Predictive Value of Computed Tomography Angiography–Determined Occlusion Type in Stent Retriever Thrombectomy

• Published in: Stroke (1970) Vol. 48; no. 10; pp. 2746 - 2752
• Main Authors: Baek, Jang-Hyun, Kim, Byung Moon, Yoo, Joonsang, Nam, Hyo Suk, Kim, Young Dae, Kim, Dong Joon, Heo, Ji Hoe, Bang, Oh Young
• Format: Journal Article
• Language: English
• Published: United States American Heart Association, Inc 01.10.2017
• Subjects: Aged; Aged, 80 and over; Computed Tomography Angiography - methods; Endovascular Procedures - methods; Female; Humans; Male; Middle Aged; Multidetector Computed Tomography - methods; Predictive Value of Tests; Retrospective Studies; Stents; Stroke - diagnostic imaging; Stroke - surgery; Thrombectomy - methods; computed tomography angiography; angiography, digital subtraction; intracranial atherosclerosis; stroke; thrombectomy
• ISSN: 0039-2499; 1524-4628; 1524-4628
• DOI: 10.1161/STROKEAHA.117.018096

BACKGROUND AND PURPOSE—We investigated whether occlusion type identified with computed tomography angiography (CTA-determined occlusion type) could predict endovascular treatment success using stent retriever (SR) thrombectomy. METHODS—Consecutive patients with stroke who underwent CTA and then endovascular treatment for intracranial large artery occlusion were retrospectively reviewed. CTA-determined occlusion type was classified into truncal-type occlusion or branching-site occlusion and compared with digital subtraction angiography–determined occlusion type during endovascular treatment. Three rapidly- and readily-assessable pre-procedural findings (CTA-determined occlusion type, atrial fibrillation, and hyperdense artery sign), which may infer occlusion pathomechanism (embolic versus nonembolic) before endovascular treatment, were evaluated for association with SR success along with stroke risk factors and laboratory results. In addition, the predictive power of the 3 pre-procedural findings for SR success was compared with receiver operating characteristic curve analyses. RESULTS—A total of 238 patients (mean age, 70.0 years; male patients, 52.9%) were included in this study. CTA-determined occlusion type corresponded adequately with digital subtraction angiography–determined occlusion type (P=0.453). Atrial fibrillation (odds ratio, 2.66; 95% confidence interval, 1.25–5.66) and CTA-determined branching-site occlusion (odds ratio, 8.20; confidence interval, 3.45–19.5) were independent predictors for SR success. For predicting SR success, the area under the receiver operating characteristic curve value for CTA-determined branching-site occlusion (0.695) was significantly greater than atrial fibrillation (0.594; P=0.038) and hyperdense artery sign (0.603; P=0.023). CONCLUSIONS—CTA-determined branching-site occlusion was significantly associated with SR success. Furthermore, among the 3 rapidly- and readily-assessable pre-procedural findings, CTA-determined branching-site occlusion had the greatest predictive power for SR success.