Cerebrospinal Fluid Aβ Biomarkers Predict Recurrent Hemorrhage and Identify Cerebral Amyloid Angiopathy in Patients With Lobar Hemorrhage

Abstract

Background: Recurrent hemorrhage represents a significant risk for patients with lobar hemorrhage and underlying cerebral amyloid angiopathy. However, it remains unknown, whether cerebrospinal fluid biomarkers of β-amyloid (Aβ) retention, predict recurrent hemorrhagic events in these patients.

Methods: In this retrospective study, we evaluated patients with first-ever lobar intracerebral hemorrhage or convexity subarachnoid hemorrhage who underwent cerebrospinal fluid analysis of Aβ₄₀, Aβ₄₂, and Aβ_42/40 ratio. Biomarker levels were compared between patients with and without recurrent hemorrhage, and optimal cutoff values were derived from receiver operating characteristic analysis to define high and low biomarker groups. Negative binomial regression models, adjusted for age, sex, and hypertension, were used to assess associations with recurrent hemorrhage rates, and Kaplan-Meier survival analysis evaluated time to first event as a sensitivity measure.

Results: Among 289 patients with lobar hemorrhage, 48 were eligible for analysis (mean age, 72.4 years; 44% women; median follow-up, 2.7 years). Patients with recurrent hemorrhage had significantly lower Aβ₄₀ and Aβ₄₂ levels (P<0.05). Those with low Aβ₄₂ levels exhibited a recurrence rate of 18.2 versus 1.7 events per 100 patient-years (incidence rate ratio, 12.7 [95% CI, 1.5-306.6]; P=0.035; sensitivity, 89%; specificity, 54%), while low Aβ₄₀ levels were associated with 63.6 versus 4.5 events per 100 patient-years (incidence rate ratio, 41.0 [95% CI, 5.8-434.4]; P<0.001; sensitivity, 44%; specificity, 95%). Combining probable cerebral amyloid angiopathy (Boston criteria version 1.5) with low Aβ₄₀ levels improved risk stratification, yielding a rule-in specificity of 75% and a rule-out sensitivity of 100%. Additionally, the Aβ_42/40 ratio demonstrated robust accuracy for identifying probable cerebral amyloid angiopathy (sensitivity, 63%; specificity, 83%).

Conclusions: Cerebrospinal fluid Aβ biomarkers are effective in predicting recurrent hemorrhage and identifying probable cerebral amyloid angiopathy in patients with lobar hemorrhage. These findings underscore the potential of disease-specific biofluid markers to improve clinical risk stratification and guide management strategies.

Keywords: amyloid β; aβ positivity; cerebral amyloid angiopathy; cerebrospinal fluid; convexity subarachnoid hemorrhage; lobar intracerebral hemorrhage; mixed intracerebral hemorrhage.

Figure 1. Study flowchart.

A total of 48 patients with lobar ICH or nonaneurysmal convexity SAH and cerebrospinal fluid Aβ biomarkers were included in the primary outcome analyses. Outcomes were the utility in predicting recurrent hemorrhage during follow‐up and the diagnostic accuracy for identifying patients with probable cerebral amyloid angiopathy according to MRI. Aβ indicates β‐amyloid; CSVD, cerebral small vessel disease; ICH, intracerebral hemorrhage; MRI, magnetic resonance imaging; and SAH, subarachnoid hemorrhage.

Figure 2. Time to first recurrent hemorrhage and stroke during follow‐up.

Kaplan–Meier curves showing the relative number patients without recurrent hemorrhage (A, C) and without recurrent stroke (B, D) according to their CSF Aβ₄₂ and Aβ₄₀ levels. Aβ indicates β‐amyloid.