Back

Blood-brain barrier dysfunction predicts cognitive trajectory after ischemic stroke

Xue, L.; Jones, O. A.; Drag, L.; Zera, K. A.; Zhu, L.; Mlynash, M.; Carmichael, N. S.; Shu, C.-H.; Biesiada, Z.; Seong, D.; Thomas, O. M.; Simmons, E. C.; Huang, E.; Berry, K.; Chung, P.; Aslan, A.; Xue, R.; Rutledge, J. E.; Oh, H. S.-H.; Yang, L. K.; James, T.; Ghanem, M.; Moran-Losada, P.; Poston, K. L.; Henderson, V. W.; Grainger, J.; Allan, S. M.; Lansberg, M.; Mormino, E. C.; Wyss-Coray, T.; Kapasi, A.; Schneider, J. A.; Smith, C. J.; Parkes, L. M.; Aghaeepour, N.; Buckwalter, M. S.

2026-03-02 neuroscience
10.64898/2026.02.27.708013 bioRxiv
Show abstract

Ischemic stroke doubles the risk of dementia.1-4 Stroke severity and location affect cognition early,5,6 but late dementia risk is not related to infarct characteristics, nor is it reduced by preventing additional strokes,3,6,7 and its mechanism is unknown. We identified a plasma proteomic signature of chronic stroke that was consistent with blood-brain barrier (BBB) dysfunction, including a 58% decrease in plasma levels of platelet-derived growth factor B and downregulation of its pathway compared to healthy controls. During 2 years of follow-up, the stroke-specific proteome was accentuated in stroke survivors who subsequently declined in the processing speed/executive function cognitive domain. To test BBB function, we performed dynamic contrast-enhanced MRI 6-9 months after stroke in an additional cohort and found 1.7-fold higher whole brain BBB leakage compared to controls. Finally, we compared autopsy tissue from people with infarcts and dementia at death to those with infarcts and no dementia. Those who died with dementia had dramatic loss of vascular mural cell coverage compared to those without dementia (median 0.7% vs. 27%). Thus, our proteomic, functional, and structural data implicate chronic BBB dysfunction in cognitive decline late after stroke, revealing potential proteomic and imaging biomarkers and, importantly, a novel target for intervention.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.