Distinct spatial patterns of perivascular spaces enlargement for multiple and Co-existing pathologies of cognitive
Woosik Kim1, Yejin Hwang2, Yelim Yang2
1School of Biomedical Engineering, Korea University, 145, Anam-ro, Seongbuk-gu, Seoul, South Korea; AI Imaging Research Center, NeuroXT, Inc., 48, Achasan-ro 17-gil, Seongdong-gu, Seoul, South Korea.
Background:
This study examines how amyloid-β and vascular pathology independently and jointly relate to regional perivascular space (PVS) burden on T2-weighted MRI.
Methods:
In 307 cognitively impaired participants retrospectively identified from the Seoul National University dementia cohort, PVS were automatically quantified in the basal ganglia (BG) and lobar white matter regions from 2D T2-weighted MRI. Amyloid-β and vascular pathology positivity were defined by [18F]Florbetaben PET and small vessel disease markers. Group differences among pathology-negative (n=36), vascular-only (n=106), amyloid-only (n=48), and mixed (n=117) subgroups, as well as amyloid-vascular interactions, were assessed using analysis of covariance and multivariable linear regression.
Results:
BG PVS were greater in participants with vascular burden than in pathology-negative participants (F=26.97, p<0.001; Cohen's d=1.28), independent of amyloid-β. Lobar PVS were higher in single-pathology than pathology-negative participants across the parietal, temporal, and occipital regions (F=8.25-18.04, Cohen's d=0.66-0.98, all p≤0.014), with no additional increase in the mixed group. Greater amyloid-β retention was associated with parietal, temporal, and occipital PVS in VB- participants (β [95% CI]=0.55 [0.19, 0.92], 0.69 [0.35, 1.02], 0.40 [0.16, 0.64], respectively). Among AB- participants, vascular burden was associated with BG PVS (β [95% CI]=0.65 [0.40, 0.89]). Multivariable regression demonstrated less-than-additive AB×VB interactions in parietal, temporal, and occipital PVS (β [95% CI]=-0.65 [-1.08, -0.23], -0.74 [-1.13, -0.35], -0.42 [-0.70, -0.14], respectively).
Conclusions:
Distinct regional PVS patterns reflect spatially selective and severity-dependent glymphatic-related structural alterations associated with amyloid-β and vascular pathologies, supporting PVS as a quantitative imaging biomarker for disentangling mixed pathways of cognitive impairment.
Insights
Vascular pathology increases basal ganglia perivascular spaces (PVS), while amyloid-β and vascular issues affect lobar PVS differently. These findings highlight PVS as a biomarker for cognitive impairment pathways.
Area of Science:
- Neuroimaging
- Neuropathology
- Biomarker Discovery
Background:
- Cognitive impairment is often associated with amyloid-β and vascular pathologies.
- Perivascular spaces (PVS) on MRI may reflect glymphatic system dysfunction.
- Understanding the interplay of these pathologies and PVS is crucial for diagnosis.
Purpose of the Study:
- To investigate the independent and combined effects of amyloid-β and vascular pathology on regional PVS burden.
- To determine if PVS patterns differ based on the presence and combination of amyloid-β and vascular pathologies.
- To assess PVS as a potential imaging biomarker in cognitively impaired individuals.
Main Methods:
- Retrospective analysis of 307 cognitively impaired participants.
- Automatic quantification of PVS in basal ganglia and lobar white matter using T2-weighted MRI.
- Classification of participants based on amyloid-β and vascular pathology status using PET and small vessel disease markers.
- Statistical analysis using ANCOVA and multivariable linear regression to assess group differences and interactions.
Main Results:
- Vascular pathology was associated with increased basal ganglia PVS, independent of amyloid-β.
- Lobar PVS were elevated in participants with single pathologies compared to those with no pathology.
- Amyloid-β retention correlated with lobar PVS in vascular pathology-negative individuals, while vascular burden correlated with basal ganglia PVS in amyloid-β-negative individuals.
- Less-than-additive interactions between amyloid-β and vascular pathology were observed for lobar PVS.
Conclusions:
- Regional PVS patterns are distinct and reflect spatially selective glymphatic alterations.
- PVS burden is associated with amyloid-β and vascular pathologies in a severity-dependent manner.
- Perivascular spaces serve as a quantitative imaging biomarker for differentiating mixed pathways contributing to cognitive impairment.
Related Concept Videos
Dementia l: Introduction
Cerebral Edema ll: Pathophysiology
Alzheimer Disease ll: Pathophysiology
Cerebral Edema l: Introduction
Imaging Studies VII: Vascular Imaging
Increased Intracranial Pressure ll: Pathophysiology


