Clinical relevance of improved microbleed detection by susceptibility-weighted magnetic resonance imaging.

BACKGROUND AND PURPOSE Susceptibility-weighted imaging (SWI) has been shown to be more sensitive in detecting cerebral microbleeds (MBs) than is conventional T2*-weighted gradient-recalled echo imaging (GRE). However, the clinical relevance of this improved detection in terms of associations with clinical measures and risk factors is unclear. We sought to determine whether associations of MBs with clinical characteristics, risk factors, white-matter hyperintensities, and lacunes were different on GRE versus SWI in memory clinic patients. METHODS One hundred forty-one patients presenting at our memory clinic were included and underwent clinical evaluation and a magnetic resonance imaging protocol including both GRE and SWI. Images were analyzed for numbers and locations of MBs and white-matter hyperintensities. In a subset of patients, apolipoprotein E status was determined. Negative binomial regression was used to assess clinical and radiologic associations with MB number. RESULTS MB prevalence was 23% on GRE and 40% on SWI. A total of 219 and 284 MBs were detected on GRE and SWI, respectively. Within groups with MBs, the median MB count was 1 (range, 1 to 144) on GRE and 2 (range, 1 to 129) on SWI (P<0.001). The increase in the number of MBs on SWI was equally distributed among brain regions. Strengths of the associations with age, sex, white-matter hyperintensities, and presence of lacunes with higher MB numbers were comparable for GRE and SWI (all P<0.05); no differential independent associations were detected. CONCLUSIONS SWI detected more MBs in more patients, irrespective of MB location. However, this enhanced detection had no improved clinical relevance in terms of associations with vascular risk factors or radiologic markers of small-vessel disease.