A Clinico-Radiological Study of Cerebral Amyloid Angiopathy-Related Inflammation.

OBJECTIVE: To describe the clinico-radiological features and long-term prognosis in patients with cerebral amyloid angiopathy-related inflammation (CAA-ri). METHODS: Twenty-eight CAA-ri patients were recruited retrospectively from 6 neurological centers. We recorded the clinico-radiological and biological data, at baseline and during follow-up. Baseline characteristics associated with relapse risk and prognosis were assessed. RESULTS: Five patients had pathologically confirmed CAA-ri whereas 23 had probable (n = 21) or possible (n = 2) CAA-ri. The mean age was 72 years; main clinical symptoms included confusion (54%), hemiparesis (36%), and aphasia (29%). Cerebral MRI disclosed a brain parenchymal lesion (89%), which was usually multifocal (82%) and bilateral (89%). It was associated with gadolinium enhancement (84%), small ischemic lesions (39%), cortical superficial siderosis (CSS; 50%), and a high number of microbleeds (mean 240 ± 277). An isolated leptomeningeal involvement was observed in 3 patients with pathological confirmation. Despite a favorable initial evolution after treatment, we observed a 42% risk of relapse, mostly within the first year (83%). After a mean follow-up of 2 years, 29% died and 25% had a marked disability. Disseminated CSS was associated with death. CONCLUSION: Despite an apparently favorable initial evolution, CAA-ri is characterized by a poor prognosis. Diagnostic criteria should consider patients with isolated leptomeningeal involvement.

Cerebral Amyloid Angiopathy and Cerebral Amyloid Angiopathy-Related Inflammation: Comparison of Hemorrhagic and DWI MRI Features.

BACKGROUND: Cerebral amyloid angiopathy (CAA) can be associated with primary vasculitis of small/medium-sized leptomeningeal and cortical arteries, called CAA-related inflammation (CAA-ri). OBJECTIVE: To compare hemorrhagic and diffusion-weighted imaging (DWI) MRI features in CAA and CAA-ri. METHODS: We prospectively scored in a consecutive CAA and CAA-ri cohort: presence/number of chronic intracerebral hemorrhage (ICH), cerebral microbleeds (CMB), and cortical superficial siderosis (CSS) on initial T2*-weighted imaging, and DWI lesions on both initial and follow-up imaging. In a subgroup, ApoE, CSF, and 18F-florbetaben-positron emission tomography (FBB-PET) were also analyzed. RESULTS: In CAA-ri, CMB presence was more frequent (100% versus 40%, p < 0.001) and CMB numbers higher (mean 137 versus 8, p < 0.001). No difference was observed for chronic ICH or CSS. DWI lesions were more frequent in acute compared to chronic CAA-ri (p = 0.025), whereas no such difference was observed between acute and chronic CAA (p = 0.18). Both ApoE4 (genotyping available in 22 CAA-ri and 48 CAA patients) carriers and homozygosity were more frequent in CAA-ri (48% versus 19% [p = 0.014] and 32% versus 2% [p < 0.001] respectively). CSF biomarker analyses (performed in 20 CAA-ri and 45 CAA patients) showed lower Aß42 levels in CAA-ri compared to CAA (median 312 versus 422 pg/mL, p = 0.0032). FBB-PET (performed in 11 CAA-ri and 20 CAA patients) showed higher standardized uptake value ratios in CAA-ri compared with CAA, only significant when the pons was used as reference (p = 0.037). CONCLUSION: Compared to CAA, CAA-ri was associated with higher CMB numbers, more frequent ApoE4 carriers and homozygotes, lower CSF Aß42 levels, and more severe amyloid load on FBB-PET.