau level and CSF tau/AB42 ratio may be valuable in differentiating FTLD from AD is confirmed by individual patient analyses. We found significantly reduced CSF tau level in FTLD, but none of the autopsy-proven patients with AD had a significantly reduced CSF total tau level. Likewise, none of the patients with FTLD had a high tau/Aβ42 ratio, although more than half of the patients with AD had a significantly elevated tau/Aβ42 ratio.

Several factors probably contribute to the inconsistent comparative CSF findings reported in previous articles. First, most of these studies included patients with a clinical diagnosis instead of autopsy-proven diagnosis, and only a few cases with a neuropathologically proven diagnosis have been reported, raising concern about the possibility of misdiagnosis. Clinical-pathologic correlation studies show that between 15% and 33% of patients with a clinical diagnosis of FTLD in fact have AD at autopsy.26-32 Nevertheless, the results of CSF total tau level in autopsy-proven patients with FTLD have been inconsistent. One study reported significantly increased CSF total tau in autopsy-proven patients with FTLD compared with normal controls8 while another study showed no significant difference between FTLD and controls.33 However, these were very small samples. Another contributing factor may be related to the pathologic diversity of FTLD, with varying numbers of patients with PiD, CBD, or FTLD-U included in these studies. Larger studies of patients with diagnoses such as these are needed to clarify the role of these specific pathologies on CSF biomarker values.

The reason that some patients with FTLD have low CSF tau remains unclear. One possible explanation is that the soluble tau level in the brain is depleted, leaving no source for CSF tau. This may be seen in the four tau-negative patients with FTLD (one case with DLDH and three cases with FTLD-U) with low CSF tau levels in this study. Another possibility is that tau is sequestered in the brain, possibly in the form of filamentous inclusions such as Pick bodies or in balloon cells.12 This finding may also be related to the topographic distribution of CSF protein concentrations, especially during early clinical stages, where the degenerating process in FTLD may be confined to small cortical areas that are remote from ventricular and lumber CSF spaces.10 It is less likely that low CSF tau level is related to longer disease duration at the time of assessment since the patients with FTLD in fact had marginally shorter disease duration than the patients with AD. Regardless of the basis for low CSF tau in FTLD, ROC curves showed that both CSF tau and tau/Aβ42 ratio were able to discriminate between FTLD and AD.

The tau/Aβ42 ratio has better sensitivity and specificity compared to CSF tau alone in distinguishing patients with FTLD from those with AD. This may be related in part to the pathogenic roles of tau and Aβ42 in FTLD and AD.34 It has been suggested that CSF total tau reflects the intensity of neuronal degeneration,35,36 while CSF Aβ42 reflects the deposition of Aβ into plaques,37,38 and phosphorylated tau reflects the hyperphosphorylation of tau with subsequent formation of tangles.39 The latter two are major neuropathologic characteristics of AD.40 In FTLD, however, amyloid plaque deposition, even if present in some cases, is not a common feature.41,10 This may be the main reason that patients with FTLD usually have higher Aβ42 level compared to AD, and why the ratio of tau/Aβ42 is superior at discriminating between FTLD and AD. Indeed, the tau/Aβ42 ratio discriminated between FTLD and AD in our cohort with a sensitivity of 79% and a specificity of 97%. According to a consensus report, a useful diagnostic biomarker should have sensitivity approaching or exceeding 80% for detecting AD and a specificity greater than 80% for distinguishing other dementias.42 The ratio of tau/Aβ42 thus appears to fulfill these requirements, because both high sensitivity and specificity were obtained at a cutoff value of 1.06.

A reliable and reproducible diagnostic biomarker also should be stable over time. Previous studies have indicated that CSF tau and Aβ42 levels remain stable in patients with AD when CSF samples are compared over an average interval of 6 to 18 months.43-45 In a cross-sectional correlation analysis with disease duration in the present study, we found no correlation between disease duration and CSF tau levels and the tau/Aβ42. However, the stability of these biomarkers in the course of FTLD remains to be documented using a within-patient longitudinal design.

The results of neuropsychological measures in this study indicate that distinct cognitive profiles also can differentiate between patients with known pathology due to FTLD and AD. Patients with FTLD showed greater deficits on tests of confrontation naming and category naming fluency, but relatively better performance on memory and visuospatial tasks than patients with AD. These findings are consistent with clinical observations that FTLD predominantly affects executive skills and language function, while memory and visuospatial functioning are more likely to be impaired in AD. Other comparative studies of patients with autopsy-confirmed FTLD and AD also report these findings.17,18,32 This may reflect the distinct regional brain pathology that underlies these conditions. MRI studies in FTLD associate orbitofrontal, dorsolateral, and anterior temporal cortical atrophy with impairment of personality, behavior, and social conduct, and with poor performance on cognitive tasks that involve executive and language functions.46,47 Other work relates cognitive measures such as these directly to the anatomic distribution of histopathologic disease in FTLD using correlation analyses.18 In contrast, patients with AD have extensive medial temporal and parietal lobe involvement that is consistent with the prominent episodic memory dysfunction and visuospatial impairments that occur in the disorder.48-52

CSF protein levels and neuropsychological measures are two promising biomarkers for distinguishing FTLD from AD because of their quantitative nature. The significant correlation of tau/Aβ42 with recall in FTLD indicates that poor memory performance is associated with an abnormally elevated tau/Aβ42 ratio. Likewise, the significant correlation of tau/Aβ42 ratio with confrontation naming in AD indicates that better naming performance is associated with an abnormally elevated tau/Aβ42 ratio. Since an elevated tau/Aβ42 ratio occurs in more than half of patients with AD but rarely in FTLD, poorer performance on recall memory and better performance on confrontation naming tasks may be a clinical indication that a patient is less likely to have FTLD.

CSF tau and tau/Aβ42 ratio have been linked directly to histopathologic disease, just as specific neuropsychological measures have been related to known pathology in FTLD.17,18,32 We sought to determine whether the combination of these two measures would be more informative than either one alone at distinguishing FTLD from AD. The combination of neuropsychological measures with CSF biomarkers in fact did not improve the ability to distinguish FTLD from AD. Although a ROC curve showed that the combination of CSF tau level with confrontation naming had a relatively higher specificity than CSF tau alone, the overall sensitivity and specificity of this combination is still less than the tau/Aβ42 ratio alone. None of the neuropsychological measures, when combined with the tau/Aβ42 ratio, improved the AUC associated with the ratio measure alone. Thus the ratio of tau/Aβ42 appears to be the most effective biomarker at discriminating FTLD from AD.

Several caveats should be kept in mind when interpreting our results. First, the CSF tau/Aβ42 ratio is a valuable biomarker at discriminating between neurodegenerative conditions with clinically detectable abnormalities such as FTLD and AD. However, no statistical difference was found between FTLD and the control group in the CSF tau level and tau/Aβ42 ratio. These biomarkers thus are useful only in the setting of an abnormal neurologic examination, and are not helpful at differentiating patients with FTLD from healthy controls. Second, both academic medical centers participating in this study used the same kit to analyze CSF and administered the same neuropsychological tests in a standard manner. However, minor methodologic differences in data analysis in the two centers can not be completely excluded. Third, most CSF samples were typically collected soon after the patients were first seen in our clinics, but there was often a delay between symptom onset and initial evaluation. Although we failed to find a correlation between CSF biomarkers and disease duration in a cross-sectional analysis, it is important to be cautious about generalizing our observations to the earliest stage of a neurodegenerative disease. In addition, not all types of pathology were included in the FTLD group, such as progressive supranuclear palsy and argyrophilic grain disease, and this may influence the CSF tau level and cognitive test results of the FTLD group. Finally, only a subset of the patients performed the cognitive tests, and the relatively small number of cases with known pathology included in this study limited the interpretability of these results. With these caveats in mind, our findings suggest that FTLD and AD have relatively distinct CSF biomarker profiles. Relative to patients with AD, patients with FTLD are characterized by lower levels of CSF tau and tau/Aβ42 ratio. The ratio of tau/Aβ42 appears to be the most sensitive and specific biomarker at discriminating FTLD from AD.


Acknowledgments

Supported in part by NIH (AG17586, AG15116, and NS44266) and the Dana Foundation.


GLOSSARY
AD -- Alzheimer disease
AUC -- area under the curve
CBD -- corticobasal degeneration
FTLD -- frontotemporal lobar degeneration
FTLD-U -- FTLD with ubiquitin-positive but tau-negative inclusions
MAPT -- microtubule-associated protein tau
MND -- motor neuron disease
PiD -- dementia with Pick bodies
ROC -- receiver operating characteristic
TPSD -- tangle predominant senile dementia


Footnotes
Disclosure: The authors report no conflicts of interest.


REFERENCES

1. Hutton M, Lendon CL, Rizzu P, et al. Association of missense and 5′-splice-site mutations in tau with the inherited dementia FTDP-17. Nature. 1998;393:702–705. [PubMed]

2. Poorkaj P, Bird TD, Wijsman E, et al. Tau is a candidate gene for chromosome 17 frontotemporal dementia. Ann Neurol. 1998;43:815–825. [PubMed]

3. Spillantini MG, Murrell JR, Goedert M, Farlow MR, Klug A, Ghetti B. Mutation in the tau gene in familial multiple system tauopathy with presenile dementia. Proc Natl Acad Sci USA. 1998;95:7737–7741. [PubMed]

4. Rademakers R, Cruts M, van Broeckhoven C. The role of tau (MAPT) in frontotemporal dementia and related tauopathies. Hum Mutat. 2004;24:277–295. [PubMed]

5. Ghetti B, Hutton ML, Wszolek ZK. Frontotemporal dementia and parkinsonism linked to chromosome 17 associated with Tau gene mutations (FTDP-17T). In: Dickson D, editor. Neurodegeneration: the molecular pathology of dementia and movement disorders. ISN Neuropath Press; Basel, Switzerland: 2003. pp. 86–102.

6. Baker M, Mackenzie IR, Pickering-Brown SM, et al. Mutations in progranulin cause tau-negative frontotemporal dementia linked to chromosome 17. Nature. 2006;442:916–919. [PubMed]

7. Mackenzie IR, Baker M, Pickering-Brown SM, et al. The neuropathology of frontotemporal lobar degeneration caused by mutations in the progranulin gene. Brain. 2006;129:3081–3090. [PubMed]

8. Arai H, Morikawa Y, Higuchi M, et al. Cerebrospinal fluid tau levels in neurodegenerative diseases with distinct tau-related pathology. Biochem Biophys Res Commun. 1997;236:262–264. [PubMed]

9. Green AJE, Harvey RJ, Thompson EJ, Rossor MN. Increased tau in the cerebrospinal fluid of patients with frontotemporal dementia and Alzheimer’s disease. Neurosci Lett. 1999;259:133–135. [PubMed]

10. Riemenschneider M, Wagenpfeil S, Diehl J, et al. Tau and Aβ42 protein in CSF of patients with frontotemporal degeneration. Neurology. 2002;58:1622–1628. [PubMed]

11. Mecocci P, Cherubini A, Bregnocchi M, et al. Tau protein in cerebrospinal fluid: a new diagnostic and prognostic marker in Alzheimer disease? Alzheimer Dis Assoc Disord. 1998;12:211–214. [PubMed]

12. Grossman M, Farmer J, Leight S, et al. Cerebrospinal fluid profile in frontotemporal dementia and Alzheimer’s disease. Ann Neurol. 2005;57:721–729. [PubMed]

13. Kramer JH, Miller BR. Alzheimer’s disease and its focal variants. Semin Neurol. 2000;20:447–454. [PubMed]

14. Johnson JK, Head E, Kim R, Starr A, Cotman CW. Clinical and pathological evidence for a frontal variant of Alzheimer disease. Arch Neurol. 1999;56:1233–1239. [PubMed]

15. Graham A, Davies R, Xuereb J, et al. Pathological proven frontotemporal dementia presenting with severe amnesia. Brain. 2005;128:597–605. [PubMed]

16. Neary D, Snowden JS, Gustafson L, et al. Frontotemporal lobar degeneration: a consensus on clinical diagnostic criteria. Neurology. 1998;51:1546–1554. [PubMed]

17. Rascovsky K, Salmon DP, Ho GJ, et al. Cognitive profiles differ in autopsy-confirmed frontotemporal dementia and AD. Neurology. 2002;58:1801–1808. [PubMed]

18. Grossman M, Libon DJ, Forman MS, et al. Distinct antemortem profiles in pathologically-defined patients with frontotemporal dementia. Arch Neurol. 2008 in press.

19. Skowronsky DM, Lee VMY, Pratico D. Amyloid pre-cursor protein and amyloid β peptide in human platelets: role of cyclooxygenase and protein kinase C. J Biol Chem. 2001;276:17036–17043. [PubMed]

20. Wechsler D. Wechsler Memory Scale-Revised. The Psychological Corporation; San Antonio: 1987.

21. Spreen O, Struss E. A Compendium of Neuropsychological Tests. Oxford University Press; New York: 1990.

22. Kaplan E, Goodglass H, Weintraub S. The Boston Naming Test. Lea and Febiger; Philadelphia: 1983.

23. Morris JC, Heyman A, Mohs RC. The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD), part I: clinical and neuropsychological assessment of Alzheimer’s disease. Neurology. 1989;39:1159–1165. [PubMed]

24. Welsh KA, Butters N, Hughes J, Mohs R, Heyman A. Detection of abnormal memory decline in mild cases of Alzheimer’s disease using CERAD neuropsychological measures. Arch Neurol. 1991;48:278–281. [PubMed]

25. McKhann GM, Albert MS, Grossman M, Miller B, Dickson D, Trojanowski JQ. Clinical and pathological diagnosis of frontotemporal dementia: report of the Work Group on Frontotemporal Dementia and Pick’s Disease. Arch Neurol. 2001;58:1803–1809. [PubMed]

26. Hodges JR, Davies RR, Xuereb JH, et al. Clinicopathological correlates in frontotemporal dementia. Ann Neurol. 2004;56:399–406. [PubMed]

27. Kertesz A, McMonagle P, Blair M, Davidson W, Munoz DG. The evolution and pathology of frontotemporal dementia. Brain. 2005;128:1996–2005. [PubMed]

28. Davies RR, Hodges JR, Kril JJ, Patterson K, Halliday GM, Xuereb JH. The pathological basis of semantic dementia. Brain. 2005;128:1984–1995. [PubMed]

29. Knopman DS, Boeve BF, Parisi JE, et al. Antemortem diagnosis of frontotemporal lobar degeneration. Ann Neurol. 2004;57:480–488. [PubMed]

30. Knibb JA, Xuereb J, Patterson K, Hodges JR. Clinical and pathological characterization of progressive aphasia. Ann Neurol. 2006;56:156–165. [PubMed]

31. Lipton AM, Cullum CM, Satumtira S, et al. Contribution of asymmetric synapse loss to lateralizing clinical deficits in frontotemporal dementias. Arch Neurol. 2001;58:1233–1239. [PubMed]

32. Forman MS, Farmer J, Johnson JK, et al. Frontotemporal dementia: clinicopathological correlations. Ann Neurol. 2006;59:952–962. [PubMed]

33. Clark CM, Xie S, Chittams J, et al. Cerebrospinal fluid tau and β-amyloid. How well do these biomarkers reflect autopsy-confirmed dementia diagnoses? Arch Neurol. 2003;60:1696–1702. [PubMed]

34. Blennow K, Hampe H. Cerebrospinal fluid markers for incipient Alzheimer’s disease. Lancet Neurol. 2003;2:605–613. [PubMed]

35. Blennow K. Cerebrospinal fluid protein biomarkers for Alzheimer’s disease. NeuroRx. 2004;1:213–225. [PubMed]

36. Hesse C, Rosengren L, Vanmechelen E, et al. Cerebrospinal fluid markers for Alzheimer’s disease evaluated after acute ischemic stroke. J Alzheimer Dis. 2000;2:199–206.
37. Fagan AM, Mintun MA, Mach RH, et al. Inverse relation between in vivo amyloid imaging load and cerebrospinal fluid Abeta42 in humans. Ann Neurol. 2006;59:512–519. [PubMed]

38. Strozyk D, Blennow K, White LR, Launer LJ. CSF Aβ42 levels correlate with amyloid-neuropathology in a population-based autopsy study. Neurology. 2003;60:652–656. [PubMed]

39. Buerger K, Ewers M, Pirttila T, et al. CSF phosphorylated tau protein correlates with neocortical neurofibrillary pathology in Alzheimer’s disease. Brain. 2006;129:3035–3041. [PubMed]

40. Glenner GG, Wong CW. Alzheimer’s disease: initial report of the purification and characterization of a novel cerebrovascular amyloid protein. Biochem Biophys Res Commun. 1984;120:885–890. [PubMed]

41. Arnold SE, Han LY, Clark CM, Grossman M, Trojanowski JQ. Quantitative neurohistological features of frontotemporal degeneration. J Struct Biol. 2000;130:271–279. [PubMed]

42. Consensus report of the Working Group on Molecular and Biochemical Markers of Alzheimer’s Disease. Neurobiol Aging. 1998;19:109–116. [PubMed]

43. Kanai M, Matsubara E, Isoe K, et al. Longitudinal study of cerebrospinal fluid levels of tau, Aβ1-40, and Aβ1-42(43) in Alzheimer’s disease: a study in Japan. Ann Neurol. 1998;44:17–26. [PubMed]

44. Andreasen N, Hess C, Davidsson P, et al. Cerebrospinal fluid β-amyloid (1-42) in Alzheimer’s disease: differences between early- and late-onset Alzheimer’s disease and stability during the course of disease. Arch Neurol. 1999;56:673–680. [PubMed]

45. Blennow K, Zetterberg H, Minthon L, et al. Longitudinal stability of CSF biomarkers in Alzheimer’s disease. Neurosci Lett. 2007;419:18–22. [PubMed]

46. Perry RJ, Hodges JR. Differentiating frontal and temporal variant frontotemporal dementia from Alzheimer disease. Neurology. 2000;54:2277–2284. [PubMed]

47. Rosen HJ, Gorno-Tempini ML, Goldman WP, et al. Patterns of brain atrophy in frontotemporal dementia and semantic dementia. Neurology. 2002;58:198–208. [PubMed]

48. Duarte A, Hayasaka S, Dua A, et al. Volumetric correlates of memory and executive function in normal elderly, mild cognitive impairment and Alzheimer’s disease. Neuroscience Letters. 2006;406:60–65. [PubMed]

49. Baron JC, Chetelat G, Desgranges B, et al. In vivo mapping of gray matter loss with voxel-based morphometry in mild Alzheimer’s disease. Neuroimage. 2001;14:298–309. [PubMed]

50. Karas GB, Burton EJ, Rombouts SA, et al. A comprehensive study of gray matter loss in patients with Alzheimer’s disease using optimized voxel-based morphometry. Neuroimage. 2003;18:895–907. [PubMed]

51. Du AT, Schuff N, Kramer JH, et al. Higher atrophy rate of entorhinal cortex than hippocampus in AD. Neurology. 2004;62:422–427. [PubMed]

52. Petersen RC, Jack CR, Xu YC, et al. Memory and MRI-based hippocampal volumes in aging and AD. Neurology. 2000;54:581–587. [PubMed]



___
Department of Neurology (H.B., L.M., P.M., C.M.C., M.G.), Department of Pathology and Laboratory Medicine (S.L., E.W., M.F., J.T., V.M.-Y.L.), Center for Neurodegenerative Disease Research (S.L., E.W., M.F., J.T., V.M.-Y.L.), Institute on Aging (S.L., J.T., V.M.-Y.L.), and Alzheimer’s Disease Center (C.M.C.), University of Pennsylvania School of Medicine, Philadelphia; Department of Neurology, Erasmus University Medical Centre (J.C.V.S., I.d.k., S.R.), Rotterdam, Netherlands; and Department of Neurology, Jinan Central Hospital, Shandong University School of Medicine (H.B.), Jinan, Shandong Province, People’s Republic of China
Address correspondence and reprint requests to Dr. Murray Grossman, Department of Neurology, 2 Gibson, Hospital of the University of Pennsylvania, 3400 Spruce Street, Philadelphia, PA 19104-4283, Email: mgrossma@mail.med.upenn.edu