• Michael Karsy, Joshua R Harmer, Jian Guan, Andrea A Brock, Vijay M Ravindra, Lee S Chung, Aleksander Tkach, Jennifer J Majersik, Min S Park, and Richard H Schmidt.
• Department of Neurosurgery, Clinical Neurosciences Center, University of Utah, Salt Lake City, UT 84132, USA.
• J Clin Neurosci. 2018 Jul 1; 53: 34-40.

AbstractMost patients with cerebral venous sinus thrombosis (CVST) treated with anticoagulation have good outcomes. We examined which factors were associated with poor outcomes after treatment. We retrospectively reviewed patients ≥18 years old who were diagnosed with CVST between 1997 and 2015. Good (modified Rankin score [mRS] ≤2) and poor outcomes were dichotomized. Demographic, historical, clinical, imaging, and treatment characteristics were compared. Eighty-nine patients received treatment for CVST (52.8% males, 74.2% Caucasian). Sixty-eight (76.4%) had good outcomes and 21 (23.6%) had poor outcomes. Poor outcome was associated with systemic or central nervous system (CNS) infection (p = 0.002), lower use of heparin-only therapy than interventional-only treatments (p = 0.003), and increased use of craniectomy (p = 0.002). Good outcomes were associated with migrainous headache on presentation (p = 0.01) and involvement of superficial cortical vessels only (p = 0.02). No prothrombotic or imaging findings correlated with poor outcome. Multivariable analysis showed that any clinical risk factor (p = 0.02) and headache (p = 0.02) predicted improved outcome whereas systemic or CNS infection (p = 0.02) and craniectomy (p = 0.02) predicted poor outcome. A published risk score showed a moderate ability to predict good outcome but not poor outcome. Overall sensitivity (23.8%), specificity (75.0%), and positive (24.0%) and negative (77.0%) predictive value suggested moderate prediction of good outcome and limited prediction of poor outcome. Rates of poor outcomes in CVST were comparable with previous investigations (23.6%), but prediction of poor outcome remains challenging in patients with CVST. Our results suggested that systemic infection and craniectomy were the most robust predictors of poor outcome.Copyright © 2018 Elsevier Ltd. All rights reserved.

26 keywords...

hide…