Significance of simulated conventional images on dual energy CT after endovascular treatment for ischemic stroke. (22nd January 2019)

• Record Type: Journal Article
• Authors: Ebashi, Ryo; Ogata, Atsushi; Nishihara, Masashi; Inoue, Kohei; Yoshioka, Fumitaka; Takase, Yukinori; Masuoka, Jun; Yakushiji, Yusuke; Irie, Hiroyuki; Hara, Hideo; Abe, Tatsuya
• Is Part Of: Journal of neurointerventional surgery. Volume 11:Number 9(2019)
• Page Start: 898
• Page End: 902
• Publication Date: 2019-01-22
• Subjects: thrombectomy -- CT; Nervous system -- Surgery -- Periodicals; Cerebrovascular disease -- Surgery -- Periodicals; 617.48
• DOI: 10.1136/neurintsurg-2018-014486
• Languages: English
• ISSNs: 1759-8478

Abstract

Abstract : Background and purpose: CT scans often reveal post-interventional cerebral hyperdensities (PCHDs) immediately after intra-arterial thrombectomy (IAT) for ischemic stroke. Dual energy CT (DE-CT) can indicate whether PCHDs are caused by hemorrhage or iodinated contrast. Hyperdense lesions, detected on DE-CT with the use of iodinated contrast, could be associated with delayed hemorrhagic transformation and poor outcome. However, the quantitative indicators in DE-CT for predicting delayed hemorrhagic transformation remain unclear. We assessed such indicators for predicting delayed hemorrhagic transformation. Material and methods: We retrospectively analyzed 52 consecutive acute ischemic stroke patients who underwent IAT. Simulated conventional CT (sCCT) images were obtained immediately after a DE-CT scan. Virtual, unenhanced, non-contrast (VNC) imaging was performed after reconstruction. Hounsfield units (HU) of the infarct areas observed on the sCCT were measured. The association of HU on sCCT with hemorrhage on VNC and delayed parenchymal hemorrhage (PH) was evaluated. Results: The HU of sCCT with hemorrhage on VNC was significantly higher than without it (377.9±385 HU vs 83.5±37.9 HU; P<0.0001). The cut-off index was 80 HU, which displayed 100% sensitivity, 63.8% specificity, 22.3% positive predictive value, and 100% negative predictive value (P=0.0001, area under the curve (AUC)=0.89). The HU with delayed PH was substantially higher than without it (250.8±382.2 HU vs 93.7±64.8 HU; P=0.01). The cut-off index was 78 HU, which showed 100% sensitivity, 61% specificity, 25% positive predictive value, and 100% negative predictive value (P=0.049, AUC=0.76). Conclusion: sCCT images on DE-CT are useful for excluding intracerebral hemorrhage and delayed PH.