Counting cavitating lacunes underestimates the burden of lacunar infarction

Background & Purpose: On brain imaging, lacunes, or cerebrospinal fluid (CSF)-containing cavities, are common and are often counted in epidemiological studies as old lacunar infarcts. The proportion of symptomatic lacunar infarcts which progress to lacunes is unknown. Non-cavitating lacunar infarcts may continue to resemble white matter lesions (WML). Methods: We identified patients with acute lacunar stroke, with or without an acute lacunar infarct on computed tomography (CT) or magnetic resonance (MR) imaging, who had follow-up imaging. A neuroradiologist classified lacunar infarcts progressing to definite or possible cavities on follow-up imaging. We tested associations between cavitation and patient-related, stroke-related and imaging-related features, including other features of small vessel disease. Results: Amongst 90 patients (mean age 67 years), any cavitation was present on follow-up imaging in 25 (28%) and definite cavitation in 18 (20%). Definite cavitation was associated with increasing time to follow-up imaging (median 228 days, range 54-1722, vs. no cavitation 72 days, range 6-1440, p=0.0003) and deep cerebral atrophy (p=0.03), but not with age, stroke severity, larger initial infarct size or other features of small vessel disease Hypertension and diabetes were negatively associated with cavitation (p=0.01 and 0.02, respectively). Conclusions: Definite cavitation occurs in 1/5 of symptomatic lacunar ischemic strokes, implying that most continue to resemble WML. Epidemiology and pathophysiology studies of lacunar stroke which have only counted lacunes as lacunar infarcts may have substantially underestimated, by as much as five times, the true burden of lacunar stroke disease.