Functional network segregation is associated with attenuated tau spreading in Alzheimer's disease

Abstract

Introduction Lower network segregation is associated with accelerated cognitive decline in Alzheimer's disease (AD), yet it is unclear whether less segregated brain networks facilitate connectivity‐mediated tau spreading. Methods We combined resting state functional magnetic resonance imaging (fMRI) with longitudinal tau positron emission tomography (PET) in 42 betamyloid‐negative controls and 81 amyloid beta positive individuals across the AD spectrum. Network segregation was determined using resting‐state fMRI–assessed connectivity among 400 cortical regions belonging to seven networks. Results AD subjects with higher network segregation exhibited slower brain‐wide tau accumulation relative to their baseline entorhinal tau PET burden (typical onset site of tau pathology). Second, by identifying patient‐specific tau epicenters with highest baseline tau PET we found that stronger epicenter segregation was associated with a slower rate of tau accumulation in the rest of the brain in relation to baseline epicenter tau burden. Discussion Our results indicate that tau spreading is facilitated by a more diffusely organized connectome, suggesting that brain network topology modulates tau spreading in AD. Highlights Higher brain network segregation is associated with attenuated tau pathology accumulation in Alzheimer's disease (AD). A patient‐tailored approach allows for the more precise localization of tau epicenters. The functional segregation of subject‐specific tau epicenters predicts the rate of future tau accumulation.

ApoE4 and Connectivity-Mediated Spreading of Tau Pathology at Lower Amyloid Levels

“…Our second main finding revealed that ApoE4 was not only associated with accelerated tau accumulation through higher Aβ levels, but also had possible synergistic effects with Aβ on tau spreading, congruent with cross-sectional evidence of higher tau PET at given level of Aβ in ApoE4 carriers . Specifically, we demonstrated that the rate of Aβ-related tau accumulation was moderated by ApoE4 across regions vulnerable to early-stage tau aggregation and spread (ie, in regions strongly connected to tau epicenters Q1-Q2) . These results are congruent with a biphasic AD pathophysiological framework, which proposes an Aβ-dependent then a later Aβ-independent tau accumulation phase, supported by mouse model evidence demonstrating that Aβ-targeting antibodies reduced early but not later tau changes .…”

“…Group demographic characteristics were compared using analyses of variance for continuous and χ 2 tests for categorical variables. ROI-specific annual tau PET change rates were estimated using linear mixed models with longitudinal tau PET SUVR values as the dependent variable and time from baseline as the independent variable, including random slope and intercept …”

“…All neuroimaging acquisition and processing is described in the eAppendix in Supplement 1. Participant-specific tau-spreading ROIs were generated by grouping 95% of regions into quartiles according to their template-based connectivity to 5% of brain regions with highest baseline tau PET (ie, the participant-specific tau epicenter, where Q1 indicates the strongest connectivity to the epicenter and Q4 indicates the weakest connectivity to the epicenter) …”

“…However, tau spreading from the medial temporal lobe to the cortex is modulated by patient-specific factors, including sex, vascular comorbidities, brain network architecture, and genetic predispositions. Aβ thresholds marking tau spreading may vary individually .…”

Longitudinal rates of atrophy and tau accumulation differ between the visual and language variants of atypical Alzheimer's disease

“…PCA and LPA differ in their most prominent sites of initial tau deposition on tau‐PET. Assuming that tau spreads trans‐synaptically, the accumulation of tau pathology over time might be facilitated when the regions with high tau burden are functionally connected to the rest of the brain, with a slower spreading if such regions are instead part of segregated networks 12 . This would lead to differences in rates of tau accumulation between PCA and LPA.…”