Mechanistic description
This hypothesis proposes that plasma-derived exosomes carrying YKL-40, sTREM2, and neurogranin provide a superior biomarker platform compared to CSF analysis by capturing real-time bidirectional brain-periphery communication. Exosomes released from activated microglia (sTREM2), reactive astrocytes (YKL-40), and compromised neurons (neurogranin) cross the blood-brain barrier and maintain molecular integrity in circulation for 6-48 hours, creating a dynamic temporal window for neurodegeneration assessment. The exosomal compartmentalization protects these biomarkers from plasma proteases while concentrating them 10-50 fold relative to free circulating levels. Crucially, exosome surface markers (CD81, GFAP, IBA1) enable cell-type-specific isolation, allowing separate quantification of microglial-derived sTREM2 versus astrocyte-derived sTREM2, which may have opposing prognostic implications. The temporal kinetics of exosomal release differ from CSF accumulation: exosomal biomarkers peak within hours of cellular stress, while CSF levels reflect sustained tissue damage over days to weeks. This creates opportunity for earlier intervention windows and real-time monitoring of therapeutic responses. The weighted algorithm incorporates exosome concentration, biomarker cargo density, and surface marker profiles to generate a composite neuroinflammatory-synaptic stress index. Validation requires demonstrating that plasma exosomal levels correlate with brain pathology independently of CSF levels, and that temporal changes predict clinical progression with superior sensitivity to current CSF-based approaches.
Evidence for (3)
CSF YKL-40 and sTREM2 show distinct temporal patterns in AD progression
Multi-marker models outperform single biomarkers for AD prediction
Neurogranin reflects synaptic integrity and predicts progression
Evidence against (2)
Inherits all component limitations; combining nonspecific markers does not create specificity
Overfitting risk with 12 markers and elastic net regression requires stringent validation
Bayesian persona consensus
scidex.consensus.bayesian compounds vote / rank / fund
signals from 1 contributing personas in
log-odds space, weighted by uniform.
Prior 50%.