Interplay between Neuronal Activity and Tau Pathology

Wetenschappelijke samenvatting proefschrift / Thesis synopsis The main objective of this thesis is to increase understanding of the neurophysiological changes related to Alzheimer's Disease (AD), particularly functional network alterations in the brain. One of the key objectives was to investigate how brain networks could serve as conduits for pathological tau spread, a key feature of AD pathology, by combining magnetoencephalography (MEG) and [18F]flortaucipir positron emission tomography (PET). Furthermore, we aimed to better understand functional connectivity (FC) changes in AD, with the goal of identifying reliable biomarkers for diagnosis. Finally, we evaluated the potential role of MEG to differentiate between different forms of dementia in a memory clinic population. The thesis is divided into three parts: Part I focused on validating FC measures in electroencephalography (EEG) and MEG. Large and well-defined cohorts of AD patients and controls were used to assess reproducibility and robustness of various FC metrics. We found that specific measures, like the amplitude envelope correlation (AEC-c) in the alpha and beta bands and phase-lag index (PLI) in the theta band, were reliable indicators of AD pathology and showed reproducible patterns of connectivity loss. These results highlight that choice of FC measure and frequency band can have large impact on the outcome of MEG and EEG studies in AD. Part II explored the relationship between tau protein and (functional) brain networks in AD, as well as synaptic function, using MEG in combination with [18F]flortaucipir PET. We demonstrated that increased tau pathology is linked to synaptic loss and dysfunction, particularly in the occipital lobe. By making use of an Epidemic Spreading Model, we provided evidence suggesting that functional networks, especially in the early stages of AD, play a significant role in tau propagation. Finally, we identified short network distance to the tau epicenter and so-called ‘hubs’ (central regions in a network) as crucial for the propagation of tau throughout the AD brain. These hubs, which are vital for brain communication and function, were found to accumulate more tau and were identified as uniquely vulnerable to AD-related neurodegeneration, signified by evidence of disproportionate disruption to hub areas in the functional networks of subjects on the AD continuum. These findings are critical to increase understanding of the AD brain, and may highlight vulnerabilities that could be targeted for early identification or treatment. Part III focused on the potential clinical applications of MEG-based neurophysiological characteristics. A machine-learning model using MEG data was developed to classify different types of dementia, achieving a balanced accuracy rate of 41% and multi-class area under the curve value of 0.75 for six-class classification. Classification primarily depended on posterior relative delta, theta and beta power and amplitude-based functional connectivity in the beta and gamma frequency band. DLB and AD subjects could be classified most accurately, while FTD subjects were most frequently misclassified. Diagnostic profiles, describing probability estimates for each of the six diagnoses, were assigned to individual subjects. These results highlight the potential of MEG to increase diagnostic accuracy in a non-invasive manner, particularly when combined with other biomarkers. Diagnostic profiles could provide an intuitive tool to clinicians and could facilitate implementation of the classifier in the differential diagnostics of dementia. Overall, this thesis highlights the importance of functional connectivity in AD, which could offer promising direction for potential therapeutic strategies, by for instance targeting neuronal dynamics to slow down tau propagation and synaptic degeneration.