Romain Bourcier1, Romain Thiaudière2, Laurence Legrand3, Benjamin Daumas-Duport2, Hubert Desal2, Grégoire Boulouis3. 1. Department of Diagnostic and Interventional Neuroradiology, Guillaume et René Laennec University Hospital, Nantes, France, romain.bourcier2@gmail.com. 2. Department of Diagnostic and Interventional Neuroradiology, Guillaume et René Laennec University Hospital, Nantes, France. 3. Department of Neuroradiology, Université Paris-Descartes, Sainte-Anne Hospital, Paris, France.
Abstract
BACKGROUND: Fluid attenuated inversion recovery (FLAIR) vascular hyperintensity (FVH) document slowed vascular flow at the level and after the occlusion site patients with acute ischemic stroke (AIS). We aimed to assess the accuracy of FVH for the confirmation and location of a large vessel occlusion (LVO). METHODS: Three radiologists reviewed the FLAIR sequence of the admission MRI exam of patients with suspected AIS at a single academic center. Readers were provided with the main clinical deficit with National Institute of Health Stroke Severity score and were asked to identify and locate an LVO when appropriate. Kappa coefficients were calculated for agreement along with diagnosis performances of FVH to recognize and locate an LVO with digital subtracted angiography (DSA) as gold standard. RESULTS: Among 125 patients screened with MRI for a suspected AIS, 96 (81%) were diagnosed with AIS and 47 (38%) patients had an anterior LVO of whom 25 (20%) had a DSA for mechanical thrombectomy. Kappa coefficients for intra- and inter-readers were good to excellent. Overall, the sensitivity and the specificity of the FVH to predict an anterior LVO was 0.98 (95% confidence interval [CI]: 0.94-1) and 0.86 (95% CI: 0.79-0.96), respectively, while PPV and NPV were 0.87 (95% CI: 0.85-0.95) and 0.98 (0.97-1), respectively. FVH also showed good to excellent accuracy for identifying M1 and M2 versus internal carotid artery occlusion site. CONCLUSION: We found that FVH demonstrated excellent diagnostic performances for the identification of LVO and its level with good to excellent reproducibility. This MRI radio marker of occlusion provides additional arguments and may speed-up the detection of potential candidates for MT.
BACKGROUND: Fluid attenuated inversion recovery (FLAIR) vascular hyperintensity (FVH) document slowed vascular flow at the level and after the occlusion site patients with acute ischemic stroke (AIS). We aimed to assess the accuracy of FVH for the confirmation and location of a large vessel occlusion (LVO). METHODS: Three radiologists reviewed the FLAIR sequence of the admission MRI exam of patients with suspected AIS at a single academic center. Readers were provided with the main clinical deficit with National Institute of Health Stroke Severity score and were asked to identify and locate an LVO when appropriate. Kappa coefficients were calculated for agreement along with diagnosis performances of FVH to recognize and locate an LVO with digital subtracted angiography (DSA) as gold standard. RESULTS: Among 125 patients screened with MRI for a suspected AIS, 96 (81%) were diagnosed with AIS and 47 (38%) patients had an anterior LVO of whom 25 (20%) had a DSA for mechanical thrombectomy. Kappa coefficients for intra- and inter-readers were good to excellent. Overall, the sensitivity and the specificity of the FVH to predict an anterior LVO was 0.98 (95% confidence interval [CI]: 0.94-1) and 0.86 (95% CI: 0.79-0.96), respectively, while PPV and NPV were 0.87 (95% CI: 0.85-0.95) and 0.98 (0.97-1), respectively. FVH also showed good to excellent accuracy for identifying M1 and M2 versus internal carotid artery occlusion site. CONCLUSION: We found that FVH demonstrated excellent diagnostic performances for the identification of LVO and its level with good to excellent reproducibility. This MRI radio marker of occlusion provides additional arguments and may speed-up the detection of potential candidates for MT.