Cerebral Amyloid Angiopathy and Risk of Dementia in Patients With Cognitive Complaint.

BACKGROUND AND

OBJECTIVES

The frequency and cognitive trajectory of cerebral amyloid angiopathy (CAA) in patients from memory clinics is uncertain. We aimed to determine whether patients with CAA have an increased risk of dementia in nondemented individuals presenting with a cognitive complaint.

METHODS

We retrospectively analyzed data of the MEMENTO prospective cohort that enrolled outpatients from 26 centers in France presenting with a cognitive complaint and a Clinical Dementia Rating (CDR) scale score ≤0.5. Participants aged >50 years who had interpretable baseline brain MRI were eligible for this study and followed every 6 months for 5 years with systematic assessment of dementia. Based on MRI analysis, participants were classified into 4 categories using the Boston criteria V2.0 and V1.5: probable CAA, possible CAA, deep/mixed small vessel disease (SVD), or controls (without imaging markers of SVD). The primary outcome was the progression to dementia (Diagnostic and Statistical Manual of Mental Disorders, Fourth Revision criteria) evaluated by an expert committee. Cox proportional hazards models were used to determine hazard ratios (HRs) and CIs for the primary outcome.

RESULTS

Of 2,323 MEMENTO participants, 2,136 (92%) were included in this study. The mean age (SD) at baseline was 71.3 (7.9) years and 1,320 (62%) were women. Probable CAA was diagnosed in 413 (19.3%) patients using the Boston criteria V2.0, and 144 (7.0%) using the Boston criteria V1.5. During a median follow-up of 5.0 years (interquartile range 3.1-5.1), 307 participants developed dementia. Compared with controls, patients with probable CAA according to the Boston criteria V2.0 (HR 1.73, 95% CI 1.23-2.42) and V1.5 (HR 2.23, 95% CI 1.56-3.20) had increased risk of dementia. After adjusting for age, sex, baseline Mini-Mental State Examination score and hippocampal volume, patients with probable CAA per Boston criteria V1.5 but not V2.0 had an increased risk of dementia compared with controls. This association was not significant when further adjusting with APOE ε4.

DISCUSSION: CAA is a common condition in outpatients with mild cognitive symptoms. Patients with probable CAA according to the Boston criteria V1.5 but not V2.0 have increased risk of dementia. Further studies are required to externally validate these findings and evaluate potential mediation effect of APOE genotype.