Detection of chronic cerebellar and cerebral cortical microinfarctions on T2 and FLAIR 1.5T MRI

Background: Detection rate of chronic cerebral cortical microinfarctions (CMI) is low. Few data are available on chronic cerebellar CMI. Our aim was to study the ability of 1.5T FLAIR and T2-weighted imaging to detect chronic CMI and compare cerebral and cerebellar CMI location. Methods: We prospectively included stroke patients with ≥1 acute cerebral or cerebellar CMI ≤10mm on DWI. Follow-up FLAIR and T2-weighted imaging with the same 1.5T-magnet was performed after six months. Images were evaluated unblinded to initial MRI to detect signal changes at the same location as initial acute CMI. Variables influencing chronic CMI detection were analyzed in uni- and multivariate analyses. Results: We analyzed 57 patients (median age 65, 44% women) with 111 acute CMI on initial MRI. These constituted 47 cerebellar and 64 cerebral acute CMI, with median DWI size of 5.9mm and initial FLAIR positivity of 70%, without difference between cerebellar and cerebral location. On 6-month follow-up MRI, chronic CMI detection on FLAIR and T2 was similar for cerebral lesions (p=0.81), whereas for cerebellar lesions T2 showed a 2.4-fold higher detection rate than FLAIR (p<0.0001). Univariate analysis using a logistic mixed model for T2 detection of chronic CMI was associated with cerebellar CMI location (p<0.0001) and hypertension history (p=0.008). Multivariate analysis confirmed the strong association between cerebellar location and T2 detection of chronic CMI (p<0.0001, AOR=7.25, 95% CI=2.83-18.56) adjusted for hypertension. Conclusion: Compared to FLAIR, T2-weighted imaging shows highest detection rates for cerebellar CMI, with cerebellar CMI location as strongest predictor of chronic CMI detection.