Association Between Cerebral Cortical Microinfarcts and Perilesional Cortical Atrophy on 3T MRI

Background and objectives: Cerebral cortical microinfarcts (CMIs) are a novel MRI marker of cerebrovascular disease (CeVD) that predicts accelerated cognitive decline. Presence of CMIs is known to be associated with global cortical atrophy, although the mechanism linking the two is unclear. Our primary objective was to examine the relation between CMIs and cortical atrophy and to establish possible perilesional atrophy surrounding CMIs. Our secondary objective was to examine the role of cortical atrophy in CMI-associated cognitive impairment.

Methods: Patients were recruited from 2 Singapore memory clinics between December 2010 and September 2013 and included if they received the diagnosis no objective cognitive impairment, cognitive impairment (with or without a history of stroke), or Alzheimer or vascular dementia. Cortical thickness, chronic CMIs, and MRI markers of CeVD were assessed on 3T MRI. Patients underwent cognitive testing. Cortical thickness was compared globally between patients with and without CMIs, regionally within individual patients with CMIs comparing brain regions with CMIs to the corresponding contralateral region without CMIs, and locally within individuals patients in a 50-mm radius of CMIs. Global cortical thickness was analyzed as mediator in the relation between CMI and cognitive performance.

Results: Of the 238 patients (mean age 72.5 years, SD 9.1 years) enrolled, 75 had ≥1 CMIs. Patient with CMIs had a 2.1% lower global cortical thickness (B = -0.049 mm, 95% confidence interval [CI] 0.091 to -0.007, p = 0.022) compared to patients without CMIs, after correction for age, sex, education, and intracranial volume. In patients with CMIs, cortical thickness in brain regions with CMIs was 2.2% lower than in contralateral regions without CMIs (B = -0.048 mm [95% CI -0.071 to -0.026], p < 0.001). In a 20-mm radius area surrounding the CMI core, cortical thickness was lower than in the area 20 to 50 mm from the CMI core (mean difference -0.06 mm [-0.10 to -0.02], p = 0.002). Global cortical thickness was a significant mediator in the relationship between CMI presence and cognitive performance as measure with the Mini-Mental State Examination (B = -0.12 [-0.22 to -0.01], p = 0.025).

Discussion: We found cortical atrophy surrounding CMIs, suggesting a perilesional effect in a cortical area many times larger than the CMI core. Our findings support the notion that CMIs affect brain structure beyond the actual lesion site.