Abstract TP222: Characteristic Phenotype and Prognostic Factors in Magnetic Resonance Perfusion Weighted Image of Cerebral Venous Thrombosis
Introduction: Cerebral venous thrombosis is a rare form disease. It can be difficult to diagnose and predict patient’s outcome. There are a few reports suggesting characteristic image phenotype and hypothesis of prognosis prediction but those are very rare and still controversial. We sought to identify image phenotype and prognostic factors associated with cerebral venous thrombosis using magnetic resonance perfusion weighted image.
Methods: We retrospectively collected patients who had cerebral venous thrombosis from January 2000 to December 2015. Patients who had underwent magnetic resonance image with perfusion weighted image, non-contrast computerized tomography and magnetic resonance venography were enrolled. Perfusion weighted image was analyzed by visual assessment. We evaluated patient outcome by checking presence of recanalization and disappearance of symptom.
Results: We collected 35 patients who had cerebral venous thrombosis. The mean age of these patients was 42.91 ± 15.47, and 57% were women. Perfusion abnormalities were observed in 25 cases. Cerebral blood volume abnormality was observed in 31 cases, cerebral blood flow abnormality in 31 cases, mean transit time abnormality in 18 cases, and time to peak abnormality in 22 cases. All perfusion abnormalities were localized in the areas adjacent to the occluded sinuses. In the univariate analyses, our study showed location of cerebral venous thrombosis was correlated with good outcome (transverse sinus: p = 0.042, sigmoid sinus: p = 0.026). Gradient echo sequences’ vessel sign (p = 0.094) and cerebral blood flow on perfusion weighted image (p = 0.055) was also related with good outcome. In the multivariate model with adjustment for variables with p<0.10 in the univariate analyses, sigmoid sinus thrombosis (Odd ratio 0.125, 95% confidence interval 0.014-0.756, p = 0.034) and increased cerebral blood flow on perfusion weighted image (Odd ratio 17.749, 95% confidence interval 1.427-544.176, p = 0.044) remained an independent predictor for good outcome.
Conclusions: Our study suggest that cerebral blood flow is able to predict good outcome in cerebral venous thrombosis. It is similar to penumbra phenomenon in ischemic stroke.
Author Disclosures: M. Park: None. W. Seo: None. O. Bang: None. C. Chung: None. K. Lee: None. G. Kim: None.
- © 2017 by American Heart Association, Inc.