Prevalence of Venous Infarction in Patients With Cerebral Venous Thrombosis: Baseline Diffusion-Weighted MRI and Follow-Up MRI.

Abstract

A venous pattern of infarction on neuroimaging is used as a clue to undiagnosed cerebral venous thrombosis (CVT); prevention of venous infarction is a goal of CVT management; and venous infarction is a factor used for clinical prognostication. Despite widespread use of the term venous infarct, the prevalence of true venous infarction is unclear. Our primary aim was to determine the prevalence of venous infarction in patients with CVT. We also measured the prevalence of diffusion abnormality without infarction, vasogenic edema, and intracranial hemorrhage. Single-center, retrospective cohort study using a registry of 110 consecutive patients admitted to hospital with cerebral venous thrombosis between 2004 and 2014. Inclusion criteria were brain magnetic resonance imaging (MRI) and contrast-enhanced venography at presentation, and repeat brain MRI ≥1 month later. Exclusion criteria were dural arteriovenous fistula, arteriovenous malformation, cavernous sinus thrombosis, or previous neurosurgical procedure. Main outcome was proportion of patients with venous infarction (irreversible ischemic injury) diagnosed using diffusion-weighted MRI at presentation, confirmed using T2-weighted fluid-attenuated inversion recovery MRI ≥1 month later, and reported with 95% CI using the Wilson score interval method. We also report the prevalence of transient diffusion MRI abnormality without infarction, vasogenic edema, and intracranial hemorrhage. Seventy-three patients met the inclusion criteria, and after exclusions, the final study population was 59 patients with median age 41 years (interquartile range, 32-57). Venous infarction occurred in 12% (7/59 [95% CI, 6%-23%]) of patients, and final infarct volume was >1 mL in only 5.1% (3/59) of patients. An additional 8% (5/59 [95% CI, 4%-18%]) of patients had a transient diffusion MRI abnormality without infarction. Prevalence of cerebral vasogenic edema and intracranial hemorrhage were 66% (39/59 [95% CI, 53%-77%]) and 54% (32/59 [95% CI, 41%-66%]), respectively. In patients with CVT, venous infarction is uncommon and venous infarcts are typically very small. Vasogenic edema and hemorrhage are more common consequences of CVT.