Cerebral microbleeds and cortical superficial siderosis in cerebral amyloid angiopathy.

Abstract

BACKGROUND: Cerebral amyloid angiopathy (CAA), an important cause of lobar intracerebral hemorrhage in the elderly, has other clinico-radiological manifestations. Susceptibility-weighted MRI (SWI) can detect other lesions characteristic of CAA such as lobar cerebral microbleeds (CMBs) and cortical superficial siderosis (cSS). These lesions can be utilized by the Modified Boston Criteria to make a radiological diagnosis of CAA. While CMBs and cSS are known markers of cortical and leptomeningeal CAA respectively, their relationship with each other and distribution across cerebral hemispheres is not known. METHODS: Retrospective clinical and radiological brain SWI data was analysed from a five-year period at Epworth Hosptical Richmond in patients who had known CAA, a haemorrhage potentially caused by CAA, or a presentation with clinical features mimicking CAA. Patients were then included if they met the Mofidied Boston Criteria for 'probable CAA', and clinical and radiological data was collected. 'Asymmetry' was defined as a cerebral hemisphere having two-thirds or more of the total CMB or cSS burden. RESULTS: From 734 patients reviewed, 59 patients met the inclusion criteria. Scatterplot analysis of lobar CMBs against cSS revealed an R² value of 0.007, suggesting that no relation between CMBs and cSS. In patients with lobar CMBs (n = 58), the distribution was asymmetrical in 40 patients (68.9%), including 19 patients (32.8%) who had lobar CMBs in only one hemisphere. In patients with cSS (n = 27), the distribution was asymmetrical in 18 patients (66.7%), including 13 patients (48.1%) who had cSS in only one hemisphere. Exploratory analysis of patients with two SWI MRIs (n = 6) found that all patients demonstrated an increase in CMBs and cSS between scans, including instances of new lesions appearing close to older lesions. CONCLUSION: CAA may not progress uniformly or symmetrically though the brain, and there may be differences between cortical and leptomeningeal CAA.