Clinical and radiological differences between patients with probable cerebral amyloid angiopathy and mixed cerebral microbleeds

BACKGROUND: The key imaging features of cerebral amyloid angiopathy (CAA) are lobar, cortical, or cortico-subcortical microbleeds, macrohaemorrhages and cortical superficial siderosis (cSS). In contrast, hypertensive angiopathy is characterized by (micro) haemorrhages in the basal ganglia, thalami, periventricular white matter or the brain stem. Another distinct form of haemorrhagic microangiopathy is mixed cerebral microbleeds (mixed CMB) with features of both CAA and hypertensive angiopathy. The distinction between the two entities (CAA and mixed CMB) is clinically relevant because the risk of haemorrhage and stroke should be well balanced if oral anticoagulation is indicated in CAA patients. We aimed to comprehensively compare these two entities. METHODS: Patients with probable CAA according to the modified Boston criteria and mixed CMB without macrohaemorrhage were retrospectively identified from our database. Comprehensive comparison regarding clinical and radiological parameters was performed between the two cohorts. RESULTS: Patients with CAA were older (78 ± 8 vs. 74 ± 9 years, p = 0.036) and had a higher prevalence of cSS (19% vs. 4%, p = 0.027) but a lower prevalence of lacunes (73% vs. 50%, p = 0.018) and deep lacunes (23% vs. 51%, p = 0.0003) compared to patients with mixed CMB. Logistic regression revealed an association between the presence of deep lacunes and mixed CMB. The other collected parameters did not reveal a significant difference between the two groups. CONCLUSIONS: CAA and mixed CMB demonstrate radiological differences in the absence of macrohaemorrhages. However, more clinically available biomarkers are needed to elucidate the contribution of CAA and hypertensive angiopathy in mixed CMB patients. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s00415-020-10038-8) contains supplementary material, which is available to authorized users.