Visualization of large vessel occlusion, clot extent, and collateral supply using volume perfusion flat detector computed tomography in acute stroke patients. (November 2019)

• Record Type: Journal Article
• Authors: Hoelter, Philip; Goelitz, Philipp; Lang, Stefan; Luecking, Hannes; Kalmuenzer, Bernd; Struffert, Tobias; Doerfler, Arnd
• Is Part Of: Acta radiologica. Volume 60:Number 11(2019)
• Page Start: 1504
• Page End: 1511
• Publication Date: 2019-11
• Subjects: Central nervous system -- interventional -- stroke -- flat detector computed tomography; Radiology, Medical -- Periodicals; Radiography, Medical -- Periodicals; Radiotherapy -- Periodicals; 616.0757
• DOI: 10.1177/0284185119836220
• Languages: English
• ISSNs: 0284-1851

Abstract

Background: Flat detector computed tomography angiography (FDCTA) can be reconstructed from volume perfusion flat detector computed tomography (VP-FDCT). Thus, CTA equivalent stroke imaging might be feasible within the angio suite. Purpose: To evaluate the diagnostic accuracy of FDCTA in detecting large vessel occlusion (LVO) and collateral supply in acute stroke patients. Material and Methods: Sixteen patients with LVO of the anterior circulation were analyzed retrospectively. Each patient underwent a multimodal CT stroke protocol, subsequent VP-FDCT, and digital subtraction angiography (DSA) for endovascular stroke therapy. Angiographic images reconstructed secondarily from VP-FDCT were evaluated with regard to visualization of LVO, Collateral Score (CS), Clot Burden Score (CBS), and image quality. Results: Image quality of FDCTA was sufficient for diagnosis with a strong correlation between CTA and FDCTA (median score: CTA = 4 ± 0.447; FDCTA = 4 ± 0.5). Detection of LVO could be achieved with high sensitivity and specificity for FDCTA and CTA (97.9%, 95% confidence interval [CI] = 96.0–99.9; 92.6%, 95% CI = 84.3–100.0 vs. 96.8%, 95% CI = 93.2–100.0; 86.3%, 95% CI = 88.2–95.2). CBS and CS assessment showed no significant difference between FDCTA and CTA for both readers (reader1CBS : P = 0.751, reader1CS : P = 0.718; reader2CBS : P = 0.164; reader2CS : P = 0.582) and an excellent interrater agreement (CTACBS ICC = 0.984, FDCTACBS ICC = 0.951; CTACS ICC = 0.754, FDCTACS ICC = 0.789). Conclusion: FDCTA, reconstructed from VP-FDCT data, allows reliable detection of ICA or MCA M1 segment occlusion comparable to CTA and may provide information concerning the clot extent with sufficient image quality.