Abstract 123: Correlation of Cerebral Microbleeds Distribution to Amyloid Burden in Patients With Spontaneous Intracerebral Hemorrhage

Abstract

Background and Purpose: The diagnosis of cerebral amyloid angiopathy (CAA) mainly depends on clinical estimation such as the location of ICH or cerebral microbleeds (CMBs). We aimed to use [11C]PiB PET to image cerebral β-amyloid deposition and validate the CAA diagnosis in patients with ICH. Methods: Twenty six patients with spontaneous ICH received cerebral [11C]PiB PET to measure the amyloid burden and 3-T magnetic resonance susceptibility weighted imaging (SWI) to analyze the distribution of CMBs. Seven patients with Alzheimer’s disease (AD) and 7 healthy subjects were included for a comparison. The lobar amyloid burden was expressed as average standardized uptake value (SUV) using the commercial software package PMOD. The number of CMBs was counted using the Microbleed Anatomical Rating Scale (MARS). Results: The amyloid burden in patients with ICH (0.72±0.27) was significantly higher than that of heathy controls (0.47±0.17, p=0.03), but lower than that of AD patients (1.31±0.38, p=0.001). The amyloid burden was not significantly different between the patients with cortico-subcortical (frontal, temporal, parietal and occipital lobes) ICH and those with deep (putamen, thalamus, caudate and cerebellum) ICH (0.77±0.32 vs. 0.67±0.20, p=0.66). CMBs were present in 23 of 26 patients (88.5%). The amyloid burden was higher in patients with CMBs located predominantly at the cortico-subcortical areas (n=8) than those with CMBs located predominantly at the deep regions (n=9) (0.96±0.30 vs. 0.54±0.13, p=0.009). The ratio of cortico-subcortical CMB number to deep CMB number was positively correlated with the amyloid burden (γ=0.65, p=0.004). Conclusion: The distribution of CMBs is a better indicator than the ICH location to predict cerebral β-amyloid deposition in patients with ICH.