AbstractBackgroundThe immune system is substantially involved in the development and progression of age‐related cognitive decline and Alzheimer’s disease (AD).MethodAs genetic and environmental factors interactively impact these conditions, we investigated how risk factors such as APOE genotype, age, and sex influence immune activation markers and AD biomarkers in cerebrospinal fluid (CSF) in elderly individuals enrolled in the Mayo Clinic Study of Aging cohort. Among cognitively unimpaired individuals aged over 65 at the baseline visit (N=298), we measured 365 CSF immune activation markers using the proximity extension assay.ResultWe found that age, sex, and diabetes status are associated with altered CSF levels of immune activation markers independently of other factors. For CSF AD biomarkers, we observed significant positive correlations between age and total tau, phosphorylated tau‐181 (p‐tau181), neurofilament light (NfL), and YKL40. APOE4 was also associated with lower Aβ42 and higher SNAP25 in CSF. We further examined whether baseline visit variables can predict cognitive decline, represented by the conversion from CDR=0 to CDR>0. We found that age, Aβ42, NfL, and REG4 were independently correlated with CDR conversion risk. When the cohort was dichotomized by their median values, older participants with lower Aβ42, higher NfL, and higher REG4 at baseline developed cognitive impairment during the follow up with a c‐index of 0.762 while age alone had a c‐index of 0.699.ConclusionTogether, our results suggest that assessing CSF immune activation markers and AD biomarkers can improve the prediction of cognitive impairment risk in the elderly.
Biological Factors Influencing [18F]MK6240 and [18F]FTP Correlation in Target Regions
AbstractBackgroundThe association between [18F]Flortaucipir (FTP) and [18F]MK6240, two commonly used tau‐PET tracers in Alzheimer’s disease (AD), varies due to distinct binding properties and off‐target signal regions. Our study aims to elucidate the biological factors influencing this association and evaluate the applicability of a common equation across different on‐target regions.Method113 individuals from the HEAD dataset (11 young, 58 cognitively unimpaired elderly, and 44 cognitively impaired) underwent [18F]MK6240, [18F]FTP and Aβ‐PET scans. Images were processed using the inferior cerebellar grey (CG) as the reference region, with subjects categorized by Aβ status. Stepwise linear fitting identified off‐target (OFF) regions impacting tracer associations within the whole cortex, Braak and temporal‐meta regions of interest (ROIs). OFF regions included were a composite mask (pallidum, putamen, caudate ‐ PPC) to address collinearity in subcortical nuclei, as well as the region with the greatest partial volume effects (PVE) on the ROI (meninges, choroid plexus, and the hotspot medial to entorhinal cortices (EC)). Akaike information criteria (AIC) analyses tested linear models with various off‐target regions.ResultPPC and PVE regions influenced tracer association in on‐target regions. PPC binding is associated with age. PPC and meninges retention, along with Aβ status interaction, demonstrated the best model fit for the whole cortex (Figure 1). Weighting factors derived from the whole cortex model were applied to Braak regions and temporal meta‐ROI. Braak III‐VI and temporal‐meta ROI showed optimal models including OFF regions and Aβ status interaction (Figure 2). Forcing the whole equation gave optimal results. For Braak I (EC), adding the OFF covariates only was the best model. For Braak II (hippocampus), the best model included OFF regions and the interaction with Aβ status (Figure 2). Forcing the whole cortex equation onto Braak I and Braak II yielded suboptimal fits.ConclusionHarmonizing tau‐PET tracers involves considering significant covariates. Integrating meningeal and age‐related off‐target regions enhances [18F]MK6240 and [18F]FTP association in on‐target regions, further improved by Aβ status inclusion. While the whole cortex model may not suit early ROI, it proves effective for Braak III and beyond. Creating an equation for tracer translation necessitates incorporating meningeal and subnuclei off‐target retention.
Comparison of tau‐PET tracers for in vivo Braak staging: the HEAD Study
AbstractBackgroundTau‐PET tracers allow for in vivo Braak staging of individuals in the Alzheimer’s disease (AD) continuum. The impact of tracers’ characteristics for Braak staging using tau‐PET remains unclear. Therefore, we performed a head‐to‐head comparison of Braak staging using first‐ and second‐generation tau‐PET tracers.MethodWe assessed 51 cognitively unimpaired (CU) and 49 cognitively impaired participants (mean [SD] age 69.4 [7.5] years) with at least two tau‐PET ligands ([18F]MK6240, [18F]AV1451, and/or [18F]RO948) at McGill University, as part of the HEAD study. We calculated standardized uptake value ratios (SUVR) in Braak‐like regions of interest (ROI) for each ligand and investigated their association using Spearman’s correlation. Thresholds defined the presence of tauopathy in each Braak ROI, which was used to assign a Braak stage to each participant. Finally, we assessed the agreement between the Braak staging provided by each tracer, both for the traditional (using separate Braak stages – 0, I, II, III, IV, V, and VI) and simplified (using joint Braak stages – 0, I‐II, III‐IV, and V‐VI) frameworks.ResultIn all Braak ROIs, we found positive correlations between the SUVR values of the three tracers (Figure 1). The strongest correlation was between [18F]MK6240 and [18F]RO948 in Braak II (r=0.94), and the weakest between [18F]AV1451 and [18F]RO948 in VI (r=0.51). Inter‐tracer agreement regarding the latest stage of abnormality was substantial or nearly perfect between pairs of tracers but moderate between the three tracers (Figure 2A). The simplified framework presented greater agreement compared to traditional Braak staging, except between [18F]MK6240 and [18F]RO948. At the ROI level, the lowest agreements were observed for Braak II and I‐II, when [18F]MK6240 and [18F]RO948 were compared to [18F]AV1451 (Figure 2B‐C). [18F]AV1451 had the highest probability of overestimating the Braak stage assigned by [18F]MK6240 and by [18F]RO948, especially at earlier stages (Figure 3).ConclusionThe ligands presented moderate to nearly perfect agreement for in vivo staging of AD severity. Off‐target binding of [18F]AV1451 to the choroid plexus might explain disagreements at early stages. Overall, slight improvements in agreements were achieved with the simplified framework, as observed in histopathological staging.
Evaluating DTI‐ALPS, an imaging surrogate for glymphatic clearance, in a large population‐based sample
Abstract Background Diffusion tensor imaging along perivascular spaces index (DTI‐ALPS), which measures diffusivity increases in the perivascular spaces along the medullary veins, is being increasingly utilized as a surrogate marker of glymphatic clearance (Taoka et. al. Jpn J Radio 2017). Lower DTI‐ALPS index means higher measured diffusivity from the perivascular space, and so it is expected to decrease with age. However, correlation with aging and dementia markers have not been systematically evaluated. The goal of this work was to automatically estimate DTI‐ALPS in a large sample and evaluate the variance explained by various demographic and imaging markers. Methods We identified 2739 participants aged 30+ years (mean 69.6, 48% female, 88% CU) in the population‐based Mayo Clinic Study of Aging with diffusion MRI. We estimated DTI‐ALPS through automated placement of regions as published previously (Liu et al. AAIC 2023) and shown in Figure 1. We examined the associations of DTI‐ALPS with demographics (age, sex), vascular risk (cardiovascular and metabolic condition score), clinical data (diagnosis, global cognition) and imaging markers (white matter hyperintensity (WMH), global amyloid load from PIB‐PET, and the temporal meta‐ROI Tau‐PET SUVR) using partial Pearson correlations and ANCOVA with adjustments for age and sex. Results DTI‐ALPS was negatively correlated with age, higher in females, declined with vascular risk, and lower in MCI/dementia groups (Figure 2). Higher WMH and higher amyloid were associated with lower DTI‐ALPS whereas Tau‐PET SUVR was not associated with DTI‐ALPS. Conclusions DTI‐ALPS can be automatically computed in large samples. Vascular risk and WMH had stronger associations with the computed DTI‐ALPS compared to AD biomarkers (amyloid and tau) suggesting that the surrogate was more likely linked to vascular injury.
Connectivity as a universal predictor of tau spreading in typical and atypical Alzheimer’s disease
AbstractBackgroundThere is a strong link between tau and progression of Alzheimer’s disease (AD), necessitating an understanding of tau spreading mechanisms. Prior research, predominantly in typical AD, suggested that tau propagates from epicenters (regions with earliest tau) to functionally connected regions. However, given the constrained spatial heterogeneity of tau in typical AD, validating this connectivity‐based tau spreading model in AD variants with distinct tau deposition patterns is crucial.MethodWe included 269 amyloid‐β‐positive (PET/CSF) individuals with clinically diagnosed atypical AD (113 posterior cortical atrophy, PCA‐AD; 83 logopenic variant primary progressive aphasia, lvPPA‐AD; 33 behavioural variant AD, bvAD; 40 corticobasal syndrome, CBS‐AD) and 68 with typical AD from 12 international cohorts, who underwent tau‐PET (54% [18F]AV1451/[18F]flortaucipir/Tauvid, 27% [18F]MK6240, 19% [18F]PI2620). Using Gaussian mixture modeling including amyloid‐β‐negative controls, cross‐sectional tau‐PET standardized uptake value ratios within Schaefer‐200 atlas regions were transformed to tau positivity probabilities. Tau epicenters were defined as the 5% regions with highest tau positivity probabilities. For each variant, the association between functional connectivity‐based distance (using the 30% strongest positive region‐to‐region connections of a group‐average connectivity matrix from ADNI elderly controls) and tau‐PET covariance (group‐average correlation per region pair) was assessed through linear regression, adjusting for age, sex, site, and Euclidean distance. Regions were categorized based on functional proximity to the epicenter (quartiles 1‐4) and tau positivity probabilities were assessed accordingly.ResultTau positivity probabilities matched clinical variants, with a posterior pattern in PCA‐AD, left‐hemispheric dominant pattern in lvPPA‐AD, widespread pattern in bvAD, sensorimotor cortex involvement in CBS‐AD, and temporo‐parietal predominance in typical AD (Figure 1). In line with this, tau epicenters were highly heterogeneous across variants (Figure 1). In all variants, greater tau‐PET covariance was associated with shorter functional connectivity‐based distance (Figure 2). We observed that regions in closer functional proximity to the epicenter exhibited higher tau positivity probabilities than regions functionally further away (p<0.05, Figure 3).ConclusionThis multi‐center study shows that the brain’s functional architecture serves as a universal predictor of tau spreading in AD. Since tau is a key driver of neurodegeneration and cognitive decline in AD, this finding holds potential for personalized medicine and defining participant‐specific endpoints in clinical trials.
Variational autoencoder latent space as a robust and pragmatic clinical classification tool for dementia
AbstractBackgroundMany proposed clinical decision support systems (CDSS) require multiple disparate data elements as input, which makes implementation difficult, and furthermore have a black‐box nature leading to low interpretability. Fluorodeoxyglucose Positron Emission Tomography (FDG‐PET) is an established modality for the diagnosis of dementia, and a CDSS that uses only an FDG‐PET image to produce a reliable and understandable result would ease both of these challenges to clinical application.MethodA deep variational autoencoder (VAE) was used to extract a latent representation of each image through prior training from FDG‐PET brain images (n=2000). This unsupervised VAE has a novel graph convolutional architecture that makes it applicable to masked template space images. A logistic regression model was used to classify each image. A parametric study of the latent space revealed the imaging features that were used in the logistic regression model to differentiate each class. A separate dataset of participants labeled with their clinical diagnosis (n=1239) was used to assess the logistic regression model’s diagnostic accuracy of cognitively unimpaired (CU) (n=679) and between common dementia subtypes: Alzheimer’s disease (AD) (n=310), Lewy body dementia (DLB) (n=151), and behavioral variant frontotemporal dementia (bvFTD) (n=99). The logistic regression classifier was evaluated using receiver‐operator characteristic (ROC‐AUC) curves.ResultROC‐AUC curves for CU and each dementia subtype are illustrated in Figure 1a. A k nearest neighbors model was used to develop a graphical representation of the latent space where nodes are images and edges are drawn between nearest neighbors, illustrated in Figure 1b. The graph of the latent space demonstrates a comparison between dementia subtypes and CU images.ConclusionIn this study, we developed a proof‐of‐concept VAE that can achieve differential diagnosis with 84.8% balanced accuracy. The model architecture provides visual interpretability, and shows that the hippocampi, parietal lobes, and cerebellum, were useful for distinguishing AD, the frontal lobe, posterior cingulate cortex, and hippocampi were useful for distinguishing DLB, and prefrontal pole, subcortical structures, and occipital lobe were useful for distinguishing bvFTD. These regions of metabolism align with clinical pathology and metabolic patterns.
AbstractBackgroundPhase four of the Alzheimer’s Disease Neuroimaging Initiative (ADNI4) began in 2023. This time‐period corresponded to MRI vendors introducing product sequences with compressed sensing (CS), cross‐vendor adoption of arterial spin‐labelling (ASL) and multi‐band slice excitation, and hardware improvements (head‐coils, increased gradient amplitudes). These advances enabled the acquisition of new imaging measures and reduced scan times. The ADNI4 MRI protocol aims to maintain longitudinal consistency across two decades of data acquisition, while adopting new technologies. Here we describe the design and justification behind the ADNI4 protocol and list the target imaging biomarkers.MethodsThe ADNI4 MRI protocol includes 9 MRI sequences; i) T1‐weighted (T1w) MP‐RAGE, ii) a compressed sensing T1w MP‐RAGE, iii) 3D T2w FLAIR, iv) structural isotropic 3D T2w v) T2*‐weighted susceptibility imaging vi) multi‐post labelling delay (PLD) 3D ASL, vii) 2D multi‐band‐multiple‐shell diffusion (dMRI), viii) 2D multi‐band EPI BOLD ix) high resolution, limited field of view, imaging centered on the hippocampus. Table 1 gives a detailed summary of the key protocol parameters and lists some of the imaging biomarkers that can be obtained from each series.ResultsStructural imaging was kept consistent across imaging sites. Variations in vendor specific sequences, software upgrades, and hardware performance capabilities, resulted in variations in advanced imaging strategies. Table 2 highlights the advanced sequences introduced in ADNI4, which will gradually be implemented as the required upgrades and licenses are obtained at each site. The main advanced sequences that are new in ADNI4 from ADNI 3 are i) CS for T1w images, ii) pseudo‐continuous‐ASL (pCASL) being implemented on all 3 vendors (GE, Siemens, Philips), iii) multi‐PLD ASL, which allows for the generation of quantitative maps of cerebral blood‐flow (CBF) and arterial transit times (ATTs), iv) and isotropic 1mm3 resolution for 3D FLAIR. When ADNI4 began, product sequences that utilized artificial intelligence were not widely available and could not be adequately tested before the rollout of ADNI4.ConclusionsThe ADNI4 database will continue to help the neuroimaging community, not only extract valuable imaging biomarkers, but provide the data necessary to test the impact of advanced imaging strategies on diagnostic accuracy and disease sensitivity.
Creating Universal Tau PET Scale (Uniτ) Parametric Images – The HEAD Study
AbstractBackgroundThe HEAD study focuses on collecting an extensive dataset from various tau‐PET tracers, aiming to establish robust anchor values, which are essential for harmonizing tau‐PET measurements. Here, we aim to showcase the capability of converting 3D tau‐PET images into a common scale using the Universal Tau‐PET Scale, Uniτ (tau), and to use these 3D images to subsequently obtain ROIs as needed.MethodsWe assessed 185 individuals across the aging and AD spectrum from the HEAD study, with [18F]Flortaucipir and [18F]MK‐6240 tau‐PET tracers. Tau‐PET SUVR images were standardized to a common 8mm FWHM, using the inferior cerebellar gray matter as the reference region. We generated Uniτ Tau‐PET 3D images using a single formula based on Meta‐temporal parameters in two steps: first, within‐tracer anchoring based on young (<25 years) and cognitively impaired individuals; second, between‐tracer anchoring using a piecewise transformation from [18F]Flortaucipir to [18F]MK‐6240, with smoothing at the piecewise inflection point. Subsequently, we extracted mean Uniτ values from these 3D images for key ROIs, including meta‐temporal, mesial, temporo‐parietal, frontal, and Braak stages III‐VI. Finally, we correlated Uniτ values across ROIs between the two tracers to evaluate the accuracy of estimates from the voxel‐wise transformation.ResultsThe original SUVR images present large differences between the two tau‐PET tracers (Figure 1). However, upon applying the Uniτ piecewise transformation with smoothing to all brain voxels, we were able to reasonably harmonize these images to the Uniτ scale, substantially reducing visual variability. Notably, the mean Uniτ values for the ROIs extracted from these harmonized 3D tau‐PET brain images demonstrated a high level of association between the two tracers (Figure 2). Furthermore, estimates generated from ROIs or extracted from our 3D parametric Uniτ model yielded identical estimates (Figure 3).ConclusionThe strong associations between tracers after directly harmonizing 3D images to Uniτ scale using the piecewise transformation with smoothing, underscore the effectiveness of the proposed method. This approach provides a reliable and standardized way to compare tau‐PET data across different tracers. Our results indicate the feasibility of harmonizing 3D tau‐PET images without relying on pre‐established ROIs, overcoming the limitation of restricting the analysis to only few brain regions.
Impact of anchoring cutoff point of tau positivity on CU young or older adults using Flortaucipir and MK‐6240 – Head Study
AbstractBackgroundTau PET provides continuous measurements of tau tangle pathology in the human brain. However, establishing cutoffs is crucial for selecting individuals for treatment in clinical trials or practice. In the absence of postmortem data, PET cutoffs must be established using statistical methods based on what is considered normal tracer uptake. In this study, we tested the impact of various methods to determine tau positivity using two different tau PET tracers in individuals scanned head‐to‐head.MethodsWe studied 147 individuals from Head‐to‐Head Harmonization of Tau Tracers in Alzheimer's Disease (HEAD) study with tau tangle PET scans with [18F]Flortaucipir and [18F]MK‐6240, and amyloid‐β (Aβ) PET. Tau deposition was measured with the standardized uptake value ratio (SUVR) of each agent in the Medial Temporal Lobe (MTL) and the Entorhinal Cortex (EC). To determine Tau positivity three different methods were used: >2.5 standard deviations (SD) than the mean of the young, >1.5 SD mean of the cognitively unimpaired (CU) and >1.5 SD mean of CUAβ‐.ResultsDemographic characteristics of the study population are reported in Table 1. Using the cutoff >2.5 SD mean of young, [18F]Flortaucipir was positive in 35 (23.8%) and 69 (46.9%) individuals in the EC and MTL, respectively. [18F]MK‐6240 was positive in 49 (33.3%) and 58 (39.5%) individuals in EC and in MTL. Using >1.5 SD mean of CUAβ‐, [18F]Flortaucipir was positive in 38 (25.9%) and 45 (30.6%) individuals in EC and MTL, while [18F]MK‐6240 was positive in 51 (34.7%) and 50 (34.0%) in EC and in MTL (Figure 1).ConclusionsOur findings indicate variations in tau positivity when employing different methods based on either the young or CUAβ‐. [18F]Flortaucipir exhibited a higher rate of positive results when the method based on young individuals was applied in the MTL. Conversely, [18F]MK‐6240 showed more consistent and generally higher positivity when other methods were used for cutoff determination and/or in the EC region. Further research with a larger sample size is required to gain a better understanding of the optimal cutoff determination methods for these tracers.
Alzheimer's disease staging with [18F]MK6240 and [18F]Flortaucipir ‐ the HEAD study
AbstractBackgroundUtilizing PET amyloid‐beta (Aβ) and tau for staging Alzheimer’s Disease (AD) has demonstrated potential in identifying individuals with varying degrees of disease severity, applicable to both clinical trials and practice. However, the diverse binding characteristics of tau tracers pose challenges to the application of this staging across different ligands. In this study, we evaluate a novel staging framework proposed by the AA working group, employing Aβ PET and either [18F]MK6240 or [18F]Flortaucipir in individuals participating in a head‐to‐head study of tau PET tracers. This will provide insights into the robustness of this framework when applied across different tau ligands.MethodWe evaluated 154 people (89 CU and 65 CI), with MK6240, Flortaucipir, Aβ‐PET, MRI. SUVRs used inferior cerebellar gray as reference. Participants were categorized in Aβ positivity(A), using visual read, and tau positivity (T) in medial temporal lobe(MTL) and neocortex(NEO). Tau MTL(TMTL) and NEO positivity were defined as average +3 SD of youngs. NEO moderate(TMOD) and high(THIGH) was defined as below and above median SUVR of NEO‐positive individuals. Individuals were grouped as: A‐T‐, A+T‐, A+TMTL+(early), A+TMOD+(intermediary), and A+THIGH+(advanced).ResultsWe found that the distribution of the number of individuals across the five AD stages was comparable when using both MK6240 and Flortaucipir tracers (Figure 1). However, upon examining the concordance within each group, we found a low concordance of 30% (3/10) between tracers in the early tau stage (A+TMTL+). Concordance increased in the intermediary stage A+TMOD+, with a rate of 65% (9/14), and further increased to 79% (11/14) in the advanced stage (A+THIGH+). The A+T‐ group exhibited the highest concordance between tracer results, at 95% (21/22).ConclusionOur findings indicate that the application of this framework across different populations using different tau tracers can lead to a compatible distribution among groups. However, in a direct comparison, we noted that this did not necessarily result in high concordance between groups, as group shifts were observed that depended on the cutoffs used (data not shown). These results emphasize the necessity for harmonization of tau PET tracers prior to their utilization for stage determination.