Behavioral Neurology and Dementia: Emerging Diagnostic and Therapeutic Biomarkers in Neurodegeneration*

Dr. Albin will review recent PET imaging work assessing dynamic regional changes in cholinergic terminal markers related to cognitive deficits in Parkinson disease (PD). Varying increased and diminished regional expression of the vesicular acetylcholine transporter (VAChT), measured with FEOBV PET, correlate with cognitive status of early and mild-moderate PD subjects. Centrocingular cholinergic deficits correlate with cognitive deficits and may predict subsequent cognitive decline. Relative upregulation of VAChT expression may be a feature of early PD and may signal relative resistance to cognitive decline, perhaps by providing compensation of cognitive deficits secondary to dopaminergic deficits.

Alzheimer's disease (AD) presents a complex genetic challenge. Our innovative Meta-PRS method, integrating whole genome sequencing (WGS) and deep learning via Graph Convolutional Neural Networks (GCNN), refines polygenic risk scores (PRS) for personalized risk assessment. Analyzing 94 AD-related variants, Meta-PRS enhances PRS accuracy by incorporating phenotype and pathway-specific clustering. We identified six key pathways influencing AD risk: Immune response, Beta-amyloid metabolism, Notch signaling, Cholesterol/lipid dysfunction, Homeostasis, and Morphogenesis. Notch signaling showed a notable impact with an OR of 4.06. The method also highlighted the Cholesterol/LDL/Obesity cluster with the highest OR at 4.23. This personalized approach surpasses traditional models, offering tailored risk profiles and paving the way for bespoke interventions based on individual genetic landscapes. Meta-PRS marks a substantial advancement in precision medicine, promising more effective, individualized treatment strategies for AD patients.

Studies of microglial gene manipulation in mouse models of Alzheimer’s disease amyloidopathy can cause unpredictable effects on various key endpoints, including amyloidosis, tauopathy, inflammation, neuritic dystrophy, neurodegeneration, and learning behavior. The presentation and discussion will focus on multiple examples in which observed results have been difficult to reconcile with predicted results based on precedent because key endpoints do not reliably track together. The pathogenesis of AD involves multiple cell types and complex events that may change with disease stage. Cell type targeting and timing of intervention are responsible for the sometimes impossibility of predicting whether any prospective therapeutic intervention should aim at increasing or decreasing the level or activity of a particular molecular target.

We leverage our institutional dataset of retinal color fundus photographs from subjects with pre-symptomatic AD and computationally advanced natural language processing models, to train and validate PreADFound, a novel BERT (bidirectional encoder representations from transformers)-style self-supervised learning convolutional neural network (CNN). In a real-world population with intact cognition, PreADFound achieved 85.11% accuracy to discriminate pre-symptomatic AD based on retinal fundus images alone. Superior and inferior vascular arcades close to the optic nerve were highlighted as potential early AD biomarkers. In our real-world population with intact cognition, PreADFound achieved 85.11% accuracy to discriminate pre-symptomatic AD based on retinal fundus images alone. Superior and inferior vascular arcades close to the optic nerve were highlighted as potential early AD biomarkers.

In recent years, there has been tremendous advances in the discovery, validation, and clinical implementation of fluid biomarkers for neurodegenerative disorders. These biomarkers are being used to improve the diagnostic workup of individuals presenting with symptoms of cognitive impairment and to facilitate the development of disease-modifying therapies. Assays to measure the core Alzheimer’s disease fluid biomarkers (Aβ42, Aβ40, total Tau, and phosphorylated Tau (pTau) in CSF are widely available and have received regulatory approval in many countries. Assays to identify these biomarkers in blood are emerging. This session will provide an update in the performance and clinical implementation of Alzheimer’s disease blood biomarkers