Neuroimaging biomarkers that distinguish between changes due to typical brain ageing and Alzheimer’s disease are valuable for determining how much each contributes to cognitive decline. Supervised machine learning models can derive multivariate patterns of brain change related to the two processes, including the Spatial Patterns of Atrophy for Recognition of Alzheimer’s Disease (SPARE-AD) and of Brain Aging (SPARE-BA) scores investigated herein. However, the substantial overlap between brain regions affected in the two processes confounds measuring them independently. We present a methodology, and associated results, towards disentangling the two.T1-weighted MRI scans of 4054 participants (48–95 years) with Alzheimer’s disease, mild cognitive impairment (MCI), or cognitively normal (CN) diagnoses from the Imaging-based coordinate SysTem for AGIng and NeurodeGenerative diseases (iSTAGING) consortium were analysed. Multiple sets of SPARE scores were investigated, in order to probe imaging signatures of certain clinically or molecularly defined sub-cohorts. First, a subset of clinical Alzheimer’s disease patients (n = 718) and age- and sex-matched CN adults (n = 718) were selected based purely on clinical diagnoses to train SPARE-BA1 (regression of age using CN individuals) and SPARE-AD1 (classification of CN versus Alzheimer’s disease) models. Second, analogous groups were selected based on clinical and molecular markers to train SPARE-BA2 and SPARE-AD2 models: amyloid-positive Alzheimer’s disease continuum group (n = 718; consisting of amyloid-positive Alzheimer’s disease, amyloid-positive MCI, amyloid- and tau-positive CN individuals) and amyloid-negative CN group (n = 718). Finally, the combined group of the Alzheimer’s disease continuum and amyloid-negative CN individuals was used to train SPARE-BA3 model, with the intention to estimate brain age regardless of Alzheimer’s disease-related brain changes.The disentangled SPARE models, SPARE-AD2 and SPARE-BA3, derived brain patterns that were more specific to the two types of brain changes. The correlation between the SPARE-BA Gap (SPARE-BA minus chronological age) and SPARE-AD was significantly reduced after the decoupling (r = 0.56–0.06). The correlation of disentangled SPARE-AD was non-inferior to amyloid- and tau-related measurements and to the number of APOE ε4 alleles but was lower to Alzheimer’s disease-related psychometric test scores, suggesting the contribution of advanced brain ageing to the latter. The disentangled SPARE-BA was consistently less correlated with Alzheimer’s disease-related clinical, molecular and genetic variables.By employing conservative molecular diagnoses and introducing Alzheimer’s disease continuum cases to the SPARE-BA model training, we achieved more dissociable neuroanatomical biomarkers of typical brain ageing and Alzheimer’s disease.

Differential effects of white matter hyperintensities and regional amyloid deposition on regional cortical thickness

Ethnic differences in the frequency of β-amyloid deposition in cognitively normal individuals

A neuroinflammatory reaction in Alzheimer disease (AD) brains involves reactive astrocytes that overexpress monoamine oxidase-B (MAO-B). 18F-(S)-(2-methylpyrid-5-yl)-6-[(3-fluoro-2-hydroxy)propoxy]quinoline (18F-SMBT-1) is a novel 18F PET tracer highly selective for MAO-B. We characterized the clinical performance of 18F-SMBT-1 PET across the AD continuum as a potential surrogate marker of reactive astrogliosis. Methods: We assessed 18F-SMBT-1 PET regional binding in 77 volunteers (76 ± 5.5 y old; 41 women, 36 men) across the AD continuum: 57 who were cognitively normal (CN) (44 amyloid-β [Aβ]-negative [Aβ-] and 13 Aβ-positive [Aβ+]), 12 who had mild cognitive impairment (9 Aβ- and 3 Aβ+), and 8 who had AD dementia (6 Aβ+ and 2 Aβ-). All participants also underwent Aβ and tau PET imaging, 3-T MRI, and neuropsychologic evaluation. Tau imaging results were expressed in SUV ratios using the cerebellar cortex as a reference region, whereas Aβ burden was expressed in centiloids. 18F-SMBT-1 outcomes were expressed as SUV ratio using the subcortical white matter as a reference region. Results: 18F-SMBT-1 yielded high-contrast images at steady state (60-80 min after injection). When compared with the Aβ- CN group, there were no significant differences in 18F-SMBT-1 binding in the group with Aβ- mild cognitive impairment. Conversely, 18F-SMBT-1 binding was significantly higher in several cortical regions in the Aβ+ AD group but also was significantly lower in the mesial temporal lobe and basal ganglia. Most importantly, 18F-SMBT-1 binding was significantly higher in the same regions in the Aβ+ CN group as in the Aβ- CN group. When all clinical groups were considered together, 18F-SMBT-1 correlated strongly with Aβ burden and much less with tau burden. Although in most cortical regions 18F-SMBT-1 did not correlate with brain volumetrics, regions known for high MAO-B concentrations presented a direct association with hippocampal and gray matter volumes, whereas the occipital lobe was directly associated with white matter hyperintensity. 18F-SMBT-1 binding was inversely correlated with Mini Mental State Examination and the Australian Imaging Biomarkers and Lifestyle's Preclinical Alzheimer Cognitive Composite in some neocortical regions such as the frontal cortex, lateral temporal lobe, and supramarginal gyrus. Conclusion: Cross-sectional human PET studies with 18F-SMBT-1 showed that Aβ+ AD patients, but most importantly, Aβ+ CN individuals, had significantly higher regional 18F-SMBT-1 binding than Aβ- CN individuals. Moreover, in several regions in the brain, 18F-SMBT-1 retention was highly associated with Aβ load. These findings suggest that increased 18F-SMBT-1 binding is detectable at the preclinical stages of Aβ accumulation, providing strong support for its use as a surrogate marker of astrogliosis in the AD continuum.

Background: Evaluating the risk of Alzheimer’s disease (AD) in cognitively normal (CN) and patients with mild cognitive impairment (MCI) is extremely important. While MCI-to-AD progression risk has been studied extensively, few studies estimate CN-to-MCI conversion risk. The Cox proportional hazards (PH), a widely used survival analysis model, assumes a linear predictor-risk relationship. Generalizing the PH model to more complex predictor-risk relationships may increase risk estimation accuracy. Objective: The aim of this study was to develop a PH model using an Xgboost regressor, based on demographic, genetic, neuropsychiatric, and neuroimaging predictors to estimate risk of AD in patients with MCI, and the risk of MCI in CN subjects. Methods: We replaced the Cox PH linear model with an Xgboost regressor to capture complex interactions between predictors, and non-linear predictor-risk associations. We endeavored to limit model inputs to noninvasive and more widely available predictors in order to facilitate future applicability in a wider setting. Results: In MCI-to-AD (n = 882), the Xgboost model achieved a concordance index (C-index) of 84.5%. When the model was used for MCI risk prediction in CN (n = 100) individuals, the C-index was 73.3%. In both applications, the C-index was statistically significantly higher in the Xgboost in comparison to the Cox PH model. Conclusion: Using non-linear regressors such as Xgboost improves AD dementia risk assessment in CN and MCI. It is possible to achieve reasonable risk stratification using predictors that are relatively low-cost in terms of time, invasiveness, and availability. Future strategies for improving AD dementia risk estimation are discussed.

Alzheimer's disease (AD) is conceptualized as a biological continuum encompassing the preclinical (clinically asymptomatic but with evidence of AD pathology) and clinical (symptomatic) phases. Using 18F-THK5351 as a tracer that binds to both tau and monoamine oxidase B (MAO-B), we investigated the changes in 18F-THK5351 accumulation patterns in AD continuum individuals with positive amyloid PET consisting of cognitively normal individuals (CNp), amnestic mild cognitive impairment (aMCI), and AD and cognitively normal individuals (CNn) with negative amyloid PET. We studied 69 individuals (32 CNn, 11 CNp, 9 aMCI, and 17 AD) with structural magnetic resonance imaging, 11C-Pittsburgh compound-B (PIB) and 18F-THK5351 PET, and neuropsychological assessment. 18F-THK5351 accumulation was evaluated with visual analysis, voxel-based analysis and combined region of interest (ROI)-based analysis corresponding to Braak neurofibrillary tangle stage. On visual analysis, 18F-THK5351 accumulation was increased with stage progression in the AD continuum. On voxel-based analysis, there was no statistical difference in 18F-THK5351 accumulation between CNp and CNn. However, a slight increase of the bilateral posterior cingulate gyrus in aMCI and definite increase of the bilateral parietal temporal association area and posterior cingulate gyrus/precuneus in AD were detected compared with CNn. On ROI-based analyses, 18F-THK5351 accumulation correlated positively with supratentorial 11C-PIB accumulation and negatively with the hippocampal volume and neuropsychological assessment. The AD continuum showed an increase in 18F-THK5351 with stage progression, suggesting that 18F-THK5351 has the potential to visualize the severity of tau deposition and neurodegeneration in accordance with the AD continuum.

The deposition of beta-amyloid (Aβ) in the brain precedes the onset of symptoms such as cognitive impairment in Alzheimer’s disease (AD); therefore, the early detection of Aβ accumulation is crucial. We previously reported the applicability of the QPLEXTM Alz plus assay kit for the prescreening of Aβ accumulation. Here, we tested the specific application of the kit in a large cohort of cognitively normal (CN) individuals of varying ages for the early detection of Aβ accumulation. We included a total of 221 CN participants with or without brain Aβ. The QPLEXTM biomarkers were characterized based on age groups (1st–3rd tertile) and across various brain regions with cerebral amyloid deposition. The 3rd tertile group (>65 years) was found to be the most suitable age group for the application of our assay kit. Receiver operating characteristic curve analysis showed that the area under the curve (AUC, discrimination power) was 0.878 with 69.7% sensitivity and 98.4% specificity in the 3rd tertile group. Additionally, specific correlations between biomarkers and cerebral amyloid deposition in four different brain regions revealed an overall correlation with general amyloid deposition, consistent with previous findings. Furthermore, the combinational panel with plasma Aβ1–42 levels maximized the discrimination efficiency and achieved an AUC of 0.921 with 95.7% sensitivity and 67.3% specificity. Thus, we suggest that the QPLEXTM Alz plus assay is useful for prescreening brain Aβ levels in CN individuals, especially those aged >65 years, to prevent disease progression via the early detection of disease initiation.

Automated detection, selection and classification of hippocampal landmark points for the diagnosis of Alzheimer’s disease

Abstract BackgroundAlzheimer's disease (AD) is hypothesised as a disconnection syndrome where degenerating white matter fibre bundles leads to deterioration in the integration and communication between brain regions. Connectomics allows the study of in vivo brain connectivity and elucidates how the disease changes the brain network. Although some studies have shown evidence of alteration of structural connectivity between AD and cognitively normal individuals (CN), a large proportion of research focused on functional connectomics in AD. Emerging connectomics studies explored the use of machine learning (ML) to distinguish brain structural connectivity differences in AD with promising results. This project aims to identify changes in the structural connectome using novel image processing techniques to generate network metrics and utilise ML to classify between AD and CN.MethodWe examined data from 143 age‐matched subjects (AD mean: 71.1 ± 2.79 and CN mean: 71.09 ± 2.72) from the Alzheimer's Disease Neuroimaging Initiative cohort 2 (ADNI2). We used magnetic resonance images (T1‐weighted and diffusion‐weighted images) combined with the latest state‐of‐the‐art imaging processing tools to generate structural connectomes. Relevant network metrics were used to measure and compare brain connectivity, while ML algorithms were used to distinguish network metrics between AD and CN.ResultWe found significant connectivity changes in clustering coefficient (p < 0.05), normalised degree variance (p < 0.0001), hierarchical complexity (p < 0.005) and rich club (p < 0.0001) in AD (table 1). We also established and compared classification performances within our ML model. Random forest yielded sensitivity of 53.06% and specificity of 82.98% (table 2) for imbalanced data (AD=49, CN=94). On balanced data (AD=CN=49), the model was 81.63% specific and 69.39% sensitive (table 3) in detecting AD.ConclusionThe results show the feasibility of a connectome analysis of structural imaging combining the latest network metrics with ML for AD detection. While previous ML studies achieved promising results with balanced data, we reported both balanced and imbalanced models. As real‐world AD data are more likely to be imbalanced, the lower performance of the ML models on imbalanced data suggests that further improvement is needed for clinical implementation.

MBI symptoms predict progression to Alzheimer's disease independent of neuropathology.

Investigating the risk of cardiovascular risk factor subgroups in cognitively normal elderly on progression to AD: A latent class approach

Increased regional white matter hyperintensity volume in objectively‐defined subtle cognitive decline and mild cognitive impairment

BackgroundHepatocyte growth factor (HGF) plays a role in neuronal survival and development, and has been implicated in neurodegenerative diseases. We sought to examine the associations of the CSF HGF with Alzheimer’s disease (AD) pathology and cognitive function.MethodsA total of 238 participants (including 90 cognitively normal (CN) and 148 mild cognitive impairment (MCI)) who had measurements of CSF HGF were included from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. Multiple linear regression models were utilized to explore the cross-sectional associations of CSF HGF with AD biomarkers (including Aβ42, pTau, and tTau proteins) in non-demented participants. Moreover, linear mixed-effects regression models were utilized to explore the longitudinal associations of HGF subgroups with cognitive function. Mediation analyses were utilized to explore the mediation effects of AD markers.ResultsMCI individuals had significantly increased CSF HGF compared with the CN individuals. Results of multiple linear regressions showed significant correlations of CSF HGF with CSF Aβ42, pTau, and tTau in non-demented participants. Higher level of baseline CSF HGF was associated with faster cognitive decline. Influences of the baseline CSF HGF on cognition were partially mediated by Aβ42, pTau, and tTau pathologies.ConclusionsHigh concentrations of HGF in CSF may be related to faster cognitive decline. The cognitive consequences of higher CSF HGF partly stem from AD pathology, which suggests that the CSF HGF may be an attractive biomarker candidate to track AD progression.

Retinal biomarkers reflecting in vivo brain Alzheimer disease (AD) pathologic abnormalities could be a useful tool for screening cognitively normal (CN) individuals at the preclinical stage of AD. To investigate the association of both functional and structural alterations of the retina with in vivo AD pathologic abnormalities in CN older adults and model a screening tool for detection of preclinical AD. This cross-sectional study included a total of 49 CN individuals, and all assessment was done at the Seoul National University Hospital, Seoul, South Korea. All participants underwent complete ophthalmic examination, including swept-source optical coherence tomography (SS-OCT) and multifocal electroretinogram as well as amyloid-β (Aβ) positron emission tomography and magnetic resonance imaging. Data were collected from January 1, 2016, through October 31, 2017, and analyzed from February 1, 2018, through June 30, 2020. For structural parameters of the retina, the thickness of the macula and layer-specific thicknesses, including peripapillary retinal nerve fiber layer and ganglion cell-inner plexiform layer measured by SS-OCT, were used for analysis. For functional parameters of the retina, implicit time and amplitude of rings 1 to 6 measured by multifocal electroretinogram were used. Of the 49 participants, 25 were women (51.0%); mean (SD) age was 70.6 (9.4) years. Compared with 33 CN individuals without Aβ deposition (Aβ-CN), the 16 participants with Aβ (Aβ+CN) showed reduced inner nasal macular thickness (mean [SD], 308.9 [18.4] vs 286.1 [22.5] μm; P = .007) and retinal nerve fiber layer thickness, particularly in the inferior quadrant (133.8 [17.9] vs 103.8 [43.5] μm; P = .003). In addition, the Aβ+CN group showed prolonged implicit time compared with the Aβ-CN group, particularly in ring 5 (41.3 [4.0] vs 38.2 [1.3] milliseconds; P = .002). AD-related neurodegeneration was correlated with the thickness of the ganglion cell-inner plexiform layer only (r = 0.41, P = .005). The model to differentiate the Aβ+CN vs Aβ-CN groups derived from the results showed 90% accuracy. The findings of this study showing both functional as well as structural changes of retina measured by multifocal electroretinogram and SS-OCT in preclinical AD suggest the potential use of retinal biomarkers as a tool for early detection of in vivo AD pathologic abnormalities in CN older adults.

Olfactory impairment is evident in Alzheimer's disease (AD); however, its precise relationships with clinical biomarker measures of tau pathology and neuroinflammation are not well understood. To determine if odor identification performance measured with the University of Pennsylvania Smell Identification Test (UPSIT) is related to in vivo measures of tau pathology and neuroinflammation. Cognitively normal and cognitively impaired participants were selected from an established research cohort of adults aged 50 and older who underwent neuropsychological testing, brain MRI, and amyloid PET. Fifty-four participants were administered the UPSIT. Forty-one underwent 18F-MK-6240 PET (measuring tau pathology) and fifty-three underwent 11C-PBR28 PET (measuring TSPO, present in activated microglia). Twenty-three participants had lumbar puncture to measure CSF concentrations of total tau (t-tau), phosphorylated tau (p-tau), and amyloid-β (Aβ42). Low UPSIT performance was associated with greater18F-MK-6240 binding in medial temporal cortex, hippocampus, middle/inferior temporal gyri, inferior parietal cortex, and posterior cingulate cortex (p < 0.05). Similar relationships were seen for 11C-PBR28. These relationships were primarily driven by amyloid-positive participants. Lower UPSIT performance was associated with greater CSF concentrations of t-tau and p-tau (p < 0.05). Amyloid status and cognitive status exhibited independent effects on UPSIT performance (p < 0.01). Olfactory identification deficits are related to extent of tau pathology and neuroinflammation, particularly in those with amyloid pathophysiology. The independent association of amyloid-positivity and cognitive impairment with odor identification suggests that low UPSIT performance may be a marker for AD pathophysiology in cognitive normal individuals, although impaired odor identification is associated with both AD and non-AD related neurodegeneration.NCT Registration Numbers: NCT03373604; NCT02831283.

Centiloid scale analysis for 18F-THK5351 PET imaging in Alzheimer's disease.

Brainstem volumetric integrity in preclinical and prodromal Alzheimer’s disease

Effect of decreased visual acuity on cortical thickness: A possible risk factor for neurodegeneration

AHEAD 3‐45 study design: A global study to evaluate the efficacy and safety of treatment with BAN2401 for 216 weeks in preclinical Alzheimer’s disease with intermediate amyloid (A3 trial) and elevated amyloid (A45 trial)

Integration of demographics, genetics, imaging and metabolomics data to identify Alzheimer’s disease patients