Advertisement

CEREBRAL MICROBLEEDS ARE ASSOCIATED WITH CEREBRAL BLOOD FLOW AND METABOLISM BUT NOT AMYLOID BURDEN OR BRAIN ATROPHY IN COGNITIVELY NORMAL ELDERLY

      Background: Cerebral microbleeds (CMBs) are commonly seen in magnetic resonance (MR) images of elderly individuals and are related to small vessel disease from cerebral amyloid angiopathy and arteriosclerosis/lipohyalinosis. We hypothesized that CMBs are associated with decreased cerebral blood flow (CBF), decreased cerebral metabolism, and cognitive deficits in very elderly individuals.
      Methods: Fifty-five normal controls (NC) (86.8 ± 2.7 years; 22 female) were recruited from the Pittsburgh Ginkgo Evaluation of Memory Study (GEMS). MR imaging for each subject was performed on a Siemens 3T Trio scanner with a Siemens GRE or syngo SWI protocol for CMB identification, structural MPRAGE sequence for hippocampal volume and region-of-interest determination, and an arterial spin labeling (ASL) sequence for CBF measures. PET imaging was performed using a Siemens/CTI ECAT HR+ PET scanner and [11 C]PiB and [18 F]FDG to determine Aβ burden and metabolism, respectively. A single reader performed all CMB identifications, designating subjects as CMB(-) (no CMB found) or CMB(+) (one or more CMBs found). Hippocampal volumes were computed using FSL software (FIRST algorithm). Regional sampling of the ASL, [11 C]PiB, and [18 F]FDG images yielded regional values of ASL CBF (mL/100g/min), [11 C]PiB Aβ burden (SUVR), and [18 F]FDG metabolism (SUVR). SUVR values were computed using cerebellum as reference. Cognitive and functional status was assessed with a battery of neuropsychiatric tests and the Clinical Dementia Rating (CDR) scale.
      Results: CMBs were found in 21 of the 55 subjects. Both the presence and number of CMBs showed significant associations with reduced regional CBF (p<0.05). The number of CMBs in CMB(+) subjects also correlated significantly with reduced metabolism in the medial temporal lobe (p<0.05). Subjects with CMBs were significantly more likely to have deficits on the Clinical Dementia Rating scale (χ 2 p-value = 0.0015). Neither the presence nor number of CMBs was found to be significantly associated with [11 C]PiB SUVR or hippocampal volume.
      Conclusions: In this cohort of very elderly NCs, CMBs are associated with reduced cerebral blood flow, reduced cerebral metabolism, and mild deficits on the CDR scale. Our findings suggest that the small vessel disease associated with CMBs can independently contribute to impaired cognition and functional status, a process likely mediated by chronic hypoperfusion and hypometabolism.