Perfusion CT for prediction of hemorrhagic transformation in acute ischemic stroke: a systematic review and meta-analysis

Autor: Kyung Won Kim, Chong Hyun Suh, Woo Yong Oh, Dong-Cheol Woo, Donghyun Kim, Jung Bin Lee, Se Jin Cho, Jong Gu Lee, Seung Chai Jung
Rok vydání: 2019
Předmět:
Zdroj: European Radiology. 29:4077-4087
ISSN: 1432-1084
0938-7994
Popis: To investigate the diagnostic performance of perfusion CT for prediction of hemorrhagic transformation in acute ischemic stroke. A computerized literature search of Ovid MEDLINE and EMBASE was conducted up to October 29, 2018. Search terms included acute ischemic stroke, hemorrhagic transformation, and perfusion CT. Studies assessing the diagnostic performance of perfusion CT for prediction of hemorrhagic transformation in acute ischemic stroke were included. Two reviewers independently evaluated the eligibility of the studies. A bivariate random effects model was used to calculate the pooled sensitivity and pooled specificity. Multiple subgroup analyses were performed. Fifteen original articles with a total of 1134 patients were included. High blood-brain barrier permeability and hypoperfusion status derived from perfusion CT are associated with hemorrhagic transformation. The pooled sensitivity and specificity were 84% (95% CI, 71–91%) and 74% (95% CI, 67–81%), respectively. The area under the hierarchical summary receiver operating characteristic curve was 0.84 (95% CI, 0.81–0.87). The Higgins I2 statistic demonstrated that heterogeneity was present in the sensitivity (I2 = 80.21%) and specificity (I2 = 85.94%). Although various perfusion CT parameters have been used across studies, the current evidence supports the use of perfusion CT to predict hemorrhagic transformation in acute ischemic stroke. • High blood-brain barrier permeability and hypoperfusion status derived from perfusion CT were associated with hemorrhagic transformation. • Perfusion CT has moderate diagnostic performance for the prediction of hemorrhagic transformation in acute ischemic stroke. • The pooled sensitivity was 84%, and the pooled specificity was 74%.
Databáze: OpenAIRE