# Imaging and fluid biomarkers for prognostic stratification in cerebral amyloid angiopathy

## TL;DR

This review discusses how imaging and fluid biomarkers can help predict outcomes in cerebral amyloid angiopathy, a brain disease linked to bleeding and cognitive decline.

## Contribution

The paper highlights recent advances in multi-modal biomarkers for individualized prognosis in cerebral amyloid angiopathy.

## Key findings

- Non-hemorrhagic imaging markers like WMHs and ePVS provide sensitive measures of microstructural damage in CAA.
- Fluid biomarkers such as NfL and GFAP reflect vascular and neuronal injury in CAA patients.
- Combining imaging and fluid biomarkers may improve prediction of hemorrhagic and cognitive outcomes.

## Abstract

Cerebral amyloid angiopathy (CAA) is a common small vessel disease characterized by Aβ deposition in cortical and leptomeningeal arteries, leading to lobar intracerebral hemorrhage and vascular cognitive impairment. Despite advances in diagnosis, prognosis remains highly heterogeneous, encompassing risks of recurrent hemorrhage and progressive cognitive decline. This review summarizes recent developments in imaging and fluid biomarkers for prognostic stratification in CAA. Imaging markers, including advanced MRI and molecular PET techniques, have evolved from traditional hemorrhagic indicators, such as cerebral micro-bleeds (CMBs) and cortical superficial siderosis (cSS), to non-hemorrhagic including white matter hyper-intensities (WMHs), and enlarged perivascular spaces (ePVS), which sensitively capture microstructural damage after using quantitative measures. Fluid biomarkers provide dynamic insights into vascular and neuronal injury, including altered plasma Aβ42/Aβ40 ratios, MMPs/TIMPs balance, and elevated neuro-filament light chain (NfL) and glial fibrillary acidic protein (GFAP) levels. Integrating these multi-modal indicators may enable individualized prediction of hemorrhagic and cognitive outcomes and inform precision management strategies. Future research should standardize quantification methods and validate multi-modal models across diverse CAA phenotypes to advance toward personalized prognostic frameworks.