INTRODUCTION:Chronic traumatic encephalopathy (CTE) is a tauopathy linked to repetitive head impacts. Factors influencing brain regional susceptibility to tau deposition and spreading remain unclear. METHODS:We used three datasets: [18F]flortaucipir positron emission tomography (PET) in 157 former professional American football players and 53 controls (DIAGNOSE CTE); cortical myelin water fractions (MWF) in 50 healthy individuals (Myelin Water Atlas); and white matter (WM) tract MWF and functional connectivity (FC) in 100 healthy individuals (Human Connectome Project). We tested associations between tau-PET uptake and covariance in football players and typical cortical gray matter (GM) MWF, WM tract MWF, and FC. RESULTS:Cortical regions with lower typical GM MWF showed higher tau-PET uptake (β = -0.399, p = 0.001). WM tracts with lower typical MWF were associated with higher tau-PET covariance (β = -0.238, p < 0.001). Higher typical FC was associated with higher tau-PET covariance (β = 0.447, p < 0.001). DISCUSSION:In former football players at risk for CTE, regional susceptibility to tau deposition may be driven by low myelin and high FC.

In Alzheimer’s disease, carriage of the ApoE4 risk allele is linked to faster tau accumulation at lower amyloid-PET levels, thereby accelerating disease progression. However, it remains unclear whether this ApoE4-facilitated transition from amyloidosis to tauopathy is mechanistically promoted by increased secretion of phosphorylated (p)tau, a key intermediate that drives the amyloid-to-tauopathy transition, or alternatively by increased ptau-driven tau aggregation. Therefore, we investigated where along the amyloid-to-tau axis ApoE4 accelerates tau aggregation and assessed i) whether ApoE4 increases ptau secretion or ii) whether ApoE4 increases ptau-associated tau aggregation. To this end, we analysed two large-scale APOE-genotyped cohorts covering the full Alzheimer’s disease spectrum (ADNI: n=201) as well as a preclinical cohort (A4-LEARN: n=200), integrating baseline amyloid-PET, plasma ptau217 and CSF ptau181 with longitudinal tau-PET. Using linear regression, we tested whether ApoE4-carriage moderates i) amyloid-PET-associated plasma ptau217 increases or ii) ptau217-associated tau spreading from local epicentres across patient-tailored tau spreading stages. All analyses were independently validated across both cohorts, including an additional replication in an ADNI subset (n=115) with available CSF ptau181 measures as an alternative marker of ptau secretion. Finally, we used logistic regression to determine ApoE4 allele count-stratified plasma ptau217 thresholds marking early pathological tau-PET increases. We found that ApoE4 did not facilitate amyloid-PET-associated ptau increases, suggesting that amyloid-related ptau secretion is not altered by ApoE4-carriage. Contrastingly, we found that plasma ptau217 elevations were linked to faster tau-PET spread from local epicentres across connected brain regions in an ApoE4-allele dose-dependent manner, independent of amyloid (ADNI/A4-LEARN: mean β=0.44/0.56, p<0.001/<0.001). Lastly, we found that a higher ApoE4 allele count was linked to lower ptau217 thresholds marking transition to tauopathy, i.e. early abnormal tau-PET increases, consistently across both samples (ADNI: 0/1/2 ApoE4 alleles=0.62/0.34/0.15pg/ml, representing ∼45% and ∼76% reductions from non-carriers; Fujirebio ptau217 assay; A4/LEARN: 0/1/2 ApoE4 alleles=0.31/0.23/0.18pg/ml, representing ∼26% and ∼42% reductions; Eli Lilly ptau217 assay). These findings suggest that ApoE4, i.e. the key genetic risk factor for sporadic Alzheimer’s disease, facilitates amyloid-dependent tau aggregation in an allele dose-dependent manner by enhancing the ptau-driven spread of fibrillar tau, leading to an earlier transition from amyloidosis to tauopathy at lower ptau217 levels. This has implications for plasma ptau-based screening approaches and therapeutic timing of anti-amyloid drugs in ApoE4 carriers: Specifically, ApoE4 carriers may require genotype-adjusted ptau thresholds to detect Alzheimer’s disease pathophysiology, as well as anti-amyloid treatment at lower ptau levels to prevent the transition to tauopathy, which ultimately drives neurodegeneration and cognitive decline.

BACKGROUND AND OBJECTIVES:Incidental hyperintense lesions on diffusion-weighted imaging (DWI) are suggested as emerging marker of cerebral small vessel disease (SVD). To further determine their role in SVD, we aimed to describe their prevalence on high-resolution DWI in 2 distinct SVD types. Second, in each SVD type, we aimed to assess incidental DWI-positive lesion distribution and associations with clinical variables. METHODS:Data from 2 hospital-based prospective cohorts in the Netherlands and Germany were used, which included patients meeting the modified Boston criteria for probable cerebral amyloid angiopathy (CAA, BIONIC study) and patients with a confirmed diagnosis of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL, VASCAMY study). In case of a stroke history, patients ≤3 months poststroke were excluded. 3T high-resolution baseline MRIs of patients with probable CAA and baseline, 18-month, and 36-month MRIs of patients with CADASIL were included. All MRI markers were rated following STRIVE-2. Within patient groups, we explored in univariable analyses the association between cardiovascular risk factors and MRI markers and incidental DWI-positive lesions and in multiple regression the association between fluid biomarkers and incidental DWI-positive lesions. RESULTS:Baseline data were available for 43 CAA patients (mean baseline age 71 ± 6 years, 44% female) and 75 CADASIL patients (mean baseline age 53 ± 9.9 years, 62% female). Cross-sectionally, incidental DWI-positive lesions were detected in 24/43 (56% [95% CI 41%-70%]) CAA patients and 16/75 (21% [95% CI 14%-32%]) CADASIL patients. In CAA, 65% of lesions were located in the cortex, whereas in CADASIL, 95% of lesions were located in the subcortical white or gray matter. In CAA patients, DWI-positive lesions were significantly associated with increased neurofilament light chain (NfL) in serum and CSF, but not with other CSF, MRI, or cardiovascular risk factors. In CADASIL patients, DWI-positive lesions were significantly associated with increased serum NfL, increased white matter hyperintensity volume and lacune presence. DISCUSSION:In CAA and CADASIL, the prevalence of incidental DWI-positive lesions is high, and lesions have disease-specific distribution, and associations with serum, CSF, and MRI biomarkers, suggesting that incidental DWI-positive lesions are a feature of SVD. Future studies should investigate their prognostic value.

Amyloid pathology drives tau accumulation, i.e., the key driver of clinical worsening in Alzheimer's disease (AD). Yet, there is considerable heterogeneity in the time of tau onset, as well as in the rates and patterns of tau accumulation, which jointly determine symptom onset and clinical trajectories. The exposure to amyloidosis is predictive of AD progression and may therefore predict tau onset and trajectories. Therefore, we investigated how the age and duration of amyloid onset influence tauopathy onset and accumulation. We included 479/390 ADNI/A4 participants with Flortaucipir tau-PET, and Florbetaben/Florbetapir amyloid-PET. Using sampled iterative local approximation, we determined subject-specific estimated onset ages of amyloid-PET positivity (centiloid>20), and tau-PET positivity (SUVR>1.3). Using robust linear regression, we investigated the associations between estimated amyloid-PET and tau-PET onset ages, the delay between amyloid and tau onset and the effect of amyloid onset on tau-PET change rates. Younger estimated age of amyloid onset predicted younger estimated age of tau onset in the temporal meta ROI (ADNI/A4, b=0.6871/0.7148, p  <0.001/0.001, Figure 1A). This result pattern was pronounced in tau vulnerable temporo-parietal regions, while sparing late Braak regions (Figure 1B). However, a younger estimated age of amyloid onset also predicted a longer delay between amyloid and tau onset, indicating that patients with young onset amyloidosis require longer to develop abnormal tau (Figure 2). By combining sliding window analyses across amyloid onset ages and bootstrapping, we identified that younger amyloid onset is linked to faster tau accumulation, with stronger involvement of parieto-frontal vs. more pronounced temporal lobe tau accumulation in individuals with later amyloid onset (Figure 3). Earlier amyloid onset predicts earlier tau onset and faster more neocortically pronounced tau accumulation. At the same time, younger amyloid onset is linked to a longer delay to tauopathy compared to individuals with older-age amyloid onset. A longer delay between amyloidosis and tauopathy in patients with earlier onset of amyloidosis may widen the window of opportunity for anti-amyloid drugs to prevent more aggressive tauopathy in these at risk individuals.

With the approval of anti-amyloid therapies in Alzheimer's disease (AD), surrogate biomarkers are urgently needed to monitor treatment effects that translate into clinical benefits. Candidate biomarkers, including amyloid-PET, tau-PET, plasma phosphorylated tau ( p -tau), and MRI-assessed atrophy, capture core pathophysiological changes in AD. While cross-sectional biomarker assessments are critical for diagnosis and staging, biomarker change rates may better reflect disease dynamics, making them more suitable for monitoring treatment efficacy. Therefore, we determined which biomarker most effectively tracks cognitive changes in AD, identifying those best suited for efficient monitoring of disease-modifying treatments. We leveraged ADNI ( N  = 108) and A4 ( N  = 151) participants with longitudinal AD biomarker data (global amyloid-PET, temporal meta tau-PET, plasma p -tau 217 , MRI-assessed cortical thickness in the AD signature region) together with cognitive assessments (ADNI: MMSE, ADAS13, CDR-SB; A4: MMSE, PACC). Linear mixed models were used to calculate change rates for biomarkers and cognition. To test whether biomarker changes track cognitive decline, linear models were applied, to test biomarker change rates as a predictor of cognitive change rates. Standardized beta values from bootstrapped linear models were extracted to compare the strengths of correlations between biomarkers and cognitive decline. For non-parametric comparisons, 95% confidence intervals (CIs) of standardized beta values were compared. Models were controlled for age, sex, education, and baseline cognition, with ADNI models additionally adjusted for clinical status. In both cohorts, changes in temporal tau-PET, plasma p -tau 217 , and MRI-assessed cortical thickness were associated with cognitive decline (ADNI: Figure 1; A4: Figure 2). Amyloid-PET changes showed no significant association with cognitive changes (ADNI: Figure 1A+F+K; A4: Figure 2A+F). Bootstrapping confirmed that tau-PET, plasma p -tau 217 , and cortical thickness track cognitive decline, but not amyloid-PET (ADNI: Figure 1E+J+O; A4: Figure 2E+J). Overlapping CIs for tau-PET and plasma p -tau 217 indicated comparable predictive accuracy. Our findings demonstrate that tau-PET and plasma p -tau 217 are robust biomarkers for monitoring cognitive changes, with plasma p -tau 217 offering a cost-effective, scalable alternative for clinical use. Changes in amyloid-PET do not reliably reflect cognitive decline, limiting its utility as a treatment monitoring tool. Although cortical thickness correlates with cognitive changes, its application is limited by pseudoatrophy and volume loss induced by anti-amyloid antibody treatments.

Background and Objectives Incidental hyperintense lesions on diffusion-weighted imaging (DWI) are suggested as emerging marker of cerebral small vessel disease (SVD). To further determine their role in SVD, we aimed to describe their prevalence on high-resolution DWI in 2 distinct SVD types. Second, in each SVD type, we aimed to assess incidental DWI-positive lesion distribution and associations with clinical variables.Methods Data from 2 hospital-based prospective cohorts in the Netherlands and Germany were used, which included patients meeting the modified Boston criteria for probable cerebral amyloid angiopathy (CAA, BIONIC study) and patients with a confirmed diagnosis of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL, VASCAMY study). In case of a stroke history, patients <= 3 months poststroke were excluded. 3T high-resolution baseline MRIs of patients with probable CAA and baseline, 18-month, and 36-month MRIs of patients with CADASIL were included. All MRI markers were rated following STRIVE-2. Within patient groups, we explored in univariable analyses the association between cardiovascular risk factors and MRI markers and incidental DWI-positive lesions and in multiple regression the association between fluid biomarkers and incidental DWI-positive lesions.Results Baseline data were available for 43 CAA patients (mean baseline age 71 +/- 6 years, 44% female) and 75 CADASIL patients (mean baseline age 53 +/- 9.9 years, 62% female). Cross-sectionally, incidental DWI-positive lesions were detected in 24/43 (56% [95% CI 41%-70%]) CAA patients and 16/75 (21% [95% CI 14%-32%]) CADASIL patients. In CAA, 65% of lesions were located in the cortex, whereas in CADASIL, 95% of lesions were located in the subcortical white or gray matter. In CAA patients, DWI-positive lesions were significantly associated with increased neurofilament light chain (NfL) in serum and CSF, but not with other CSF, MRI, or cardiovascular risk factors. In CADASIL patients, DWI-positive lesions were significantly associated with increased serum NfL, increased white matter hyperintensity volume and lacune presence.Discussion In CAA and CADASIL, the prevalence of incidental DWI-positive lesions is high, and lesions have disease-specific distribution, and associations with serum, CSF, and MRI biomarkers, suggesting that incidental DWI-positive lesions are a feature of SVD. Future studies should investigate their prognostic value.

INTRODUCTION:Amyloid-induced tauopathy drives clinical decline in Alzheimer's disease (AD). Because age and sex shape tau trajectories, defining patient-centered amyloid thresholds for tauopathy onset could facilitate pre-tauopathy AD identification and aid treatment decisions and prognosis. METHODS:By including two samples (Alzheimer's Disease Neuroimaging Initiative [ADNI, n = 301]; and 18F-AV-1451-A05 [A05, n = 143]), we explored whether age and sex affect tauopathy transition and determined patient-centered amyloid positron emission tomography (PET) thresholds that mark tauopathy onset. RESULTS:We found a consistent amyloid PET × age interaction on global tau PET increase in men (ADNI/A05: p = 0.0078/0.018), with younger men showing faster amyloid-associated tau accumulation. We then established patient-centered, amyloid PET-inferred tauopathy transition cut-offs. Women reached this transition at lower amyloid PET levels, and these cutoffs predicted both earlier onset and accelerated cognitive decline (p < 0.001). DISCUSSION:This study highlights the effect of age and sex on the amyloid-to-tauopathy transition, establishes patient-centered amyloid PET thresholds for tauopathy onset, and links these thresholds to accelerated cognitive decline. HIGHLIGHTS:Younger age is related to faster amyloid-related tau accumulation in men. We defined a series of amyloid positron emission tomography (PET) thresholds to enable patient-centered inference of amyloid-related tauopathy. Crossing the amyloid PET-defined tauopathy phase is associated with more progressive tau deposition and cognitive decline.

INTRODUCTION We investigated how plasma biomarkers (phosphorylated tau 217 [ptau(217)], glial fibrillary acidic protein [GFAP], neurofilament light chain [NfL]) relate to imaging markers of small vessel disease (SVD) and Alzheimer's disease (AD), and cognition in memory clinic patients. METHODS 76 memory clinic patients underwent plasma biomarker assessment, neuropsychological testing, and 3T MRI. SVD burden was assessed using white matter hyperintensity (WMH) volume, mean skeletonized mean diffusivity (MSMD), and fiber density. AD-related neurodegeneration was captured by AD-signature cortical thickness and fiber-bundle cross-section. Findings were validated in 41 Alzheimer's Disease Neuroimaging Initiative (ADNI) participants with amyloid-/tau-positron emission tomography (PET). RESULTS Associations varied between biomarkers. NfL showed strongest associations with SVD burden, ptau(217) with AD-related neurodegeneration, while GFAP was linked to both. SVD markers were associated with processing speed, whereas AD markers were most associated with memory. DISCUSSION NfL relates to SVD burden, while ptau(217) remains most sensitive to AD-related biomarkers. GFAP's dual associations suggest overlapping biological processes. Together, coexisting SVD should be considered when interpreting plasma biomarkers in memory clinic patients.

Understanding factors influencing Alzheimer's disease (AD) progression is crucial for optimising treatment timing and targets. A major genetic risk factor, the Apolipoprotein E ε4 allele (ApoE4), is associated with earlier tau pathology accumulation and spread at lower amyloid‐beta (Aβ) levels (Steward, JAMA Neurol, 2023). However, the mechanisms underlying this association remain unclear (Figure 1A). Therefore, we assessed how ApoE4 accelerates Aβ‐related tau aggregation. Specifically, we investigated whether ApoE4 promotes Aβ‐driven secretion of phospho tau ( p ‐tau) or ptau dependent tau aggregation, and determined whether ApoE4 promotes tau pathology in an allele dose‐dependent manner. We analysed data from APOE ‐genotyped AD‐spectrum participants in the ADNI ( n = 201) and A4 cohorts ( n = 200), integrating cross‐sectional fluid biomarker measures (plasma ptau 217 , CSF ptau 181 ) and longitudinal Flortaucipir tau‐PET and Florbetaben/Florbetapir amyloid‐PET. Using linear regression, we assessed whether the interaction between amyloid‐PET and ApoE4 allele dosage influences plasma ptau 217 , and replicated this analysis with CSF ptau 181 in an ADNI subset ( n = 115). Secondly, to investigate whether ApoE4 enhances tau fibrilisation and spread, we calculated annual tau‐PET SUVR accumulation rates across a connectivity‐based tau spreading stages, using our prior methodology (e.g. Franzmeier, Sci Adv, 2020). Linear regressions tested the interaction between ptau 217 (or CSF ptau 181 ) and ApoE4 allele count on connectivity‐mediated tau‐PET accumulation in four connectivity stages that capture progressive tau spread. ApoE4 allele dosage did not moderate the relationship between amyloid‐PET and plasma ptau 217 in either sample (Figure 1B, ADNI: β=0.13, p = 0.32; A4: β=‐0.20, p = 0.17) nor between amyloid‐PET and CSF ptau 181 in ADNI subsample (Figure 1B, b=‐.16, p = 0.42). However, a significant ApoE4 allele dose effect was observed in moderating the relationship between plasma ptau 217 and tau‐PET accumulation across connectivity stages independent of amyloid burden (Figure 1C, ADNI: Q1–4 mean β=0.44, Q1‐4 p <0.001; A4: Q1‐4 mean β = 0.56, Q1,2,4 p <0.001, Q3 p <0.05), with the strongest effect in individuals carrying two ApoE4 alleles. ApoE4 exerts an allele dose‐dependent effect on ptau induced tau aggregation, driving accelerated tau spreading at lower Aβ levels. This suggests that attenuating soluble ptau increases in ApoE4 carriers may mitigate downstream tau fibrilisation and delay dementia onset, highlighting the potential of personalised therapeutic approaches.

Neurodegenerative 4-repeat (4R) tauopathies commonly manifest as progressive supranuclear palsy (PSP). PSP patients show elevated PI-2620-PET in subcortical 4R tau predilection sites (e.g., globus pallidus), suggesting PI-2620-PET as a promising 4R tau neuroimaging candidate. However, optimal quantification of PI-2620-PET in 4R tauopathies remains challenging, as conventional cerebellar tau-PET reference regions also accumulate 4R tau. We aimed to use unbiased image-derived input function (IDIF) PET data to determine an optimized PET reference region for in vivo quantification of 4R tau. We obtained 60-minute dynamic PI-2620-PET in 54 PSP Richardson Syndrome (PSP-RS) patients and 19 healthy controls (HC), applying IDIF-modeling using carotid timeseries to assess unbiased PI-2620-PET binding and determine total distribution volume (VT). Through an iterative approach, we intensity-normalized VT-images against white-matter regions in the Hammers brain atlas, identifying regions where intensity-normalized pallidum PET values showed the largest PSP-RS vs. HC differences. White-matter regions with strongest PSP-RS vs. HC differences surviving multiple-comparison correction were summarized into a single reference region spanning bilateral temporo-orbital white-matter. This ROI was then used to determine SUVRs using conventional 20-40 minute PI-2620-PET data in PSP-RS, a PSP-non-RS validation sample ( n  = 63), as well as non-tau disease controls (i.e., alpha-synucleinopathies, n  = 20; Alzheimer's disease, n  = 23). Using PI-2620 SUVRs obtained with the temporo-orbital white-matter reference, we detected strong PSP-RS vs. HC group differences in basal ganglia SUVRs using voxel-wise comparisons ( p  <0.001, FWE-cluster corrected). Similar basal ganglia differences were detected for PSP-non-RS vs. HC, but not for alpha-syn (no group differences) or AD vs. HC (cortical AD-like group differences). In contrast, minimal group differences were found using a conventional inferior cerebellar grey matter reference region. Our findings strongly suggest temporo-orbital white-matter is superior to inferior cerebellum as a reference region for PI-2620-PET imaging in 4R tauopathies, due to increased sensitivity and purported specificity for 4R tau.

Lewy body pathology consisting of aggregated alpha-Synuclein (a-Syn) is the hallmark pathology in Parkinson’s disease, yet a-Syn aggregates are also commonly observed post-mortem as a co-pathology in Alzheimer’s disease (AD) patients. Preclinical research has shown that a-Syn can amplify Ab-associated tau seeding and aggregation, hence a-Syn co-pathology may contribute to the Ab-induced progression of tau pathology in AD. To address this, we combined a novel CSF-based RT-QuIC seed-amplification assay to determine a-Syn positivity, with PET-neuroimaging in a large patient cohort ranging from cognitively normal to dementia, to determine whether a-Syn co-pathology accelerates Ab-driven tau accumulation. In 261 Ab-positive vs. 272 Ab-negative subjects ranging from cognitively normal to dementia we employed amyloid-PET, Flortaucipir tau-PET and a CSF-based a-Syn RT-QuIC assay for in vivo detection of abnormal a-Syn aggregation. A subset of 136 Ab-positive vs. 102 Ab-negative subjects had longitudinal tau-PET across ∼2.5years. Using linear regression, we tested whether a-Syn positivity was linked to stronger Ab-related tau aggregation (i.e. interaction a-Syn x amyloid-PET on tau-PET). Prevalence of a-Syn positivity rose across increasing clinical severity and was particularly pronounced in Ab+ (i.e. CN/MCI/Dementia=20/23/47%) vs. Ab- subjects (i.e. CN/MCI/Dementia=15/10/29%), suggesting that a-Syn co-pathology is more common in clinically advanced AD (chi-squared-test, p<0.001). When testing the interaction between a-Syn and global amyloid-PET, we found that a-Syn positivity was associated with stronger Ab-related tau deposition (Figure 1A, p<0.001), and faster Ab-related tau accumulation rates (Figure 1B, p=0.010) in typical tau vulnerable brain regions (i.e. temporal meta ROI), adjusting for age and sex. Regional analyses confirmed that higher regional amyloid-PET was associated with stronger temporal-lobe tau deposition (Figure 2A) and faster tau accumulation in downstream regions (Figure 2B) in a-Syn positive individuals. In addition, there was an independent effect of a-Syn positivity on faster temporal lobe tau accumulation rates controlling for age, sex and global amyloid, suggesting that a-Syn may also independently contribute to tau aggregation (Figure 3). a-Syn co-pathology as detected by CSF seed-amplification assays is more common at clinically advanced AD and related to faster Ab-related tau aggregation. This suggests that a-Syn co-pathology may actively contribute to AD-related tau accumulation and therefore contribute to dementia development.

In Alzheimer’s disease (AD), cortical tau aggregation is a strong predictor of cortical brain atrophy as shown by MRI and PET studies, particularly driving the degeneration of neuronal somata in the grey matter. However, tau’s physiological role is to stabilize microtubules within axons in the brain’s white matter (WM) pathways. Therefore, tau’s white-to-grey-matter translocation and aggregation in neurofibrillary tangles close to neuronal somata may induce WM degeneration through destabilization of axonal microtubule integrity. To address this, we determined whether cortical tau predicted faster atrophy of connected WM tracts in AD. We included from ADNI cohort 37 amyloid-PET negative (Aβ-) cognitively normal (CN) participants and 88 amyloid-PET positive (Aβ+) participants across the AD-spectrum (i.e. CN/MCI/Dementia = 50/28/10), with baseline amyloid-PET, longitudinal tau-PET and longitudinal structural MRI data. For replication, we included baseline amyloid-PET, tau-PET and MRI data of 321 CN-Aβ+ subjects from the A4 cohort. T1-weighed MRIs were segmented into grey and white matter and non-linearly normalized to MNI space using CAT12. The cortical Brainnetome Atlas and a diffusion imaging-based tractography atlas were applied tau-PET and MRI data to i) determine cortical tau-PET accumulation rates within Brainnetome ROIs, and ii) assess WM volume changes within fiber tracts connected to each cortical ROI. Statistical regression-models were adjusted for age, sex, WM hyperintensity volume, global amyloid, intracranial volume, and APOE4-status. In ADNI, higher baseline temporo-parietal tau-PET predicted faster volume reductions in connected WM tracts (Figure 1A), especially pronounced in Aβ+ subjects (Figure 1B). Similarly, faster tau accumulation was strongly linked to widespread WM volume reductions in connected fiber tracts (Figure 2A; Figure 2B), particularly exacerbated in Aβ+ APOE4 carriers (Figure 3A) compared to non-carriers (Figure 3B). These results suggest that tau accumulation and WM degeneration are parallel processes in AD, modulated by APOE4. In the A4 validation sample of preclinical AD patients, we detected congruent associations between inferio-temporal tau-PET increase and reduced WM volume in connected fiber tracts (Figure 4). Cortical tau aggregation is associated with progressive WM atrophy in connected fiber tracts throughout the brain, suggesting that tau accumulation triggers axonal degeneration, which may induce neuronal disconnection and dysfunction and thereby contributing to AD progression.

Aggregated alpha-Synuclein (αSyn) is a hallmark pathology in Parkinson’s disease but also one of the most common co-pathologies in Alzheimer’s disease (AD). Preclinical studies suggest that αSyn can exacerbate tau aggregation, implying that αSyn co-pathology may specifically contribute to the Aβ-induced aggregation of tau that drives neurodegeneration and cognitive decline in AD. To investigate this, we combined a novel CSF-based seed-amplification assay (SAA) to determine αSyn positivity with amyloid- and tau-PET neuroimaging in a large cohort ranging from cognitively normal individuals to those with dementia, examining whether αSyn co-pathology accelerates Aβ-driven tau accumulation and cognitive decline. In 284 Aβ-positive and 308 Aβ-negative subjects, we employed amyloid-PET, Flortaucipir tau-PET, and a CSF-based αSyn seed-amplification assay (SAA) to detect in vivo αSyn aggregation. CSF p-tau181 measures were available for 384 subjects to assess earliest tau abnormalities. A subset of 155 Aβ-positive and 135 Aβ-negative subjects underwent longitudinal tau-PET over approximately 2.5 years. Using linear regression models, we analyzed whether αSyn SAA positivity was linked to stronger Aβ-related increases in baseline fluid and PET tau biomarkers, faster Aβ-driven tau-PET increase, and more rapid cognitive decline. αSyn SAA positivity was more common in Aβ + vs. Aβ- subjects and increased with clinical severity (p < 0.001). Most importantly, αSyn positivity was also associated with greater amyloid-associated CSF p-tau181 increases (p = 0.005) and higher tau-PET levels in AD-typical brain regions (p = 0.006). Longitudinal analyses confirmed further that αSyn positivity was associated with faster amyloid-related tau accumulation (p = 0.029) and accelerated amyloid-related cognitive decline, potentially driven driven by stronger tau pathology. Our findings suggest that αSyn co-pathology, detectable via CSF-based SAAs, is more prevalent in advanced AD and contributes to the development of aggregated tau pathology thereby driving faster cognitive decline. This highlights that a-Syn co-pathology may specifically accelerate amyloid-driven tau pathophysiology in AD, underscoring the need to consider αSyn in AD research and treatment strategies.

INTRODUCTION:With anti-amyloid beta (Aβ) therapies approved for Alzheimer's disease (AD), surrogate biomarkers are needed to monitor clinical treatment efficacy. Therefore, we systematically compared longitudinal changes in A/T/N biomarkers (amyloid-positron emission tomography [PET], tau-PET, plasma phosphorylated tau at threonine 217 [p-tau217], and magnetic resonance imaging) for tracking cognitive changes. METHODS:We analyzed longitudinal biomarker and cognitive change rates from the Alzheimer's Disease Neuroimaging Initiative (N = 141) and Anti-Amyloid Treatment in Asymptomatic Alzheimer's (A4) and Longitudinal Evaluation of Amyloid Risk and Neurodegeneration (LEARN) (N = 151), estimated using linear mixed models. Using linear models, we tested biomarker changes as predictors of cognitive changes, comparing predictive strengths across biomarkers using bootstrapping. RESULTS:Tau-PET, plasma p-tau217, and cortical thickness changes accurately tracked change rates in Mini-Mental State Examination, Alzheimer's Disease Assessment Scale-Cognitive Subscale 13-item version, Clinical Dementia Rating-Sum of Boxes, and Preclinial Alzheimer Cognitive Composite scores. In contrast, amyloid-PET change rates were not linked to cognitive changes. DISCUSSION:Plasma p-tau217 offers a cost-effective AD-specific alternative to tau-PET and could potentially be implemented for monitoring cognitive changes in AD trials, while amyloid-PET lacks utility. Cortical thickness changes accurately track cognitive changes but may be confounded by pseudo-atrophy in anti-Aβ treatments. HIGHLIGHTS:Longitudinal changes in tau-PET, plasma p-tau217, cortical thickness - but not amyloid-PET - effectively track cognitive decline. Cortical thickness may be confounded by pseudo-atrophy in anti-Aβ trials. Plasma p-tau217 is a robust and cost-effective alternative to tau-PET as an AD-specific surrogate biomarker for monitoring cognitive changes.

The link between amyloidosis and tauopathy in Alzheimer’s Disease (AD) remains unclear. Both in-vitro and in-vivo studies have shown that amyloid-beta (Aβ) induces neuronal hyperexcitability and since tau spreads trans-synaptically in an activity-dependent manner, Aβ-related hyperactivity may promote tau propagation and serve as a treatment target to slow disease progression. However, the temporal dynamics of this mechanism across the disease course remains uncharacterized — an essential step for informing mechanistic models and interventions. We previously demonstrated that regional Aβ is associated with hyperconnectivity from tau epicentres to posterior regions in preclinical AD. Thus, we hypothesized that Aβ-induced hyperexcitability and associated hypermetabolism may promote early-stage tau accumulation and spread. From ADNI, we included subjects across the early AD continuum with baseline amyloid-PET and FDG-PET as a marker for neuronal activity and tau-PET ∼4.8 years later ( N =237). Regional FDG-PET SUVRs were z -score-transformed relative to age- and sex-matched Aβ-negative controls (centiloid < 0). We computed within-subject correlations of regional centiloid and FDG-PET. Subject-specific mediation models were used to test whether higher Aβ promotes later tau accumulation via FDG-PET increases. We employed a sliding-window approach to quantify the mean mediation effect across different centiloid and tau levels. Higher regional centiloid is significantly associated with stronger regional metabolism (mean r =0.14, p <0.001). Further, future tau-PET is best predicted by linear models including Aβ and FDG and their interaction term (Figure 1). We found an overall significant mediation effect of baseline Aβ through FDG-PET on tau-PET ∼4.8 years later (ACME=0.013, p <0.001, T (236)=3.75). Assessing the temporal dynamics of the mediation effect throughout the disease course showed a significant positive mediation pathway between 12.8-21.2 centiloids and 1.0-1.22 tau-PET SUVR in the temporal meta-ROI (Figure 2). Subjects within this centiloid range show significantly increased FDG-PET metabolism, promoting downstream tau-PET increases (Figure 3). As expected, the FDG-tau association decreases as the disease progresses, attenuating the mediation effect. Our results suggest that early-stage Aβ-related hypermetabolism drives tau accumulation already at low centiloid levels, preceding tau positivity in the temporal meta-ROI and decreases as disease progresses. Trials targeting neuronal hyperactivity may be most effective at early-stage Aβ deposition and low fibrillar tau.

Cerebral small vessel disease (SVD) is highly prevalent in older adults and a key comorbidity in neurodegenerative conditions such as Alzheimer's disease. While conventional MRI markers (e.g. white matter hyperintensities) capture late SVD stages, diffusion MRI metrics are sensitive to early SVD-related brain changes, making them key candidates as surrogate endpoints in clinical trials. The diffusion MRI marker “peak width of skeletonized mean diffusivity” (PSMD) has been used and validated in various SVD studies. Here, we developed this marker further and propose PSMD-2, which is specifically tailored for longitudinal assessment. Improvements over the prior version include a) skeletonization using free-water corrected diffusion maps, b) using a within-subject template and a common skeleton across all timepoints, and c) allowing flexible use of multiple diffusion parameters (see Figure 1). For PSMD-2 development, we included n  = 120 subjects with familial SVD (CADASIL from the DiViNAS study; 3T Philips, 2-years follow-up). For validation, we included n  = 17 CADASIL patients (VASCAMY study; 3T Siemens, 18-months follow-up) and n  = 72 patients with a history of lacunar stroke (“sporadic” SVD from the SCANS study; 1.5T GE, 2-years follow-up). PSMD-2 was benchmarked against the conventional PSMD version via estimating required sample sizes for a hypothetical clinical trial using PSMD or PSMD-2 as surrogate endpoints (power=80%, alpha=5%, mean change=30%). PSMD-2 closely tracked disease progression in all samples and yielded smallest sample size estimates in the CADASIL development dataset (24.3% reduction) as well as in both validation datasets (CADASIL: 31.0% reduction; sporadic SVD: 18.6% reduction). Validation in cerebral amyloid angiopathy (CAA) and instrumental validation addressing repeatability and reproducibility are ongoing. PSMD-2 provides improved sensitivity to detect SVD-related brain changes compared to the previous PSMD version, leading to smaller required sample sizes for clinical trials. Thus, PSMD-2 is a key candidate as surrogate endpoint in SVD-targeting trials, for monitoring disease progression and for capturing comorbid vascular brain changes in AD in longitudinal settings.

BACKGROUND:Tau accumulation drives neurodegeneration and cognitive decline in Alzheimer's Disease (AD) and preclinical research suggests that tau spreads transsynaptically across connected neurons. We translated tau spreading models to human neuroimaging data, showing that tau pathology spreads from circumscribed epicenters to connected regions in AD, following the architecture of functional brain networks. To further determine whether the topology of brain networks influences tau spreading dynamics, we investigated whether functional hubs (i.e. regions with strong inter-regional connections) accelerate tau spread in AD. Specifically, we hypothesized that more efficient communication from tau epicenters towards hubs that cross-link large-scale brain networks (connector hubs) rather than hubs that interconnect neighboring regions (local hubs) accelerates amyloid-related tau accumulation and cognitive decline (Figure 1). METHOD:Longitudinal tau/amyloid-PET and cognitive data from two independent cohorts covering the AD spectrum (ADNI/A4 n = 325/220) were analyzed to examine amyloid-driven spatiotemporal tau accumulation patterns and cognitive decline. Structural- and functional-connectivity templates from healthy controls were used to model the connectional efficiency of subject-level tau epicenters (i.e. 10% of brain regions with highest baseline tau-PET) towards connector/local hubs (Figure 2). Using robust regression, we then tested whether more efficient communication of subject-level tau epicenters to connector vs. local hubs accelerated global tau accumulation, cognitive decline, and tau dissemination across networks. RESULT:Supporting our hypotheses, we found that the effect of higher baseline amyloid-PET on faster global tau-PET increases was moderated by more efficient communication of tau epicenters towards connector relative to local hubs (ADNI/A4: β = 0.31/0.40, p <0.001/0.03), such that subjects with stronger epicenter communication to connector hubs showed an amplified effect of amyloid on global tau accumulation rates (Figure 3A). The same interaction models also predicted faster cognitive decline (ADNI/A4: β = -0.49/-0.34, p <0.001/0.04, Figure 3B), and larger extents of tau dissemination across functional networks (ADNI/A4: β = 0.6/0.36, p <0.001/0.04). All p-values were FDR-corrected. CONCLUSION:Brain network topology shapes spatiotemporal tau accumulation rates and cognitive trajectories in AD. Specifically, stronger communication of tau epicenters with connector hubs that are characterized by widespread cross-network connections amplifies amyloid-related tau accumulation. This suggests that brain network architecture has a profound modulating impact on tau aggregation and disease progression in AD.

The myelin sheath around axons is of fundamental importance for signal transduction. Myelin is reduced in white matter hyperintensities (WMH), which occur in both small vessel disease (SVD) and Alzheimer’s disease (AD), giving rise to the question to what extent myelin is reduced in these diseases. Here, we employed an advanced MRI based method to assess myelin independently from a major confounding factor, i.e. iron-related signal, in monogenic small vessel disease (i.e. CADASIL) and in mild cognitive impairment (MCI) & AD dementia. We included 62 CADASIL subjects, 11 with MCI or AD dementia, and 22 elderly controls (HC). Using 3D-T2-star-weighted multi-echo gradient-echo MRI, we performed susceptibility source separation of diamagnetic (|χ-negative|, e.g. myelin) and paramagnetic (χ-positive, e.g. iron) sources (Shin et al., 2021). Mean diffusivity (MD) was calculated from diffusion tensor imaging. We extracted |χ-negative| and MD values within WMH, normal appearing white matter (NAWM), and two ROIs including the left anterior thalamic radiation and forceps minor (genu) of the corpus callosum, i.e. two strategic tracts for processing speed. Subject-level difference-scores between CADASIL or MCI/dementia patients and group-averaged HC were derived as abnormality scores. Voxel-based WMH frequencies were mapped for each group. Group differences in |χ-negative| and MD values were tested using linear regressions, controlled for χ-positive scores, age, sex, and education. Voxel-wise proportions of WMH are mapped for each group in Figure 1. For CADASIL, |χ-negative| values were decreased, and MD values increased in each ROI compared to the HC group, with intermediate values for the MCI/dementia group (Figure 2 A&B). Worse |χ-negative| and MD alterations were observed in WMH compared to NAWM (Figure 2 C-F) in each disease group. Lower |χ-negative| values were associated with higher MD in WMH (r=-0.57, Figure 3) and NAWM (r=-0.56). |χ-negative| values showed a marked decrease in WMH of CADASIL patients, suggesting myelin loss in SVD, with less pronounced myelin reduction present in MCI/AD. |χ-negative| values were only moderately associated with MD, suggesting that each provides complimentary information. Our results encourage future studies to test the cognitive consequences of myelin loss in SVD and MCI.

BACKGROUND:Cerebral small vessel disease (SVD) is highly prevalent in older adults and a key comorbidity in neurodegenerative conditions such as Alzheimer's disease. While conventional MRI markers (e.g. white matter hyperintensities) capture late SVD stages, diffusion MRI metrics are sensitive to early SVD-related brain changes, making them key candidates as surrogate endpoints in clinical trials. The diffusion MRI marker "peak width of skeletonized mean diffusivity" (PSMD) has been used and validated in various SVD studies. Here, we developed this marker further and propose PSMD-2, which is specifically tailored for longitudinal assessment. Improvements over the prior version include a) skeletonization using free-water corrected diffusion maps, b) using a within-subject template and a common skeleton across all timepoints, and c) allowing flexible use of multiple diffusion parameters (see Figure 1). METHOD:For PSMD-2 development, we included n = 120 subjects with familial SVD (CADASIL from the DiViNAS study; 3T Philips, 2-years follow-up). For validation, we included n = 17 CADASIL patients (VASCAMY study; 3T Siemens, 18-months follow-up) and n = 72 patients with a history of lacunar stroke ("sporadic" SVD from the SCANS study; 1.5T GE, 2-years follow-up). PSMD-2 was benchmarked against the conventional PSMD version via estimating required sample sizes for a hypothetical clinical trial using PSMD or PSMD-2 as surrogate endpoints (power=80%, alpha=5%, mean change=30%). RESULT:PSMD-2 closely tracked disease progression in all samples and yielded smallest sample size estimates in the CADASIL development dataset (24.3% reduction) as well as in both validation datasets (CADASIL: 31.0% reduction; sporadic SVD: 18.6% reduction). Validation in cerebral amyloid angiopathy (CAA) and instrumental validation addressing repeatability and reproducibility are ongoing. CONCLUSION:PSMD-2 provides improved sensitivity to detect SVD-related brain changes compared to the previous PSMD version, leading to smaller required sample sizes for clinical trials. Thus, PSMD-2 is a key candidate as surrogate endpoint in SVD-targeting trials, for monitoring disease progression and for capturing comorbid vascular brain changes in AD in longitudinal settings.

Neuroimaging studies have revealed age and sex-specific differences in Alzheimer’s disease (AD) trajectories. However, how age and sex modulate tau spreading remains unclear. Thus, we investigated how age and sex modulate the amyloid-beta (Aβ)-induced accumulation and spreading of tau pathology from local epicenters across connected brain regions. We included 313 ADNI participants (female/male, n = 167/146), i.e. 110 cognitively normal (CN) Aβ-negative, and 203 Aβ-positive subjects across the AD spectrum (i.e. CN/MCI/Dementia, n = 98/70/35) with baseline amyloid-PET and longitudinal Flortaucipir tau-PET. Annual tau-PET change rates for 200 cortical regions of the Schaefer atlas were calculated. Sex-specific resting-state fMRI-connectivity templates across the 200 Schaefer regions were determined in independent Aβ-negative controls (female/male, n = 118/82) to determine the connectivity of tau epicenters to the rest of the brain. Using linear regression, we investigated interactions between age, sex and Aβ on tau accumulation and spread, controlling for APOE4-status and diagnosis. Higher Aβ (i.e. centiloid) predicted faster tau accumulation, where this association was pronounced in younger individuals (i.e. age x centiloid interaction, b = -3.64, p<0.001, Fig. 1A). This age x centiloid interaction was stronger in men (b = -4.82, p<0.001, Fig. 1B) vs. women (b = -1.67, p = 0.029, Fig. 1C), suggesting that younger age promotes Aβ-related tau accumulation predominantly in men. Bootstrapping analysis further confirmed this effect (Fig. 1D). In Aβ+, epicenters with highest baseline tau-PET showed a similar temporal-lobe distribution in men and women (Fig. 2A&B), yet epicenter connectivity to the rest of the brain was stronger in men vs. women (Fig. 2C). Stronger connectivity of tau epicenters to the rest of the brain was linked to faster tau accumulation especially in younger Aβ+ subjects (i.e. interaction age x epicenter connectivity, b = 4.41, p<0.001, Fig. 3A). However, this effect was clearly driven by men (b = 6.13, p<0.001, Fig. 3B) and not observed when tested in women only (b = 1.55, p = 0.252, Fig. 3C). Aβ drives faster tau accumulation and this effect is particularly strong at younger age and even further pronounced in men, whose tau epicenters are more densely interconnected with the rest of the brain. Together, age and sex have clear modulating effects on tau spreading, and heterogeneous AD trajectories may be partly arisen due to sex-specific differences in brain network architecture.