Key PointsQuestion
What is the extent and the role of tau, measured by 18F-AV1451 positron emission tomography, in patients with subcortical vascular cognitive impairment?
Findings
In this cross-sectional study of patients with subcortical vascular cognitive impairment, amyloid-β positivity and cerebral small vessel disease score were each independently associated with increased AV1451 uptake in the medial temporal and inferior temporal regions, respectively. Level of AV1451 uptake mediated the associations of amyloid-ß or cerebral small vessel disease with cognitive impairment.
Meaning
The findings suggest that in subcortical vascular cognitive impairment, both amyloid-β and cerebral small vessel disease are independently associated with increased tau accumulation; furthermore, tau burden played a pivotal role because it was the final common pathway for cognitive impairment in patients with subcortical vascular cognitive impairment.
Importance
Amyloid-β (Aβ), tau, and cerebral small vessel disease (CSVD), which occasionally coexist, are the most common causes of cognitive impairments in older people. However, whether tau is observed in patients with subcortical vascular cognitive impairment (SVCI), as well as its associations with Aβ and CSVD, are not yet established. More importantly, the role of tau underlying cognitive impairments in SVCI is unknown.
Objective
To investigate the extent and the role of tau in patients with SVCI using 18F-AV1451, which is a new ligand to detect neurofibrillary tangles in vivo.
Design, Setting, and Participants
This cross-sectional study recruited 64 patients with SVCI from June 2015 to December 2016 at Samsung Medical Center, Seoul, Korea. The patients had significant ischemia on brain magnetic resonance imaging, defined as periventricular white matter hyperintensity at least 10 mm and deep white matter hyperintensity at least 25 mm. We excluded 3 patients with SVCI owing to segmentation error during AV1451 positron emission tomography analysis.
Main Outcomes and Measures
We calculated CSVD scores based on the volumes of white matter hyperintensities, numbers of lacunes, and microbleeds using magnetic resonance imaging data. The presence of Aβ was assessed using fluorine 18–labeled (18F) florbetaben positron emission tomography. Tau was measured using 18F-AV1451 positron emission tomography. We determined the spreading order of tau by sorting the regional frequencies of cortical involvement. We evaluated the complex associations between Aβ, CSVD, AV1451 uptake, and cognition in patients with SVCI.
Results
Of the 61 patients with SVCI, 44 (72.1%) were women and the mean (SD) age was 78.7 (6.3) years. Patients with SVCI, especially patients with Aβ-negative SVCI, showed higher AV1451 uptake in the inferior temporal areas compared with normal control individuals. In patients with SVCI, Aβ positivity and CSVD score were each independently associated with increased AV1451 uptake in the medial temporal and inferior temporal regions, respectively. Involvement frequency of AV1451 uptake in the fusiform gyrus, inferior temporal, and precuneus regions were higher than that in the parahippocampal region. In patients with SVCI, higher AV1451 uptake in the inferior temporal and medial temporal regions correlated with worse language and general cognitive function. In patients with SVCI, Aβ positivity and CSVD score each correlated with worse general cognitive function, which was completely mediated by AV1451 uptake in the entorhinal cortex and inferior temporal gyrus, respectively.
Conclusions and Relevance
Our findings suggest that in SVCI, both Aβ and CSVD were independently associated with increased tau accumulation. Furthermore, tau burden played a pivotal role because it was the final common pathway for the cognitive impairment in patients with SVCI.
Alzheimer disease (AD) and cerebral small vessel disease (CSVD) are the most common causes of cognitive impairments in older people.1 Previous studies show that AD pathologies, such as amyloid-β (Aβ) plaques and neurofibrillary tangles (NFT), and CSVD have a strong association.2 With the rapid development of molecular imaging, it is now possible to detect such AD pathologies in vivo. In fact, previous studies from our group suggested that approximately 30% of subcortical vascular cognitive impairment (SVCI), characterized by extensive CSVD, has significant Aβ burden on positron emission tomography (PET).3
AV1451, a new ligand for paired helical filament tau,4 is able to detect NFT in vivo and is reported to be correlated with Braak NFT stage, cortical atrophy, and cognitive status.5,6 However, the associations between AV1451 uptake and CSVD have not been fully investigated in patients with SVCI. Considering that Aβ triggers NFT spread according to the amyloid cascade model,7 increased AV1451 uptake might be driven by Aβ burden in patients with SVCI. Alternatively, increased AV1451 uptake may exist in the absence of Aβ because pathological studies have shown that tau burden was elevated in patients with acute ischemic stroke and in mice with chronic hypoperfusion.8,9
More importantly, the role of tau in SVCI is not yet fully understood. AV1451 PET studies have shown that in AD, tau, rather than Aβ, is more tightly correlated with cognition6 and that there is at least some evidence to suggest that tau accumulation may occur after stroke. Previous studies showed that imaging markers, such as cortical thickness, partially mediated the association of CSVD or Aβ with cognition.10-12 In the meantime, tau is reported to have close correlations with brain structural changes or cognition. Thus, CSVD and Aβ leading to cognitive impairment may both be mediated by increased tau burden. In other words, tau may be the final common pathway for cognitive symptoms.
In this study, we investigated the extent and role of tau in patients with SVCI using AV1451 PET. We hypothesized that CSVD burden, measured by a combined score of white matter hyperintensities (WMH), lacunes, and microbleeds, would be associated with increased AV1451 uptake regardless of Aβ positivity on PET. We also hypothesized that AV1451 uptake mediates the associations of CSVD or Aβ with cognition.
We recruited 64 patients with SVCI and 9 patients with AD at Samsung Medical Center, Seoul, Korea, from June 2015 to December 2016. Twenty patients with amnestic mild cognitive impairment (aMCI) and 20 amyloid PET-negative normal control (NC) individuals were recruited from Gangnam Severance Hospital, Seoul, Korea, from January 2015 to April 2016. Patients had to meet the following criteria to be diagnosed as having SVCI: (1) subjective cognitive complaint by the patient or caregiver; (2) objective cognitive impairment less than the 16th percentile of the norm in any domain including language, visuospatial, memory, or frontal function on neuropsychological tests; (3) significant ischemia on brain magnetic resonance imaging (MRI), defined as periventricular WMH at least 10 mm and deep WMH at least 25 mm; and (4) focal neurologic symptoms or signs. Alzheimer disease was diagnosed based on National Institute on Aging–Alzheimer’s Association research criteria for probable AD.13 To be diagnosed as having aMCI, patients had to meet Petersen criteria and show objective memory impairment less than the 1.5 SD of the norm in at least 1 memory test. In this study, to compare with SVCI, we combined AD and aMCI into Alzheimer disease cognitive impairment (ADCI). Normal control individuals were all characterized by the following: (1) no history of neurologic or psychiatric disorders and (2) normal cognitive function determined using neuropsychological tests (greater than 1.5 SD of the norm). No patient in the AD, aMCI,or NC groups had significant ischemia on brain MRI, defined as periventricular WMH at least 10 mm and deep WMH at least 25 mm.
All participants were assessed through clinical interviews and neurologic examinations, and clinical diagnosis was established by consensus among a multidisciplinary team. All patients underwent laboratory testing, including complete blood cell count, blood chemistry, vitamin B12/folate analysis, syphilis serology, thyroid function tests, and apolipoprotein E genotyping. We excluded participants who showed structural lesions including territorial cerebral infarction, cortical stroke, brain tumor, hippocampal sclerosis, or vascular malformation on brain MRI. We further excluded 3 patients with SVCI, 2 with ADCI, and 1 NC participant owing to segmentation error during AV1451 PET analysis. Thus, the final sample analyzed in this study included 61 patients with SVCI, 27 patients with ADCI, and 19 NC individuals.
This study was approved by the institutional review boards of Samsung Medical Center and Gangnam Severance Hospital. We obtained informed written consent from all participants.
All participants underwent neuropsychologic tests using a standardized neuropsychologic battery, the Seoul Neuropsychological Screening Battery.14 We calculated composite scores for language, visuospatial, memory, and frontal-executive functions.15 The language function was based on the Boston naming test (range, 0-60). The visuospatial score was based on the Rey Complex Figure Test (range, 0-36). The memory score was calculated by summing scores from verbal memory tests (Seoul Verbal Learning Test immediate recall, delayed recall, and recognition score) and visual memory tests (Rey Complex Figure Test immediate recall, delayed recall, and recognition score; range, 0-144). The frontal-executive score was calculated by summing scores from the category word generation test, phonemic word generation test, and Stroop color-reading test (range, 0-55). Global cognitive function was assessed using the Mini-Mental State Examination (range, 0-30). The norms for each of these tests were based on assessments of 1067 normal Korean participants.
We acquired standardized T2, 3-dimensional T1 turbo field echo images, 3-dimensional fluid-attenuated inversion recovery, and T2-weighted gradient echo MRIs from all participants at Samsung Medical Center using a 3.0-T MRI scanner (Philips 3.0T Achieva; Philips Healthcare) as previously described.10
Assessment of CSVD Scores
We quantified overall CSVD burden using a new scoring system because the previously proposed scoring system16 could not well categorize SVD burden in patients with SVCI. We calculated CSVD scores (0-3) by adding scores for lacune (1, if greater than median value of 4 lacunes), microbleed (1, if greater than median value of 4 microbleeds), and WMH (1, if greater than median value of 40.6 mL).
Lacunes were defined as small lesions (≤15 mm and ≥3 mm in diameter) with low signal on T1-weighted images, high signal on T2-weighted images, and a perilesional halo on 80 axial slices of fluid-attenuated inversion recovery images as proposed by Wardlaw et al.17 Cerebral microbleeds were defined as lesions 10 mm or less in diameter on 20 T2-weighted gradient echo MRI axial slides using criteria proposed by Wardlaw et al.17 We quantified WMH volume (in milliliters) on fluid-attenuated inversion recovery images using an automated method.
18F-Florbetaben PET Acquisition and Analysis
Patients underwent fluorine 18–labeled (18F) florbetaben PET at Samsung Medical Center using a Discovery STe PET/computed tomography scanner (GE Medical Systems) in 3-dimensional scanning mode that examined 47 slices of 3.3-mm thickness spanning the entire brain. Computed tomography images were acquired using a 16-slice helical computed tomography (140 KeV, 80 mA; 3.75-mm section width) for attenuation correction. For 18F-florbetaben PET, a 20-minute emission PET scan with dynamic mode (consisting of 4 × 5-minute frames) was performed 90 minutes after injection of approximately 300 MBq of 18F-florbetaben. Three-dimensional PET images were reconstructed in a 128 × 128 × 48 matrix with 2 × 2 × 3.27-mm voxel size using the ordered-subsets expectation maximization algorithm (iteration, 4 and subset, 20).
We defined amyloid PET to be positive when visual assessment of florbetaben PET was scored as 2 or 3 on the brain Aβ plaque load scoring system.18 Our visual assessment matched well with a binarized global florbetaben PET binding (standardized uptake value ratio [SUVR] cutoff, 1.407), as comparison of the 2 methods resulted in high accuracy of 94.4% (sensitivity, 91.5% [43 of 47] and specificity, 96.7% [58 of 60]) (eFigure 1 in the Supplement).
18F-AV-1451 PET Acquisition and Analysis
All patients underwent 18F-AV-1451 PET at Gangnam Severance Hospital using a Biograph molecular computed tomography PET/computed tomography scanner (Siemens Medical Solutions). At 80 minutes after intravenous bolus injections of approximately 280 MBq of 18F-AV1451, tau PET images were acquired for 20 minutes. Prior to the PET scan, we applied a head holder to minimize head motion and acquired brain computed tomography images for attenuation correction. Three-dimensional PET images were reconstructed in a 256 × 256 × 223 matrix with 1.591 × 1.591 × 1-mm voxel size using the ordered-subsets expectation maximization algorithm (iteration, 6 and subset, 16).
Positron emission tomographic images were coregistered to individual MRIs, which were normalized to a T1-weighted MRI template. We applied normalized parameters to transform coregistered PET images into the MRI template. Standardized uptake value ratios were calculated using cerebellar gray matter as a reference region. Then, SUVR images were spatially smoothed with an 8-mm Gaussian kernel. Data processing was performed using SPM, version 8 (SPM) through Matlab, version 2014b (MathWorks).
Cortical Spreading Order of AV1451 Uptake
To determine the order of cortical spreading, we assumed that regions with earlier appearance of pathology would show increased binding in a greater number of participants, as suggested by a previous study.5 We parcellated the whole cerebral cortex into 25 regions in each hemisphere. We selected 12 NC individuals whose AV1451 SUVR values in the entorhinal cortex were less than 1.20 to obtain regional mean and standard deviation values for SUVR. Then, regional z scores of tau PET were calculated for the regions that showed AV1451 SUVR greater than 1.0 in patients with SVCI (green, yellow, or red in Figure 1A) and the regions with z score values greater than 2.5 were considered to have extensive tau accumulation. We sorted these regions in descending order by number of participants whose regional z scores were greater than 2.5 for each region.
For descriptive statistics, we used the χ2 test and analysis of variance followed by Bonferroni post hoc analysis to compare groups.
To compare differences in AV1451 uptake between groups, we performed multiple linear regression after controlling for age. We also performed multiple linear regression to evaluate voxelwise associations of Aβ or CSVD scores with AV1451 uptake. We entered Aβ positivity, CSVD score, and age as independent variables. We determined an uncorrected P value of less than .01 to be significant, and the P value was 2-sided.
To assess differences of regional involvement frequency across regions, we used a bootstrapping method with 1000 resamples to derive 95% confidence intervals and standard errors. For all combinations of regional pairs, asymptotic P values were calculated and corrected for multiple comparisons using Bonferroni method.
To evaluate the associations between AV1451 SUVR and the top 5 regions of interests that showed early involvement of tau and cognition among patients with SVCI, we performed multiple linear regression analyses after controlling for age and education. To correct for multiple regions and multiple cognitive domains tested, false discovery rate (FDR) corrected P less than .05 was considered significant. To evaluate the association of Aβ or CSVD with cognition, we performed multiple linear regression analyses after controlling for age and education. Then, path analyses were used to evaluate whether regional AV1451 SUVR mediated associations between Aβ and cognition or between CSVD score and cognition among patients with SVCI, after controlling for age and education (eFigure 2 in the Supplement).
Statistical analyses were conducted with Predictive Analysis Software Statistics, version 22 software (SPSS Inc), and Amos, version 18.0 software (SPSS Inc), was used for all path analyses using maximum likelihood estimation.
Demographic Characteristics of Participants
Of the 61 patients with SVCI, 44 (72.1%) were women and the mean (SD) age was 78.7 (6.3) years. Patients with SVCI were older and had a lower educational level than patients with ADCI. In patients with SVCI, 44.3% (27 of 61) were amyloid PET positive, whereas in patients with ADCI, 74.1% (20 of 27) (63.2% [12 of 19] of aMCI and 100% [8 of 8] of AD) were amyloid PET positive. The mean Mini-Mental State Examination score did not differ between patients with SVCI and patients with ADCI (Table 1).
AV1451 Distribution in NC, SVCI, and ADCI
Figure 1A shows mean AV1451 uptakes in NC, SVCI, and ADCI. Compared with NC individuals, both Aβ-negative and Aβ-positive patients with SVCI showed higher AV1451 uptake, especially in the bilateral inferior temporal areas (Figure 1B). Compared with NC individuals, patients with ADCI showed high AV1451 uptake in more extended areas including the bilateral temporal, parietal, and frontal areas. Furthermore, compared with SVCI, patients with ADCI showed high AV1451 uptake in the bilateral temporal, parietal, and frontal areas (Figure 1B).
Association Between Aβ, CSVD Score, and AV1451 in SVCI
In SVCI, Aβ positivity was associated with higher AV1451 uptake in the bilateral medial temporal regions independent of age and CSVD score (uncorrected P < .01) (Figure 1C). Higher CSVD score was associated with higher AV1451 uptake in the bilateral inferior temporal regions independent of age and Aβ positivity (Figure 1D). However, the significance did not survive when we applied FDR-corrected P < .05 for multiple comparison correction.
Amyloid-ß positivity was associated with higher AV1451 SUVR in several regions of interest including the entorhinal (β= 0.16; SE = 0.06; P = .01) and parahippocampal (β= 0.10; SE = 0.05; P = .03) regions after controlling for age and CSVD score. The CSVD score was associated with higher AV1451 SUVR in several regions of interest including the inferior temporal (β= 0.08; SE = 0.03; P = .01), entorhinal (β= 0.08; SE = 0.03; P = .017), parahippocampal (β= 0.05; SE = 0.02; P = .02), and fusiform (β= 0.06; SE = 0.03; P = .04) gyri after controlling for age and Aβ positivity.
Cortical Spreading Order of AV1451 in SVCI
In total SVCI, AV1451 accumulation was most frequently observed in the entorhinal cortex (59.1%) followed by the fusiform gyrus (42.7%), inferior temporal gyrus (42.6%), precuneus (36.3%), and parahippocampal gyrus (34.4%) (eTable 1 in the Supplement). Bootstrapping with 1000 resamples showed that the involvement frequencies of AV1451 in the entorhinal cortex, fusiform, inferior temporal gyri, and precuneus were significantly higher than that in the parahippocampal gyrus (Figure 2). When total patients with SVCI were stratified into Aβ-negative SVCI and Aβ-positive SVCI, each group of patients showed similar results with total SVCI (Figure 2; eTable 1 in the Supplement). Several typical SVCI cases prominently involving the inferior temporal regions while sparing parahippocampal region are shown in eFigure 3 in the Supplement.
Associations Between AV1451 SUVR and Cognition in SVCI
We evaluated the top 5 regions that showed earliest involvement of AV1451: entorhinal cortex, fusiform gyrus, inferior temporal gyrus, precuneus, and parahippocampal gyrus. In NC individuals, there were no significant correlations in any region of interest. However, in patients with SVCI, higher AV1451 uptake in all 5 regions was correlated with worse cognitive function in terms of language and general cognition. In addition, higher AV1451 uptake in the precuneus correlated with worse frontal function. The associations between regional AV1451 and other cognitive domains did not show significant results (Table 2). Subgroup analyses of Aβ-negative patients with SVCI and Aβ-positive patients with SVCI showed similar results (eTable 2 in the Supplement).
In patients with SVCI, Aβ positivity was associated with worse general cognitive function (β = −3.38; SE = 1.46; P = .03) and higher CSVD score was associated with worse frontal (β = −3.35; SE = 1.61; P = .04) and general cognitive function (β = −1.48; SE = 0.72; P = .05) (eTable 3 in the Supplement). To explore the complex associations between Aβ, CSVD, AV1451, and cognition, we built path models using factors that showed significant associations. We previously showed that Aβ was most associated with AV1451 uptake in the entorhinal cortex while CSVD was most associated with AV1451 uptake in the inferior temporal gyrus. Thus, we analyzed whether AV1451 uptake in the entorhinal cortex mediated the association between Aβ and general cognitive function and whether AV1451 uptake in the inferior temporal gyrus mediated the association between CSVD and general cognitive function. Path analyses showed that AV1451 uptake in the entorhinal cortex completely mediated the associations between Aβ and general cognitive function. In addition, AV1451 uptake in the inferior temporal gyrus completely mediated the association between CSVD score and general cognitive function. The path analyses showed goodness of fit to the data (Figure 3; eTable 4 in the Supplement).
We explored the distribution and clinical significance of tau burden using AV1451 in a well-characterized cohort of patients with SVCI and ADCI via comprehensive biomarker studies. The major findings of our study were as follows. First, CSVD burden did increase tau accumulation regardless of Aβ status. Second, in patients with SVCI AV1451 uptake mediated the associations of Aβ or CSVD score with cognitive impairment. Taken together, our findings suggest that both Aβ and CSVD are independently associated with increased tau accumulation, which is the final common pathway for cognitive impairment in patients with SVCI.
Our first conclusion that CSVD burden influenced tau accumulation regardless of Aβ is supported by several findings. First, compared with Aβ-negative NC individuals, Aβ-negative patients with SVCI had increased AV1451 uptake in the bilateral inferior temporal regions. This result strongly suggests that CSVD increases brain tau accumulation independent of Aβ negative status. Second, even after controlling for Aβ positivity, SVCI showed that CSVD score was associated with increased AV1451 uptake in the inferior temporal areas. Third, in patients with SVCI, the cortical spreading order of AV1451 showed that AV1451 involvement in the fusiform gyrus, inferior temporal, and precuneus regions preceded that in the parahippocampal region, which is different from the previously reported spreading order of NFT in typical patients with AD.5 However, the lack of increased AV1451 uptake in the hippocampus may reflect the technical issues of this ligand such as nonspecific retention of AV1451 in the choroidal plexus and partial volume effect due to hippocampal atrophy.6 Our findings are consistent with previous pathological studies showing that WMH is associated with the extension of NFT,19 although some studies suggested that CSVD does not promote NFT.20,21
There are several pieces of evidence that support Aβ independent pathway for brain phosphorylated tau accumulation. Studies of acute stroke identify transient increases in cerebrospinal fluid tau levels without concomitant increases in either amyloid-β42 levels.8 In addition, chronic cerebral hypoperfusion enhances tau hyperphosphorylation in rodents.22 One hypothesis suggests that ischemia alters the interactions between glycogen synthase kinase 3 and protein phosphatase 2A, leading to an Aβ independent phosphorylation of intracellular tau proteins.23 Additionally, it is possible that AV1451 might bind to molecules other than paired helical filament tau such as aging-related tau astrogliopathy, degenerating white matter, or degenerating neurons. Increased tau binding may occur as the result of a disrupted blood-brain barrier around ischemia.24 Finally, it is possible that any process that damages neurons, including ischemia, infarction, and other processes, may lead to increased tau accumulation.
We found that positive Aβ was associated with increased AV1451 uptake only in the medial temporal regions of patients with SVCI and did not extend to the isocortical regions. This finding suggests that some level of AD pathology exists in patients with SVCI. A previous study of preclinical AD showed higher uptakes of AV1451 in the medial and inferior temporal areas than in Aβ-negative cognitively normal individuals while preclinical AD showed less AV1451 uptake in the frontal and parietal regions than AD.25,26 Studies of dementia with Lewy bodies showed higher uptake of AV1451 compared with NC individuals but lower uptake of AV1451 compared with AD.27,28
Our final major finding was that in patients with SVCI, AV1451 uptake was correlated with cognitive impairment, and AV1451 mediated the associations of Aβ or CSVD score with cognition. This suggests that tau played a pivotal role because it was the final common pathway for cognitive impairment in patients with SVCI. According to the hypothetical model of temporal ordering of biomarkers in the AD spectrum, changes in Aβ lead to tau accumulation, which in turn results in structural changes and finally cognitive impairment.7 Our results also indicate that sequential biomarker changes in SVCI demonstrated that CSVD as well as Aβ led to tau accumulation, which in turn affected cognitive impairment. Interestingly, AV1451 inversely correlated with scores for language and general cognition, but not memory function, in patients with SVCI. The role of tau in patients with SVCI, therefore, might be different from that in the AD spectrum, in which tau is reported to be associated with memory.6,29 It may be possible that vascular pathologies and associated neuroinflammation modify neurotoxic processes triggered by tau. Further longitudinal studies are needed to investigate the complex associations between tau and structural or functional network changes.
There are several limitations in our study. First, we lacked pathological data to confirm AV1451-binding targets in patients with SVCI. Second, we had a relatively small sample size, and our results regarding the effects of Aβ or CSVD on AV1451 at uncorrected P less than .01 level did not survive after FDR correction. Future studies with more participants will be required to confirm these observations. Third, we could not consider the effects of other pathologies, including other AD-related changes (increases in soluble assemblies of Aβ and tau), microinfarcts, or possible combined degenerative dementia (dementia with lewy bodies and frontotemporal lobar degeneration) pathologies, which are also associated with cognitive impairments. Finally, our study population included patients with significant vascular burden, which may limit the generalizability of our data to other populations.
In SVCI, Aβ and CSVD are independently associated with increased tau accumulation in the medial temporal and inferior temporal regions, respectively. In addition, tau burden plays a pivotal role in the pathophysiology of cognitive impairment in SVCI. Our results add to the growing evidence that tau may be the final common pathway for injury and disease in the brain. The important role of tau underlying cognitive impairments in SVCI emphasizes the need for molecular imaging with the expectation that future treatment may target the disease-specific protein.
Corresponding Author: Sang Won Seo, MD, PhD, Department of Neurology, Sungkyunkwan University School of Medicine, Samsung Medical Center, 50 Ilwon-dong, Gangnam-gu, Seoul 135-710, Republic of Korea (sangwonseo@empal.com).
Accepted for Publication: February 23, 2018.
Correction: This article was corrected on June 25, 2018, to remove the Open Access designation.
Published Online: May 14, 2018. doi:10.1001/jamaneurol.2018.0975
Author Contributions: Dr Seo had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: H. Kim, Moon, Na, Seo.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: H. Kim, Park, Lee, Choi, Seo.
Critical revision of the manuscript for important intellectual content: H. Kim, Cho, Y. Jang, H. Jang, Y. Kim, K. Kim, Ryu, Moon, Weiner, Jags, Rabinovici, DeCarli, Lyoo, Na, Seo.
Statistical analysis: H. Kim, Park, Seo.
Obtained funding: H. Kim, Na, Seo.
Administrative, technical, or material support: H. Kim, Park, Cho, H. Jang, K. Kim, Ryu, Choi, Moon, Jags, DeCarli, Lyoo, Seo.
Supervision: Ryu, Weiner, Na, Seo.
Conflict of Interest Disclosures: None reported.
Funding/Support: This research was supported by National Research Foundation of Korea grants funded by the Korean government (NRF-2015R1C1A2A01053281 and NRF-2017R1A2B2005081) and the Brain Research Program through the National Research Foundation of Korea funded by the Ministry of Science, Information and Communication Technology, and Future Planning (2016M3C7A1913844).
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
2.Schneider
JA, Arvanitakis
Z, Bang
W, Bennett
DA. Mixed brain pathologies account for most dementia cases in community-dwelling older persons.
Neurology. 2007;69(24):2197-2204.
PubMedGoogle ScholarCrossref 3.Lee
JH, Kim
SH, Kim
GH,
et al. Identification of pure subcortical vascular dementia using 11C-Pittsburgh compound B.
Neurology. 2011;77(1):18-25.
PubMedGoogle ScholarCrossref 4.Xia
CF, Arteaga
J, Chen
G,
et al. [(18)F]T807, a novel tau positron emission tomography imaging agent for Alzheimer’s disease.
Alzheimers Dement. 2013;9(6):666-676.
PubMedGoogle ScholarCrossref 5.Cho
H, Choi
JY, Hwang
MS,
et al. In vivo cortical spreading pattern of tau and amyloid in the Alzheimer disease spectrum.
Ann Neurol. 2016;80(2):247-258.
PubMedGoogle ScholarCrossref 6.Johnson
KA, Schultz
A, Betensky
RA,
et al. Tau positron emission tomographic imaging in aging and early Alzheimer disease.
Ann Neurol. 2016;79(1):110-119.
PubMedGoogle ScholarCrossref 7.Jack
CR
Jr, Knopman
DS, Jagust
WJ,
et al. Tracking pathophysiological processes in Alzheimer’s disease: an updated hypothetical model of dynamic biomarkers.
Lancet Neurol. 2013;12(2):207-216.
PubMedGoogle ScholarCrossref 8.Hesse
C, Rosengren
L, Vanmechelen
E,
et al. Cerebrospinal fluid markers for Alzheimer’s disease evaluated after acute ischemic stroke.
J Alzheimers Dis. 2000;2(3-4):199-206.
PubMedGoogle ScholarCrossref 9.Qiu
L, Ng
G, Tan
EK, Liao
P, Kandiah
N, Zeng
L. Chronic cerebral hypoperfusion enhances Tau hyperphosphorylation and reduces autophagy in Alzheimer’s disease mice.
Sci Rep. 2016;6:23964.
PubMedGoogle ScholarCrossref 10.Kim
HJ, Yang
JJ, Kwon
H,
et al. Relative impact of amyloid-β, lacunes, and downstream imaging markers on cognitive trajectories.
Brain. 2016;139(pt 9):2516-2527.
PubMedGoogle ScholarCrossref 11.Villeneuve
S, Reed
BR, Wirth
M,
et al. Cortical thickness mediates the effect of β-amyloid on episodic memory.
Neurology. 2014;82(9):761-767.
PubMedGoogle ScholarCrossref 12.Villeneuve
S, Reed
BR, Madison
CM,
et al. Vascular risk and Aβ interact to reduce cortical thickness in AD vulnerable brain regions.
Neurology. 2014;83(1):40-47.
PubMedGoogle ScholarCrossref 13.McKhann
GM, Knopman
DS, Chertkow
H,
et al. The diagnosis of dementia due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease.
Alzheimers Dement. 2011;7(3):263-269.
PubMedGoogle ScholarCrossref 14.Kang
Y, Na
DL. Seoul Neuropsychological Screening Battery (SNSB). Incheon, South Korea: Human Brain Research & Consulting; 2003.
15.Ahn
HJ, Chin
J, Park
A,
et al. Seoul Neuropsychological Screening Battery-dementia version (SNSB-D): a useful tool for assessing and monitoring cognitive impairments in dementia patients.
J Korean Med Sci. 2010;25(7):1071-1076.
PubMedGoogle ScholarCrossref 16.Staals
J, Makin
SD, Doubal
FN, Dennis
MS, Wardlaw
JM. Stroke subtype, vascular risk factors, and total MRI brain small-vessel disease burden.
Neurology. 2014;83(14):1228-1234.
PubMedGoogle ScholarCrossref 17.Wardlaw
JM, Smith
EE, Biessels
GJ,
et al; Standards for Reporting Vascular Changes on Neuroimaging (STRIVE v1). Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration.
Lancet Neurol. 2013;12(8):822-838.
PubMedGoogle ScholarCrossref 18.Barthel
H, Gertz
HJ, Dresel
S,
et al; Florbetaben Study Group. Cerebral amyloid-β PET with florbetaben (18F) in patients with Alzheimer’s disease and healthy controls: a multicentre phase 2 diagnostic study.
Lancet Neurol. 2011;10(5):424-435.
PubMedGoogle ScholarCrossref 19.Polvikoski
TM, van Straaten
EC, Barkhof
F,
et al. Frontal lobe white matter hyperintensities and neurofibrillary pathology in the oldest old.
Neurology. 2010;75(23):2071-2078.
PubMedGoogle ScholarCrossref 20.Thal
DR, Attems
J, Ewers
M. Spreading of amyloid, tau, and microvascular pathology in Alzheimer’s disease: findings from neuropathological and neuroimaging studies.
J Alzheimers Dis. 2014;42(suppl 4):S421-S429.
PubMedGoogle ScholarCrossref 21.Esiri
MM, Joachim
C, Sloan
C,
et al. Cerebral subcortical small vessel disease in subjects with pathologically confirmed Alzheimer disease: a clinicopathologic study in the Oxford Project to Investigate Memory and Ageing (OPTIMA).
Alzheimer Dis Assoc Disord. 2014;28(1):30-35.
PubMedGoogle ScholarCrossref 22.Sun
J, Wang
H, Chen
LN,
et al. Remarkable impairment of Wnt/β-catenin signaling in the brains of the mice infected with scrapie agents.
J Neurochem. 2015;136(4):731-740.
PubMedGoogle ScholarCrossref 23.Song
B, Ao
Q, Wang
Z,
et al. Phosphorylation of tau protein over time in rats subjected to transient brain ischemia.
Neural Regen Res. 2013;8(34):3173-3182.
PubMedGoogle Scholar 24.Hartz
AM, Bauer
B, Soldner
EL,
et al. Amyloid-β contributes to blood-brain barrier leakage in transgenic human amyloid precursor protein mice and in humans with cerebral amyloid angiopathy.
Stroke. 2012;43(2):514-523.
PubMedGoogle ScholarCrossref 25.Wang
L, Benzinger
TL, Su
Y,
et al. Evaluation of tau imaging in staging Alzheimer disease and revealing interactions between β-amyloid and tauopathy.
JAMA Neurol. 2016;73(9):1070-1077.
PubMedGoogle ScholarCrossref 26.Tosun
D, Landau
S, Aisen
PS,
et al; Alzheimer’s Disease Neuroimaging Initiative. Association between tau deposition and antecedent amyloid-β accumulation rates in normal and early symptomatic individuals.
Brain. 2017;140(5):1499-1512.
PubMedGoogle ScholarCrossref 27.Gomperts
SN, Locascio
JJ, Makaretz
SJ,
et al. Tau positron emission tomographic imaging in the lewy body diseases.
JAMA Neurol. 2016;73(11):1334-1341.
PubMedGoogle ScholarCrossref 28.Kantarci
K, Lowe
VJ, Boeve
BF,
et al. AV-1451 tau and β-amyloid positron emission tomography imaging in dementia with Lewy bodies.
Ann Neurol. 2017;81(1):58-67.
PubMedGoogle ScholarCrossref 29.Schöll
M, Lockhart
SN, Schonhaut
DR,
et al. PET imaging of tau deposition in the aging human brain.
Neuron. 2016;89(5):971-982.
PubMedGoogle ScholarCrossref