Abstract
Background
We evaluated the relationship between florbetapir-F18 positron emission tomography (FBP PET) and cerebrospinal fluid (CSF) biomarkers.
Methods
Alzheimer’s Disease Neuroimaging Initiative (ADNI)-GO/2 healthy control (HC), mild cognitive impairment (MCI), and Alzheimer’s disease (AD) dementia subjects with clinical measures and CSF collected ±90 days of FBP PET data were analyzed using correlation and logistic regression.
Results
In HC and MCI subjects, FBP PET anterior and posterior cingulate and composite standard uptake value ratios correlated with CSF amyloid beta (Aβ1-42) and tau/Aβ1-42 ratios. Using logistic regression, Aβ1-42, total tau (t-tau), phosphorylated tau181P (p-tau), and FBP PET composite each differentiated HC versus AD. Aβ1-42 and t-tau distinguished MCI versus AD, without additional contribution by FBP PET. Total tau and p-tau added discriminative power to FBP PET when classifying HC versus AD.
Conclusion
Based on cross-sectional diagnostic groups, both amyloid and tau measures distinguish healthy from demented subjects. Longitudinal analyses are needed.
Keywords: Alzheimer’s disease, florbetapir positron emission tomography, cerebrospinal fluid, mild cognitive impairment, Alzheimer’s Disease Neuroimaging Initiative, biomarkers
1. Background
Hallmark neuropathological lesions of Alzheimer’s disease (AD) at autopsy are amyloid beta (Aβ) protein deposition in plaques and hyperphosphorylated tau deposition in neurofibrillary tangles [1]. However, data from the National Institute on Aging (NIA) Alzheimer’s Disease Centers collected from 2005 to 2010 found ranges for sensitivity of 70.9% to 87.3% and specificity of 44.3% to 70.8% when clinical diagnoses of possible and probable AD dementia are compared with postmortem histopathology diagnosis [2]. Florbetapir-F18 positron emission tomography (FBP PET) for estimating beta-amyloid neuritic plaque density was Food and Drug Administration (FDA)-approved in April 2012 and has high sensitivity (96%; 95% CI [confidence interval] 80%–100%) and specificity (100%; 95% CI 78%–100%) versus autopsy within 1 year [3]. Another positron emission tomography (PET) radiotracer used to quantify amyloid deposits in the brain in research settings is Pittsburgh compound B (PiB) [4, 5]. Cerebrospinal fluid (CSF) levels of Aβ1-42, total tau (t-tau), and phosphorylated tau181P (p-tau) [6] are additional research tools with ongoing efforts to standardize across laboratories and patients [7, 8].
A model of the temporal order in which clinically measurable AD biomarkers become abnormal throughout the progression of AD has been proposed by Jack and colleagues [9]. According to this model, abnormal CSF Aβ1-42 and amyloid PET findings are detected earliest, followed by CSF tau and other biomarker types. Deposition of Aβ into plaques appears very early in the disease process during the asymptomatic stages prior to AD dementia. In contrast, elevated tau levels are downstream biomarkers that become strikingly more abnormal closer to the development of clinical symptoms [9]. Evidence continues to accumulate in support of this model [10–12]. Fagan and colleagues reported a similar CSF biomarker phenotype in patients with very mild AD symptoms (Clinical Dementia Rating [CDR]=0.5) versus patients with more advanced AD (CDR>1) [13].
There is no consensus for antemortem staging of AD clinical phases using biomarker thresholds and where the progression of neuropathological changes is hypothesized to be on a continuum beginning with a long asymptomatic period and culminating in dementia [14, 15]. Further, symptom severity is influenced by multiple factors, such as age [16], premorbid functioning [17], education [18], cognitive reserve [14], apolipoprotein E epsilon 4 (APOE4) allele carrier status [19], and certain concurrent medical conditions [20]. Thus, there may be a discrepancy between the presence and degree of AD neuropathology with the expression of AD symptoms on an individual basis. These challenges underscore the need for additional tools, such as AD clinical biomarkers, to aid the accurate diagnosis and staging of AD across the continuum of clinical progression [21].
The CSF Aβ1-42 and tau analytes and amyloid PET neuroimaging as adjunctive biomarkers for diagnosis of AD are not commonly used in clinical practice but have the potential to significantly impact accuracy of a clinical diagnosis. There is a small amount of emerging literature about their relationship to each other across the spectrum of disease progression. Studies of the amyloid brain deposits assessed with PiB PET and CSF levels of Aβ1-42 found an inverse relationship between them, no relationship between PiB and CSF t-tau or p-tau, and discordance with clinical diagnosis where some healthy controls had evidence of amyloid positive status by both PiB and CSF Aβ1-42 [4, 5]. Binary classification using PiB PET and CSF-Aβ1-42 overlapped in 96.4% [4].
We explored cross-sectional relationships between FBP PET and CSF biomarkers among groups of healthy control (HC), mild cognitive impairment (MCI), and AD dementia subjects enrolled in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) using approaches not previously reported. We measured correlations between regional and composite FBP PET values and CSF Aβ1-42, t-tau, and p-tau, and their ratios in diagnostic groups. We used logistic regression to compare composite FBP PET values with CSF Aβ1-42, t-tau, and p-tau in distinguishing between diagnostic groups including evaluating for additive contributions by the other biomarker type.
2. Methods
2.1. Subjects and study design
Data used in the preparation of this article were obtained from the ADNI database (adni.loni.usc.edu). The ADNI was launched in 2003 by the NIA, the National Institute of Biomedical Imaging and Bioengineering (NIBIB), the FDA, private pharmaceutical companies and non-profit organizations, as a $60 million, 5-year public-private partnership. The primary goal of ADNI has been to test whether serial magnetic resonance imaging (MRI), PET, other biological markers, and clinical and neuropsychological assessment can be combined to measure the progression of MCI and early AD. Determination of sensitive and specific markers of very early AD progression is intended to aid researchers and clinicians to develop new treatments and monitor their effectiveness, as well as lessen the time and cost of clinical trials.
The Principal Investigator of this initiative is Michael W. Weiner, MD, VA Medical Center and University of California – San Francisco. ADNI is the result of efforts of many co-investigators from a broad range of academic institutions and private corporations, and subjects have been recruited from over 50 sites across the United States and Canada. The initial goal of ADNI was to recruit 800 subjects, but ADNI has been followed by ADNI-GO and ADNI-2. To date these 3 protocols have recruited over 1500 adults, ages 55 to 90, to participate in the research, consisting of cognitively normal older individuals, people with early or late MCI, and people with early AD. The follow-up duration of each group is specified in the protocols for ADNI-1, ADNI-2, and ADNI-GO. Subjects originally recruited for ADNI-1 and ADNI-GO had the option to be followed in ADNI-2. For up-to-date information, see www.adni-info.org.
Data were downloaded in August 2012 from ADNI-GO/2 which included FBP PET scans. Participants were recruited from outpatient memory clinics. Clinical diagnoses were assigned to participants by the site investigators and reassessed at each visit. Normal age-matched control subjects showed no signs of depression, MCI, or dementia (www.adni-info.org). Participants with MCI were required to present education-adjusted ranges on the Logical Memory II subscale from the Wechsler Memory Scale-Revised: ≥ 16 years of education – 9 to 11 for early MCI, ≤ 8 for late MCI; 8 to 15 years of education – 5 to 9 for early MCI, ≤4 for late MCI; 0 to 7 years of education – 3 to 6 for early MCI, ≤ 2 for late MCI. Additionally, participants with MCI had Mini-Mental State Examination (MMSE) scores between 24 and 30 (inclusive), a CDR of 0.5 with a Memory Box score ≥ 0.5, and preserved activities of daily living. Participants with AD dementia met the National Institute of Neurological and Communicative Disorders and Stroke - Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) criteria for probable AD. At subsequent visits, diagnoses were categorized as HC, MCI, or AD. For this cross-sectional analysis, we selected all HC, MCI, and AD dementia subjects who had clinical measures, diagnoses, and CSF analyte levels within ±90 days of their FBP PET scans.
2.2. Clinical measures
The following clinical measures were included to describe the sample: Estimated Verbal Intelligence Quotient (EVIQ), Functional Activities Questionnaire (FAQ), Geriatric Depression Scale (GDS), Neuropsychiatric Inventory-Questionnaire (NPI-Q), 11- and 13-item versions of the cognitive subscale of the Alzheimer’s Disease Assessment Scale (ADAS-Cog11; ADAS-Cog13), and MMSE.
2.3. Biomarker variables
2.3.1. Florbetapir-F18 positron emission tomography
FBP PET data for all subjects were analyzed using a semi-automatic method, which includes spatial normalization to a standard template in the Talairach space [3]. Standard uptake value ratios (SUVRs) using whole cerebellum as the reference region were calculated for 6 FBP PET regions of interest (ROI): posterior cingulate, precuneus, parietal, temporal, anterior cingulate, frontal; and the composite, which is their mean SUVR. The 6 target ROIs were defined in a previous study,[22] in which PET uptake was increased in AD subjects compared with control subjects. Raw FBP PET data were initially pre-processed at the Laboratory of Neuroimaging at the University of California, Berkeley (http://resource.loni.ucla.edu/research/data-interpretation/).
2.3.2. Cerebrospinal fluid measures
Samples were analyzed using the Luminex® xMAP® platform (Austin, TX) and Innogenetics/Fujirebio AlzBio3 immunoassay kits (Gent, Belgium) by the ADNI Core Laboratory at the University of Pennsylvania Medical Center. The following variables were determined: Aβ1-42, t-tau, p-tau, t-tau/Aβ1-42 ratio, and p-tau/Aβ1-42 ratio.
2.4. Genotyping
A blood sample for genomic deoxyribonucleic acid extraction was obtained at enrollment for all study participants. The APOE4 genotyping on these samples was performed by Illumina® (San Diego, CA).
2.5. Statistical analyses
Pearson correlation coefficients were calculated among 5 CSF and 7 FBP PET variables by diagnostic group. Demographic and other clinical characteristics were compared among 3 diagnostic groups with Chi-square/Fisher’s exact test for categorical characteristics and analysis of variance for continuous variables. A significance cutoff of P≤.0014 based on Bonferroni correction was applied (i.e., taking into account 35 correlations for each diagnostic group).
Logistic regression modeling assessed relationships between clinical diagnosis with CSF variables (not ratios) and the FBP PET composite SUVR. The likelihood ratio test was used to examine whether adding CSF biomarkers to the model, which regresses clinical diagnosis on FBP PET composite SUVR, significantly improved model fit, and vice versa. Analyses were adjusted for the following subject demographics: APOE4 carrier status (binary); age at FBP PET scan; gender; and EVIQ. Data are expressed with bolded P-value notation for analyses meeting the statistical significance threshold after Holm-Bonferroni correction [23] for multiple comparisons (i.e., taking into account 30 analyses). All regression analyses were done separately for 3 pairs of diagnoses: HC versus MCI, MCI versus AD, and HC versus AD. For all analyses, statistical significance was defined as P≤.05, except where corrections were applied.
3. Results
3.1. Subject characteristics
A total of 577 subjects underwent FBP PET scans and had clinical diagnoses available within ±90 days of the scan. Of these, 344 subjects had all data points available for FBP PET, CSF, clinical diagnosis, age, and EVIQ, as well as sex and APOE4 status, and were the basis of this analysis. These 344 subjects consisted of 97 HC, 226 MCI, and 21 AD dementia subjects; mean ages were 74.5 (±5.6) years in HC, 71.4 (±7.5) years in MCI, and 74.0 (±10.0) years in AD dementia subjects (Table 1). Neuropsychiatric assessment scale scores differed significantly (P ≤.05) among groups, with AD dementia subjects most severely affected (Table 1).
Table 1.
Subject demographics and neuropsychiatric assessment.
| HC (n=97) | MCI (n=226) | AD Dementia (n=21) | P-value* | |
|---|---|---|---|---|
| Mean age, years (SD) | 74.5 (5.6) | 71.4 (7.5) | 74.0 (10.0) | .002 |
| Male sex, n (%) | 52 (53.6) | 126 (55.8) | 13 (61.9) | .781 |
| Race, n (%) | .968 | |||
| American Indian or Alaskan Native | 0 | 1 (0.4) | 0 | |
| Asian | 1 (1.0) | 3 (1.3) | 0 | |
| Native Hawaiian or other Pacific Islander | 0 | 2 (0.9) | 0 | |
| Black or African American | 3 (3.1) | 6 (2.7) | 0 | |
| White | 92 (94.8) | 207 (91.6) | 21 (100.0) | |
| Multiracial | 1 (1.0) | 5 (2.2) | 0 | |
| Unknown | 0 | 2 (0.9) | 0 | |
| APOE4 allele carrier, n (%) | <.001 | |||
| No | 76 (78.4) | 128 (56.6) | 7 (33.3) | |
| Yes | 21 (21.6) | 98 (43.4) | 14 (66.7) | |
| Mean education, years (SD) | 16.4 (2.6) | 16.1 (2.6) | 15.8 (2.8) | .382 |
| AmNART error rate, mean (SD) | 10.2 (8.4) | 11.8 (8.4) | 16.3 (10.6)** | .011 |
| FAQ, mean (SD) | 0.2 (0.7) | 2.4 (3.7)††† | 12.9 (7.0)†††‡‡‡ | <.001 |
| EVIQ, mean (SD) | 118.8 (8.0) | 117.2 (8.0) | 113.1 (10.4)††‡ | .012 |
| GDS, mean (SD, n) | 0.7 (1.1, 94) | 1.8 (1.5, 205)††† | 2.0 (1.2, 18)††† | <.001 |
| NPI, mean (SD, n) | 0.4 (1.1, 95) | 2.0 (2.9, 226)††† | 2.7 (3.0, 21)††† | <.001 |
| ADAS-Cog11, mean (SD) | 6.2 (3.1) | 8.9 (4.3)††† | 19.6 (6.2)†††‡‡‡ | <.001 |
| ADAS-Cog13, mean (SD) | 9.7 (4.5) | 14.2 (6.6)††† | 30.3 (8.0)†††‡‡‡ | <.001 |
| MMSE, mean (SD, n) | 29.0 (1.2, 94) | 28.2 (1.7, 207)††† | 22.8 (1.7, 18)†††‡‡‡ | <.001 |
Abbreviations: AD = Alzheimer’s disease; ADAS-Cog11 = Alzheimer’s Disease Assessment Scale, 11-item cognitive subscale; ADAS-Cog13 = Alzheimer’s Disease Assessment Scale, 13-item cognitive subscale; AmNART = American National Adult Reading Test; APOE4 = apolipoprotein E epsilon 4; EVIQ = Estimated Verbal Intelligence Quotient; FAQ = Functional Activities Questionnaire; GDS = Geriatric Depression Scale; HC = healthy controls; MCI = mild cognitive impairment; MMSE = Mini Mental State Examination; n = number of subjects; SD = standard deviation.
P-values from analysis of variance model for continuous variables; from Chi-Square/Fisher’s exact test for categorical variables.
P-value ≤.05 versus HC
P-value<.01 versus HC
P-value<.001 versus HC (P-values versus HC are only indicated in the MCI and AD dementia columns to avoid repetition)
P-value ≤.05 versus MCI
P-value<.001 versus MCI
3.2. Correlation analyses of biomarker variables by diagnostic group
Pearson’s correlation coefficients were assessed between FBP PET SUVR and CSF biomarkers. The highest statistically significant (P≤.05, Bonferroni corrected) correlations were between FBP PET anterior cingulate, posterior cingulate, and composite SUVRs with CSF Aβ1-42, t-tau/Aβ1-42 ratio, and p-tau/Aβ1-42 ratio for HC and MCI groups (Table 2).
Table 2.
Pearson correlation coefficients between FBP PET SUVR and CSF biomarker levels by diagnostic group.
| CSF Biomarkers | Posterior Cingulate | Precuneus | Parietal | Temporal | Anterior Cingulate | Frontal | Composite |
|---|---|---|---|---|---|---|---|
| HC group (n=97) | |||||||
| Aβ1-42 | −0.661* | −0.374* | −0.364* | −0.325* | −0.629* | 0.338* | −0.681* |
| t-tau | 0.346* | 0.054 | 0.073 | 0.042 | 0.388* | 0.065 | 0.392* |
| p-tau | 0.219 | 0.068 | 0.083 | 0.059 | 0.288 | 0.096 | 0.286 |
| t-tau/Aβ1-42 ratio | 0.600* | 0.181 | 0.185 | 0.146 | 0.603* | 0.162 | 0.643* |
| p-tau/Aβ1-42 ratio | 0.562* | 0.260 | 0.266 | 0.228 | 0.613* | 0.253 | 0.635* |
|
| |||||||
| MCI group (n=226) | |||||||
| Aβ1-42 | −0.651* | −0.326* | −0.286* | −0.267* | −0.662* | −0.287* | −0.697* |
| t-tau | 0.557* | 0.190 | 0.139 | 0.154 | 0.560* | 0.178 | 0.573* |
| p-tau | 0.559* | 0.268* | 0.200 | 0.214* | 0.533* | 0.245* | 0.558* |
| t-tau/Aβ1-42 ratio | 0.624* | 0.222* | 0.171 | 0.173 | 0.620* | 0.198 | 0.644* |
| p-tau/Aβ1-42 ratio | 0.661* | 0.301* | 0.241* | 0.241* | 0.638* | 0.267* | 0.678* |
|
| |||||||
| AD Dementia group (n=21) | |||||||
| Aβ1-42 | −0.375 | −0.215 | −0.252 | −0.208 | −0.580 | −0.193 | −0.563 |
| t-tau | −0.058 | 0.404 | 0.404 | 0.428 | 0.180 | 0.417 | 0.082 |
| p-tau | 0.173 | 0.391 | 0.433 | 0.437 | 0.256 | 0.438 | 0.235 |
| t-tau/Aβ1-42 ratio | 0.002 | 0.371 | 0.400 | 0.397 | 0.322 | 0.383 | 0.224 |
| p-tau/Aβ1-42 ratio | 0.170 | 0.268 | 0.331 | 0.309 | 0.317 | 0.306 | 0.297 |
Abbreviations: Aβ1-42 = beta-amyloid protein; AD = Alzheimer’s disease; CSF = cerebrospinal fluid; FBP PET = florbetapir-F18 positron emission tomography; HC = healthy controls; MCI = mild cognitive impairment; n = number of subjects; SUVR = standard uptake value ratio; p-tau = phosphorylated tau181P; t-tau = total tau.
P-value ≤.0014 based on Bonferroni correction
Although significant correlations between CSF tau measures and FBP PET variables were seen, the values of the correlation coefficients were relatively lower unless CSF tau was in a ratio with Aβ1-42. Correlations between both t-tau and p-tau and several FBP PET variables did reach statistical significance in the MCI group. In the AD dementia group, no significant correlations were observed (Table 2).
3.3. Regression analyses of biomarker variables
After Holm-Bonferroni correction, logistic regression modeling of biomarkers found no variables that statistically significantly differentiated HC from MCI (Table 3). Amyloid biomarkers alone (FBP PET and CSF Aβ1-42) significantly distinguished between diagnostic groups when comparing HC and AD dementia groups (FBP PET, P=.0002; CSF Aβ1-42, P=.0007). CSF t-tau significantly differentiated AD dementia from both HC (P<.0001) and MCI groups (P=.0003), and CSF p-tau distinguished between HC and AD dementia groups (P=.0001).
Table 3.
Logistic regression analyses of clinical diagnostic group on CSF and FBP PET variables, adding 1 biomarker to the other to determine an additive contribution in distinguishing among groups.
| HC vs MCI | MCI vs AD | HC vs. AD | ||||
|---|---|---|---|---|---|---|
|
| ||||||
| Chi-Square (df) | P-value | Chi-Square (df) | P-value | Chi-Square (df) | P-value | |
| Test FBP PET without CSF | 5.7502 (1) | .0165 | 8.7197 (1) | .0031 | 14.3044 (1) | .0002 |
| Test of CSF Aβ1-42 when added to FBP PET | 0.5176 (1) | .4718 | 4.6972 (1) | .0302 | 1.9339 (1) | .1643 |
| Test of CSF t-tau when added to FBP PET | 1.6375 (1) | .2007 | 6.9011 (1) | .0086 | 10.7866 (1) | .0010 |
| Test of CSF p-tau when added to FBP PET | 0.3143 (1) | .5751 | 2.0160 (1) | .1556 | 9.5094 (1) | .0020 |
| Test Aβ1-42 without FBP PET | 0.6942 (1) | .4047 | 9.9783 (1) | .0016 | 11.5618 (1) | .0007 |
| Test of FBP PET when added to CSF Aβ1-42 | 5.5238 (1) | .0188 | 1.7866 (1) | .1813 | 3.5791 (1) | .0585 |
| Test CSF t-tau without FBP PET | 4.5379 (1) | .0332 | 13.2332 (1) | .0003 | 15.2843 (1) | <.0001 |
| Test of FBP PET when added to CSF t-tau | 2.6964 (1) | .1006 | 2.8014 (1) | .0942 | 5.4003 (1) | .0201 |
| Test CSF p-tau without FBP PET | 2.2812 (1) | .1310 | 6.1506 (1) | .0131 | 14.5239 (1) | .0001 |
| Test of FBP PET when added to CSF p-tau | 3.8047 (1) | .0511 | 5.1704 (1) | .0230 | 7.4634 (1) | .0063 |
Abbreviations: Aβ1-42 = beta-amyloid protein; AD = Alzheimer’s disease; CSF = cerebrospinal fluid; df = degrees of freedom; FBP PET = florbetapir-F18 positron emission tomography; HC = healthy controls; MCI = mild cognitive impairment; p-tau = phosphorylated tau181P; t-tau = total tau; vs = versus.
FBP PET includes all 6 regions of interest, and CSF includes all 5, unless otherwise specified. P-values that meet statistical significance with Holm-Bonferroni-corrected cutoff are in bold.
Table 3 also shows the effect of adding CSF or FBP PET variables to the other biomarker type to assess any additional contribution to differentiating diagnostic groups (where the reported P-values represent the impact of just the additional information). No significant gain in differentiation was observed when testing FBP PET variables in the presence of CSF variables for any group comparison. However, adding CSF t-tau or CSF p-tau to FBP PET significantly improved differentiation between HC and AD dementia groups.
4. Discussion
This cross-sectional analysis explored relationships between 2 types of AD biomarkers, amyloid PET imaging (FBP PET) and CSF analytes (Aβ1-42, t-tau, and p-tau), for their ability to differentiate clinical diagnostic group status among HC, MCI, and AD dementia subjects in ADNI. Both amyloid-related biomarkers were highly correlated with each other. Overall, the amyloid-related biomarkers were not appreciably different with respect to categorical clinical classification in that adding one to the other in logistic regressions did not improve classification.
Specifically, in logistic regression analyses, neither CSF Aβ1-42 nor FBP PET distinguished HC and MCI, probably because amyloid pathology in those who could later progress to clinical AD had already manifested. However, CSF Aβ1-42 and FBP PET each distinguished HC from AD groups, as did CSF t-tau and p-tau. Additionally, CSF t-tau also significantly differentiated AD dementia from MCI, and CSF p-tau distinguished between HC and AD dementia groups.
These findings with CSF tau are consistent with CSF tau abnormalities manifesting later and progressively in the disease, as compared to amyloid plaque, which exhibits substantial deposition by the time patients present with MCI [9].
CSF Aβ1-42 but not FBP PET significantly distinguished MCI from AD dementia groups; however, FBP PET was close to the threshold applied by the Holm-Bonferroni correction for the multiple comparisons method, and it is possible that a better-powered study might have found a different result. Once a person has positive binary status the rate of amyloid SUVR increase is slower during MCI and dementia stages than in the decades before MCI [15].
We found a number of statistically significant correlations between the biomarker types, especially those that involved beta-amyloid. Although significant correlations between CSF tau measures and FBP PET variables were seen, the values of the correlation coefficients were relatively lower unless CSF tau was in a ratio with CSF Aβ1-42.
Within the HC and MCI groups, we found some strong and significant correlations for FBP PET with CSF Aβ1-42, with the anterior and posterior cingulate ROIs and composite SUVRs being the most notable. This is consistent with the known neuroanatomical progression pattern of AD where cingulate gyri are affected early with beta-amyloid plaque. In the AD dementia group, the highest correlations were between CSF Aβ1-42 and FBP PET, but no correlations reached statistical significance. However, it needs to be considered that the sample size for the AD dementia group was much smaller than the other groups.
Interestingly, CSF t-tau provided differentiation in the comparisons of HC versus AD dementia and MCI versus AD dementia, but not HC versus MCI. This suggests that amyloid-related biomarkers are informative as adjunctive tests for establishing an AD diagnosis since the associated pathology starts long before clinical symptoms appear, while tau may be more helpful for staging because it accumulates in the later stages of the disease, as has been described previously. While CSF Aβ1-42 changes are observed 5 to 10 years before conversion of MCI to AD dementia, CSF t-tau and p-tau seem to be markers of later stage pathology [24]. Thomann and colleagues associated changes in CSF t-tau and p-tau with neurodegenerative changes in MCI subjects who converted to early AD dementia [25]. Alternatively, some studies have suggested that tau abnormalities at the cellular level may begin in the asymptomatic period before or simultaneously with amyloid [26], but our current clinical biomarker methodologies may not be targeted or sensitive enough to detect those [27].
Doré and colleagues recently described longitudinal (18- and 36-months) relationships among Aβ deposition, cortical thickness, and memory [28]. They reported a faster rate of gray matter atrophy in the temporal cortex and hippocampi and greater episodic memory impairment in clinically unimpaired individuals who were amyloid positive on PiB PET than those who were amyloid negative [28]. A longitudinal study published by the Australian Imaging Biomarkers and Lifestyle (AIBL) research group estimated that it takes 19.2 years (95% CI 16.8–22.5) for subjects to progress from the threshold of PiB PET positivity to amyloid levels observed in AD dementia [15]. After the emergence of symptoms of AD, the rate of Aβ deposition slowed and then plateaued at the dementia stage [15]. Additionally, a study of 401 ADNI subjects found that reduction in the CSF Aβ1-42 level becomes dynamic early, whereas changes in CSF t-tau levels and adjusted hippocampal volumes occur later and may be biomarkers of downstream pathophysiological processes [29]. However, a study by Driscoll and colleagues in non-demented individuals did not observe a correlation between the level of amyloid load and longitudinal brain volume changes [30].
The generalizability of our results to the broader population is uncertain and potentially limited by the study sample. We used data from ADNI-GO and ADNI-2 cohorts, which represents a selected convenience sample including subjects with amnestic MCI, but also higher education and cognitive reserve. Compared to the ADNI cohort, the population-based sample in the Mayo Clinic Study of Aging (MCSA) [31] was older and less educated, and had lower MMSE scores and less frequent family history of AD. The rate of hippocampal volume decline was larger in ADNI subjects compared with MCSA, suggesting more advanced brain pathology in ADNI subjects [31]. Additionally, analyzing early and late MCI subjects as 1 group might have affected our findings. Further, because ADNI used a central laboratory to test CSF, the lack of standardization of CSF AD biomarker measurements across clinical sites and assays may limit the applicability of our results to clinical practice. Finally, the analyses presented here are based on cross-sectional and not longitudinal data. Prospective, longitudinal studies are needed to confirm or refute our findings. The strengths of our study are the relatively large HC and MCI sample sizes and the combination of CSF and FBP PET measures where most prior work was reported using PiB PET.
In conclusion, we found some unique characteristics, but also considerable overlap between CSF and FBP PET measures when assessing their ability to distinguish among pairs of HC, MCI, and AD dementia groups. We report both composite and ROI correlations for FBP PET with CSF. Our findings of differences in differentiation of AD stages by amyloid versus tau biomarkers might aid in the development of further diagnostic and staging tools for AD.
Research in Context.
Systematic Review
The authors reviewed the currently available literature on florbetapir positron emission tomography and cerebrospinal fluid biomarkers in Alzheimer’s Disease and combined their findings with their clinical experience in this patient population.
Interpretation
The authors found some unique characteristics but also considerable overlap between cerebrospinal fluid and florbetapir positron emission tomography measures when assessing their ability to distinguish among pairs of healthy control, mild cognitive impairment, and Alzheimer’s Disease groups using a variety of analytic methods. These findings of differences in differentiation of Alzheimer’s disease stages by amyloid versus tau biomarkers might aid in the development of further diagnostic and staging tools for Alzheimer’s Disease.
Future Directions
Prospective, longitudinal studies are needed to confirm the results of the presented retrospective cross-sectional analyses.
Acknowledgments
Data collection and sharing for this project were funded by the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (National Institutes of Health Grant U01 AG024904) and DOD ADNI (Department of Defense award number W81XWH-12-2-0012). ADNI is funded by the National Institute on Aging, the National Institute of Biomedical Imaging and Bioengineering, and through generous contributions from the following: Alzheimer’s Association; Alzheimer’s Drug Discovery Foundation; BioClinica, Inc.; Biogen Idec Inc.; Bristol-Myers Squibb Company; Eisai Inc.; Elan Pharmaceuticals, Inc.; Eli Lilly and Company; F. Hoffmann-La Roche Ltd and its affiliated company Genentech, Inc.; GE Healthcare; Innogenetics, N.V.; IXICO Ltd.; Janssen Alzheimer Immunotherapy Research & Development, LLC.; Johnson & Johnson Pharmaceutical Research & Development LLC.; Medpace, Inc.; Merck & Co., Inc.; Meso Scale Diagnostics, LLC.; NeuroRx Research; Novartis Pharmaceuticals Corporation; Pfizer Inc.; Piramal Imaging; Servier; Synarc Inc.; and Takeda Pharmaceutical Company. The Canadian Institutes of Health Research is providing funds to support ADNI clinical sites in Canada. Private sector contributions are facilitated by the Foundation for the National Institutes of Health (www.fnih.org). The grantee organization is the Northern California Institute for Research and Education, and the study is coordinated by the Alzheimer’s Disease Cooperative Study at the University of California, San Diego. ADNI data are disseminated by the Laboratory for Neuroimaging at the University of Southern California. The authors would like to thank Vicki Poole Hoffmann, full-time employee of Eli Lilly and Company, for careful review of the manuscript; Alexandra Heinloth, full-time employee of inVentiv Health Clinical; for writing assistance; Jia Sun, full-time employee of BC Forward, for assistance with acquiring the data and careful review of the manuscript; Terri Tucker, Sree Lakshmi, Angela Lorio, and Harini Muthyala (all full-time employees of inVentiv Health Clinical) for editorial assistance; and Linda Tabas (full-time employee of Eli Lilly and Company) for project management assistance. Eli Lilly and Company contracted inVentiv Health Clinical, LLC, for writing and editorial assistance.
6. List of abbreviations
- Aβ
amyloid beta
- AD
Alzheimer’s disease
- ADAS-Cog11
Alzheimer’s Disease Assessment Scale, 11-item cognitive subscale
- ADAS-Cog13
Alzheimer’s Disease Assessment Scale, 13-item cognitive subscale
- ADNI
Alzheimer’s Disease Neuroimaging Initiative
- AIBL
Australian Imaging Biomarkers and Lifestyle
- APOE4
apolipoprotein E epsilon 4
- CDR
Clinical Dementia Rating
- CI
confidence interval
- CSF
cerebrospinal fluid
- EVIQ
Estimated Verbal Intelligence Quotient
- FAQ
Functional Activities Questionnaire
- FDA
Food and Drug Administration
- FBP PET
florbetapir-F18 positron emission tomography
- GDS
Geriatric Depression Scale
- HC
healthy control
- MCI
mild cognitive impairment
- MCSA
Mayo Clinic Study of Aging
- MMSE
Mini-Mental State Examination
- MRI
magnetic resonance imaging
- NIA
National Institute on Aging
- NIBIB
National Institute of Biomedical Imaging and Bioengineering
- NINCDS-ADRDA
National Institute of Neurological and Communicative Disorders and Stroke - Alzheimer’s Disease and Related Disorders Association
- NPI-Q
Neuropsychiatric Inventory-Questionnaire
- p-tau
phosphorylated tau181P
- PET
positron emission tomography
- PiB
Pittsburgh compound B
- ROI
region of interest
- SUVR
standard uptake value ratio
- t-tau
total tau
Footnotes
Posters related to this manuscript were presented at the 2013 annual meetings of the Alzheimer’s Association International Conference (13–18 July 2013, Boston) and the American Neurological Association (13–15 October 2013, New Orleans); an oral presentation was delivered at the Academy of Psychosomatic Medicine (13–16 November 2013, Tucson).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Honjo K, Black SE, Verhoeff NP. Alzheimer’s disease, cerebrovascular disease, and the beta-amyloid cascade. Can J Neurol Sci. 2012;39:712–728. doi: 10.1017/s0317167100015547. [DOI] [PubMed] [Google Scholar]
- 2.Beach TG, Monsell SE, Phillips LE, Kukull W. Accuracy of the clinical diagnosis of Alzheimer disease at National Institute on Aging Alzheimer Disease Centers, 2005–2010. J Neuropathol Exp Neurol. 2012;71:266–273. doi: 10.1097/NEN.0b013e31824b211b. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Clark CM, Pontecorvo MJ, Beach TG, Bedell BJ, Coleman RE, Doraiswamy PM, et al. Cerebral PET with florbetapir compared with neuropathology at autopsy for detection of neuritic amyloid-beta plaques: a prospective cohort study. Lancet Neurol. 2012;11:669–678. doi: 10.1016/S1474-4422(12)70142-4. [DOI] [PubMed] [Google Scholar]
- 4.Ewers M, Insel P, Jagust WJ, Shaw L, Trojanowski JQ, Aisen P, et al. CSF biomarker and PIB-PET-derived beta-amyloid signature predicts metabolic, gray matter, and cognitive changes in nondemented subjects. Cereb Cortex. 2012;22:1993–2004. doi: 10.1093/cercor/bhr271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Fagan AM, Mintun MA, Mach RH, Lee SY, Dence CS, Shah AR, et al. Inverse relation between in vivo amyloid imaging load and cerebrospinal fluid Abeta42 in humans. Ann Neurol. 2006;59:512–519. doi: 10.1002/ana.20730. [DOI] [PubMed] [Google Scholar]
- 6.Rosenmann H. CSF biomarkers for amyloid and tau pathology in Alzheimer’s disease. J Mol Neurosci. 2012;47:1–14. doi: 10.1007/s12031-011-9665-5. [DOI] [PubMed] [Google Scholar]
- 7.Vanderstichele H, Bibl M, Engelborghs S, Le BN, Lewczuk P, Molinuevo JL, et al. Standardization of preanalytical aspects of cerebrospinal fluid biomarker testing for Alzheimer’s disease diagnosis: a consensus paper from the Alzheimer’s Biomarkers Standardization Initiative. Alzheimers Dement. 2012;8:65–73. doi: 10.1016/j.jalz.2011.07.004. [DOI] [PubMed] [Google Scholar]
- 8.Teunissen CE, Verwey NA, Kester MI, van Uffelen K, Blankenstein MA. Standardization of assay procedures for analysis of the CSF biomarkers amyloid beta((1-42)), tau, and phosphorylated tau in Alzheimer’s disease: Report of aninternational workshop. Int J Alzheimers Dis. 2010 doi: 10.4061/2010/635053. pii: 635053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jack CR, Jr, Knopman DS, Jagust WJ, Petersen RC, Weiner MW, Aisen PS, et al. Tracking pathophysiological processes in Alzheimer’s disease: an updated hypothetical model of dynamic biomarkers. Lancet Neurol. 2013;12:207–216. doi: 10.1016/S1474-4422(12)70291-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Blennow K, Dubois B, Fagan AM, Lewczuk P, de Leon MJ, Hampel H. Clinical utility of cerebrospinal fluid biomarkers in the diagnosis of early Alzheimer’s disease. Alzheimers Dement. 2014 doi: 10.1016/j.jalz.2014.02.004. pii: S1552-5260(14)00066-1. Epub ahead of print. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Vergara C, Ordóñez-Gutiérrez L, Wandosell F, Ferrer I, Del Rio JA, Gavin R. Role of PrP expression in tau protein levels and phosphorylation in Alzheimer’s disease evolution. Mol Neurobiol. 2014 Jun 26; doi: 10.1007/s12035-014-8793-7. Epub ahead of print. [DOI] [PubMed] [Google Scholar]
- 12.Wildsmith KR, Schauer SP, Smith AM, Arnott D, Zhu Y, Haznedar J, et al. Identification of longitudinally dynamic biomarkers in Alzheimer’s disease cerebrospinal fluid by targeted proteomics. Mol Neurodegener. 2014;9:22. doi: 10.1186/1750-1326-9-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fagan AM, Roe CM, Xiong C, Mintun MA, Morris JC, Holtzman DM. Cerebrospinal fluid tau/beta-amyloid(42) ratio as a prediction of cognitive decline in nondemented older adults. Arch Neurol. 2007;64:343–349. doi: 10.1001/archneur.64.3.noc60123. [DOI] [PubMed] [Google Scholar]
- 14.SantaCruz KS, Sonnen JA, Pezhouh MK, Desrosiers MF, Nelson PT, Tyas SL. Alzheimer disease pathology in subjects without dementia in 2 studies of aging: the Nun Study and the Adult Changes in Thought Study. J Neuropathol Exp Neurol. 2011;70:832–840. doi: 10.1097/NEN.0b013e31822e8ae9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Villemagne VL, Burnham S, Bourgeat P, Brown B, Ellis KA, Salvado O, et al. Amyloid beta deposition, neurodegeneration, and cognitive decline in sporadic Alzheimer’s disease: a prospective cohort study. Lancet Neurol. 2013;12:357–367. doi: 10.1016/S1474-4422(13)70044-9. [DOI] [PubMed] [Google Scholar]
- 16.Dukart J, Mueller K, Villringer A, Kherif F, Draganski B, Frackowiak R, et al. Relationship between imaging biomarkers, age, progression and symptom severity in Alzheimer’s disease. Neuroimage Clin. 2013;3:84–94. doi: 10.1016/j.nicl.2013.07.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Temple V, Jozsvai E, Konstantareas MM, Hewitt TA. Alzheimer dementia in Down’s syndrome: the relevance of cognitive ability. J Intellect Disabil Res. 2001;45(Part 1):47–55. doi: 10.1046/j.1365-2788.2001.00299.x. [DOI] [PubMed] [Google Scholar]
- 18.Fritsch T, McClendon MJ, Smyth KA, Lerner AJ, Chen CH, Petot GJ, et al. Effects of educational attainment on the clinical expression of Alzheimer’s disease: results from a research registry. Am J Alzheimers Dis Other Demen. 2001;16:369–376. doi: 10.1177/153331750101600606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Schutte DL, Reed D, Decrane S, Ersig AL. Saitohin and APOE polymorphisms influence cognition and function in persons with advanced Alzheimer Disease. Dement Geriatr Cogn Disord. 2011;32:94–102. doi: 10.1159/000329542. [DOI] [PubMed] [Google Scholar]
- 20.Potter GG, Steffens DC. Contribution of depression to cognitive impairment and dementia in older adults. Neurologist. 2007;13:105–117. doi: 10.1097/01.nrl.0000252947.15389.a9. [DOI] [PubMed] [Google Scholar]
- 21.Nelson PT, Kukull WA, Frosch MP. Thinking outside the box: Alzheimer-type neuropathology that does not map directly onto current consensus recommendations. J Neuropathol Exp Neurol. 2010;69:449–454. doi: 10.1097/NEN.0b013e3181d8db07. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Fleisher AS, Chen K, Liu X, Roontiva A, Thiyyagura P, Ayutyanont N, et al. Using positron emission tomography and florbetapir F18 to image cortical amyloid in patients with mild cognitive impairment or dementia due to Alzheimer disease. Arch Neurol. 2011;68:1404–1411. doi: 10.1001/archneurol.2011.150. [DOI] [PubMed] [Google Scholar]
- 23.Holm S. A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics. 1979;6:65–70. [Google Scholar]
- 24.Buchhave P, Minthon L, Zetterberg H, Wallin AK, Blennow K, Hansson O. Cerebrospinal fluid levels of beta-amyloid 1-42, but not of tau, are fully changed already 5 to 10 years before the onset of Alzheimer dementia. Arch Gen Psychiatry. 2012;69:98–106. doi: 10.1001/archgenpsychiatry.2011.155. [DOI] [PubMed] [Google Scholar]
- 25.Thomann PA, Kaiser E, Schonknecht P, Pantel J, Essig M, Schroder J. Association of total tau and phosphorylated tau 181 protein levels in cerebrospinal fluid with cerebral atrophy in mild cognitive impairment and Alzheimer disease. J Psychiatry Neurosci. 2009;34:136–142. [PMC free article] [PubMed] [Google Scholar]
- 26.Musiek ES, Holtzman DM. Origins of Alzheimer’s disease: reconciling cerebrospinal fluid biomarker and neuropathology data regarding the temporal sequence of amyloid-beta and tau involvement. Curr Opin Neurol. 2012;25:715–720. doi: 10.1097/WCO.0b013e32835a30f4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Braak H, Zetterberg H, Del Tredici K, Blennow K. Intraneuronal tau aggregation precedes diffuse plaque deposition, but amyloid-beta changes occur before increases of tau in cerebrospinal fluid. Acta Neuropathol. 2013;126:631–641. doi: 10.1007/s00401-013-1139-0. [DOI] [PubMed] [Google Scholar]
- 28.Doré V, Villemagne VL, Bourgeat P, Fripp J, Acosta O, Chetélat G, et al. Cross-sectional and longitudinal analysis of the relationship between Abeta deposition, cortical thickness, and memory in cognitively unimpaired individuals and in Alzheimer disease. JAMA Neurol. 2013;70:1–9. doi: 10.1001/jamaneurol.2013.1062. [DOI] [PubMed] [Google Scholar]
- 29.Jack CR, Jr, Vemuri P, Wiste HJ, Weigand SD, Aisen PS, Trojanowski JQ, et al. Evidence for ordering of Alzheimer disease biomarkers. Arch Neurol. 2011;68:1526–1535. doi: 10.1001/archneurol.2011.183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Driscoll I, Zhou Y, An Y, Sojkova J, Davatzikos C, Kraut MA, et al. Lack of association between 11C-PiB and longitudinal brain atrophy in non-demented older individuals. Neurobiol Aging. 2011;32:2123–2130. doi: 10.1016/j.neurobiolaging.2009.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Whitwell JL, Wiste HJ, Weigand SD, Rocca WA, Knopman DS, Roberts RO, et al. Comparison of imaging biomarkers in the Alzheimer Disease Neuroimaging Initiative and the Mayo Clinic Study of Aging. Arch Neurol. 2012;69:614–622. doi: 10.1001/archneurol.2011.3029. [DOI] [PMC free article] [PubMed] [Google Scholar]