High-resolution CT angiography for follow-up evaluation of intracranial aneurysms treated with the Woven EndoBridge device: comparison with digital subtraction angiography Authors: Kenichiro Suyama, Ichiro Nakahara, Shoji Matsumoto, Junpei Koge, Jun Tanabe, Akiko Hasebe, Eiji Fujiwara, Shiho Tanaka, Tatsunori Mase, Yuichi Hirose Published: 2025 ABSTRACT AbstractPurposeDigital subtraction angiography (DSA) is the gold standard for follow-up evaluation of intracranial aneurysms treated with the Woven EndoBridge (WEB) device. This study aimed to assess the efficacy of high-resolution CT angiography (HR-CTA) as a less invasive alternative by comparing its diagnostic performance with that of DSA.MethodsThis single-center retrospective study analyzed the angiographic and clinical data of patients treated with the WEB device for cerebral aneurysms between January 2021 and December 2024. Patients who underwent HR-CTA within 2 weeks before or after follow-up DSA were included. Occlusion status was assessed using the Bicêtre Occlusion Scale Score (BOSS) and binary classification. The concordance rate between HR-CTA and DSA was evaluated.ResultsA total of 54 eligible examinations were identified. Using the BOSS, 46 examinations were concordant, resulting in an agreement rate of 85.2%. The Cohen’s κ coefficient was 0.81 (95% CI 0.69 to 0.93), indicating a very high level of agreement. All discordant cases resulted from HR-CTA overestimating occlusion status; however, HR-CTA accurately identified aneurysm remnants. Univariate analyses identified BOSS 0’ as the only significant factor contributing to discrepancies. In the binary evaluation, all 54 examinations were fully concordant (κ=1.00, 95% CI 1.00 to 1.00).ConclusionsHR-CTA demonstrated a high concordance rate with DSA for evaluating occlusion status after WEB placement. Its reliable assessment of aneurysm remnants suggests HR-CTA could serve as a practical and less invasive alternative to DSA in follow-up evaluations.Keywords: Aneurysm, CT Angiography, Flow Diverter, Angiography WHAT IS ALREADY KNOWN ON THIS TOPIC Digital subtraction angiography (DSA) is the gold standard for evaluating occlusion status after Woven EndoBridge (WEB) placement, but its invasiveness poses risks. Digital subtraction angiography (DSA) is the gold standard for evaluating occlusion status after Woven EndoBridge (WEB) placement, but its invasiveness poses risks. WHAT THIS STUDY ADDS High-resolution CT angiography (HR-CTA) demonstrated a high concordance rate with DSA in assessing occlusion status after WEB placement, supporting its potential as a less invasive follow-up imaging modality. High-resolution CT angiography (HR-CTA) demonstrated a high concordance rate with DSA in assessing occlusion status after WEB placement, supporting its potential as a less invasive follow-up imaging modality. All discordant cases resulted from HR-CTA overestimating occlusion status; however, HR-CTA accurately identified aneurysm remnants. All discordant cases resulted from HR-CTA overestimating occlusion status; however, HR-CTA accurately identified aneurysm remnants. Despite its tendency to overestimate occlusion, HR-CTA reliably detects residual aneurysms requiring intervention, ensuring appropriate identification while potentially reducing the need for routine DSA. Despite its tendency to overestimate occlusion, HR-CTA reliably detects residual aneurysms requiring intervention, ensuring appropriate identification while potentially reducing the need for routine DSA. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY Future research should focus on optimizing HR-CTA techniques to improve accuracy and minimize overestimation. Future research should focus on optimizing HR-CTA techniques to improve accuracy and minimize overestimation. Introduction Treatment of intracranial aneurysms has advanced significantly with the development of various endovascular techniques. Among these, the Woven EndoBridge (WEB; TerumoNeuro, Aliso Viejo, CA) device is one of the treatment options for managing wide-neck bifurcation aneurysms. Its efficacy and safety have been well-documented in multiple studies, demonstrating favorable outcomes with a low complication rate.1,3 Digital subtraction angiography (DSA) is currently considered the gold standard for follow-up of intracranial aneurysms treated with the WEB device. DSA provides high-resolution images and dynamic blood flow information. However, DSA is an invasive procedure that involves risks, such as thromboembolic complications, hemorrhage at the puncture site, and nerve injury. Recent advances in imaging technology have led to the development of CT angiography (CTA) using high-resolution CT (HR-CT), which provides a non-invasive alternative for vascular imaging.4,6 CTA provides detailed images of blood vessels without the need for catheterization, thereby reducing the associated risks and discomfort for patients. The use of HR-CTA for follow-up after WEB placement has the potential to offer diagnostic accuracy comparable to that of DSA, while minimizing invasiveness. This study aimed to evaluate the efficacy of HR-CTA in the follow-up of intracranial aneurysms treated with the WEB device by comparing its diagnostic performance with that of DSA. Specifically, we sought to determine whether HR-CTA could serve as a less invasive alternative to DSA. Patients and methods Patient selection This single-center retrospective study evaluated the angiographic and clinical data of consecutive patients treated with the WEB device for cerebral aneurysms between January 2021 and December 2024. At our institution, patients with unruptured aneurysms treated with the WEB undergo follow-up DSA at 3 or 6 months and 12 months after treatment. Additionally, patients with ruptured aneurysms undergo early follow-up DSA between 2 and 4 weeks after WEB placement in addition to the standard follow-up schedule. The inclusion criteria of this study were as follows: patients who had undergone (1) WEB placement more than 2 weeks prior, and (2) HR-CTA within 2 weeks before or after follow-up DSA. Patients with recurrent aneurysms who were previously treated using coils were excluded from the study. The comparison between CTA and DSA was conducted on a per-examination basis, with some patients undergoing multiple examinations. Baseline data (sex and age), aneurysm characteristics (location, size, and aneurysm status), and the WEB device characteristics (size, shape, and volume) were obtained and reviewed retrospectively from medical records. Imaging technique and postprocessing for DSA DSA was performed using a biplane angiographic system (Alphenix Biplane (INFX-8000V); Canon Medical Systems, Otawara, Japan) with a 1024×1024 image matrix and a field of view (FOV) ranging from 1.5 to 12 inches. Standard anteroposterior and lateral projections, working-angle views used during treatment, three-dimensional rotational angiography (3DRA), and HR contrast-enhanced cone-beam CT imaging were included in the DSA protocol. A transfemoral or transradial approach was used to perform selective injections of the internal carotid artery, common carotid artery, or vertebral arteries, as appropriate, using a 4 French (4F) catheter. Anteroposterior and lateral views were acquired with a 7–8 mL bolus of iopamidol (Iopamidol300; Fuji Pharma, Tokyo, Japan) injected at 5–6 mL/s. 3DRA was performed using a 15–18 mL injection at 3 mL/s. The imaging parameters for 3DRA were as follows: tube voltage, 95 kV and tube current, 200 mA. The working views used during treatment were obtained from 3DRA. Additionally, HR contrast-enhanced cone-beam CT was performed using a fivefold diluted contrast agent. The imaging parameters for cone-beam CT were as follows: tube voltage, 90 kV; tube current, 155 mA; and slice thickness, 0.3 mm. Occlusion status was comprehensively evaluated using standard anteroposterior and lateral views, working-angle views, 3DRA, and contrast-enhanced cone-beam CT. The Alphenix Workstation Pro (Canon Medical Systems) was used for image processing and analysis. Imaging technique and postprocessing for HR-CTA HR-CTA was performed using a 160-detector row HR-CT scanner (Aquilion Precision, Canon Medical Systems) with helical scanning. The scan parameters were as follows: section thickness, 0.25 mm; tube voltage, 120 kV; automatic exposure control tube current SD, 5; pitch factor, 0.625; reconstructed FOV size, 100–200 mm; and gantry rotation speed, 1 s/rotation. The volume of the contrast medium was calculated based on each patient’s body weight. The injection rate was set at 30 mg I/kg/s, and iopamidol (Iopamiron 370; Bayer HealthCare, Osaka, Japan) was injected intravenously as a bolus over 13–15 s. This was followed by an intravenous bolus injection of 30 mL of physiological saline solution at the same rate as the contrast medium. To reduce metal artifacts, Single Energy Metal Artifact Reduction (SEMAR) combined with Advanced intelligent Clear-IQ Engine (AiCE), both developed by Canon Medical Systems, was applied in all cases. SEMAR is a single-energy metal artifact reduction algorithm, and it was applied during post-processing. AiCE is a deep learning reconstruction algorithm, and we used AiCE for brain CTA for image reconstruction. In some cases, the application of SEMAR AiCE resulted in decreased image quality. Therefore, the decision to use SEMAR AiCE-processed images was made on a case-by-case basis, and unprocessed images were used when necessary. Occlusion status was comprehensively assessed using 3D and multiplanar reconstruction (MPR) images, including axial, coronal, and sagittal views. The ZIOSTATION2 workstation (Ziosoft, Tokyo, Japan) was used for the image processing and analysis. Data analysis, including the statistical analysis Occlusion status was evaluated using two methods: the Bicêtre Occlusion Scale Score (BOSS),7 a six-grade scale, and a binary classification to determine whether adequate occlusion was achieved. Adequate occlusion was defined as a BOSS score of 0, 0’, or 2. All DSA and HR-CTA images were independently evaluated in a random order by two neurointerventionalists, each with more than 10 years of experience. Interobserver agreement was assessed using the Cohen’s κ coefficient, interpreted as follows: 0.01–0.20 (poor), 0.21–0.40 (fair), 0.41–0.60 (moderate), 0.61–0.80 (good), and 0.81–1.00 (excellent).8 The Cohen’s κ coefficients for interobserver agreement were calculated separately for DSA and HR-CTA. In cases of disagreement, the two interventionalists reached a consensus. Data are presented as mean±SD or median and IQR for continuous variables and frequencies for categorical variables. Intermodality agreement between DSA and HR-CTA was assessed using the Cohen’s κ coefficient. DSA was used as the gold standard for evaluating the occlusion status, and the sensitivity and specificity of HR-CTA were calculated for each BOSS grade. A bubble plot was utilized to visualize the agreement based on the BOSS. Univariate analyses were performed using Student’s t-test or Fisher’s exact test to identify factors contributing to the discrepancies between HR-CTA and DSA. To ensure data independence and minimize bias from multiple examinations, only one examination per patient was included in the univariate analysis. For patients who underwent multiple sets of examinations, the chronologically first set was used to reduce selection bias. Statistical significance was set at P<0.05. In cases of disagreement between DSA and HR-CTA, we investigated the regions of aneurysmal blood flow that were inadequately visualized on HR-CTA. Aneurysmal blood flow was categorized into three regions: (1) within the WEB device, (2) around the neck, and (3) outside the WEB device (online supplemental figure S1). All analyses were conducted using DSA as the gold standard to evaluate the performance of HR-CTA in assessing occlusion status. All statistical analyses were performed using R 4.3.3 (The R Project for Statistical Computing, www.r-project.org). Results During the study period, 152 aneurysms in 147 patients were treated with the WEB device. Among them, 54 eligible examinations were identified, covering 47 aneurysms in 46 patients. The baseline characteristics are presented in online supplemental table S1. The average age of patients was 63.5±12.5 years, and the median maximum dome diameter was 6.5 mm. Among the 47 aneurysms, 18 (28.3%) were ruptured aneurysms. Among the 54 examinations, the mean interval from treatment to examination was 3.7±3.5 (range 0.5–12) months, and the mean interval between HR-CTA and DSA was 1.9±2.3 (range 1–14) days. Five of the examinations involved the use of stents. Among these, four were Neuroform Atlas (Stryker, Prtage, MI, USA), and one was LVIS Jr (TerumoNeuro). Based on the BOSS, an overall agreement between DSA and HR-CTA (intermodality agreement) was observed in 46 of 54 examinations, resulting in an agreement rate of 85.2%. The Cohen’s κ coefficient was 0.81 (95% CI 0.69 to 0.93), indicating a very high level of agreement. The correlation between DSA and HR-CTA is shown in a bubble plot diagram (figure 1), which demonstrates concordance among most cases. In this bubble plot, all discordant cases were located above the diagonal, indicating that HR-CTA overestimated the occlusion status compared with DSA in every discordant case. Bubble plot illustrating the correlation between HR-CTA and DSA evaluations based on the BOSS. The x-axis represents the HR-CTA scores, and the y-axis represents the DSA scores. The size of each bubble corresponds to the frequency of cases with the respective HR-CTA and DSA scores. Larger bubbles indicate a higher frequency of cases. The numbers inside each bubble denote the case count for that specific score pairing. Concordant cases between HR-CTA and DSA are located along the diagonal, and the bubble plot demonstrates that most cases exhibit concordance. BOSS, Bicêtre Occlusion Scale Score; DSA, digital subtraction angiography; HR-CTA, high-resolution CT angiography. The sensitivity and specificity of HR-CTA for each BOSS grade, with DSA as the reference standard, are shown in online supplemental table S2. The sensitivity was particularly low for BOSS 0’. Table 1 shows the results of univariate analyses performed to identify the factors contributing to the discrepancy. The only significant factor identified was the BOSS 0’, which had a significantly lower concordance rate. The aneurysm height tended to be a contributing factor to the discrepancy. However, the difference was not statistically significant. The number of discordant cases was too small to perform a statistically valid multivariate analysis. Therefore, multivariate analysis was not performed. Table 1Factors contributing to the discrepancy between HR-CTA and DSA in aneurysm occlusion assessmentAgreement(n=39)Disagreement(n=7)P valueAge, years64.7±12.856.7±8.70.12Female21 (53.8%)3 (42.9%)0.69Dome, mm7.0 (5.9–8.6)5.9 (5.5–6.5)0.17Height, mm5.2 (4.2–6.2)4.4 (3.9–4.6)0.06Width, mm5.3 (4.1–6.5)4.6 (4.3–5.1)0.59Neck, mm3.9 (3.3–4.6)3.6 (3.3–4.1)0.46WEB volume, cm385 (59–154)85 (59–99)0.41Stent3 (7.7%)0 (0%)1BOSS grade 09 (23.1%)0 (0%)0.32 0’0 (0%)5 (71.4%)< 0.001 116 (41.0%)0 (0%)0.08 25 (12.8%)1 (14.3%)1 30 (0%)0 (0%)NA 1+39 (23.1%)1 (14.3%)1To ensure data independence and minimize bias from multiple examinations, only one examination per patient was included in the univariate analysis. For patients who underwent multiple sets of examinations, the chronologically first set was used to reduce selection bias.BOSS, Bicêtre Occlusion Scale Score; DSA, digital subtraction angiography; HR-CTA, high-resolution CT angiography; WEB, Woven EndoBridge. Factors contributing to the discrepancy between HR-CTA and DSA in aneurysm occlusion assessment To ensure data independence and minimize bias from multiple examinations, only one examination per patient was included in the univariate analysis. For patients who underwent multiple sets of examinations, the chronologically first set was used to reduce selection bias. BOSS, Bicêtre Occlusion Scale Score; DSA, digital subtraction angiography; HR-CTA, high-resolution CT angiography; WEB, Woven EndoBridge. There were eight cases of discrepancy, all of which involved HR-CTA overestimating the occlusion status compared with DSA. The BOSS grades of DSA for the discrepant cases were as follows: BOSS 0’, five; BOSS 2, two; and BOSS 1+3, one. We investigated the regions where HR-CTA failed to visualize aneurysmal flow and found that, in all cases, HR-CTA was unable to depict the flow around the neck. Based on the binary evaluation, complete concordance (54/54) was observed between DSA and HR-CTA. The Cohen’s κ coefficient was 1.00 (95% CI 1.00 to 1.00), indicating complete concordance. A heatmap was used to evaluate the correlation between DSA and HR-CTA results (online supplemental figure S2). Representative cases are illustrated in figures2 3. Additional representative cases are provided in onlinesupplemental figures S3S6. Images of an unruptured right middle cerebral artery aneurysm treated with the WEB (SL 11/6). DSA (A, B, C) and HR-CTA (D, E, F) are performed 1 year after treatment. 3D rotational angiography (A, B) and cone-beam CT (C) from DSA show residual flow in the aneurysm neck (black arrowhead) and within the WEB (white arrowhead) (BOSS 1+3). Similarly, 3D images (D, E) and an MPR image (F) from HR-CTA also show residual flow in the aneurysm neck (white arrow) and within the WEB (white arrowhead) (BOSS 1+3). The DSA and HR-CTA results are concordant. BOSS, Bicêtre Occlusion Scale Score; DSA, digital subtraction angiography; HR-CTA, high-resolution CT angiography; MPR, multiplanar reconstruction; 3D, three-dimensional; WEB, Woven EndoBridge. Images of an unruptured basilar artery aneurysm treated with the WEB (SL 6/3). DSA (A, B) and HR-CTA (C, D) are performed 3 months after treatment. 3D rotational angiography (A) and cone-beam CT (B) from DSA show a convex-shaped remnant at the neck (white arrowhead) (BOSS 0’). In contrast, a 3D image (C) and an MPR image (D) from HR-CTA show complete occlusion (BOSS 0), failing to detect the convex-shaped remnant at the neck. The DSA and HR-CTA results are discrepant because HR-CTA overestimated the occlusion status. BOSS, Bicêtre Occlusion Scale Score; DSA, digital subtraction angiography; HR-CTA, high-resolution CT angiography; MPR, multiplanar reconstruction; 3D, three-dimensional; WEB, Woven EndoBridge. Interobserver agreement was also evaluated using the BOSS and binary evaluation. Based on the BOSS, agreement between the two raters was achieved in 48 out of 54 DSA examinations (88.9%, κ=0.84, 95% CI 0.72 to 0.95) and in 46 out of 54 HR-CTA examinations (85.2%, κ=0.80, 95% CI 0.67 to 0.93), respectively. Based on the binary evaluation, agreement between the two raters was achieved in 52 out of 54 DSA examinations (96.3%, κ=0.93, 95% CI 0.82 to 1.00) and in 51 out of 54 HR-CTA examinations (94.4%, κ=0.89, 95% CI 0.77 to 1.00), respectively. Discussion Two previous studies have reported on the usefulness of CTA for the follow-up of aneurysms treated using the WEB.9 10 Raoult et al reported a Cohen’s κ coefficient of 0.61 between CTA and DSA based on a four-grade evaluation in 16 examinations, and 1.00 based on a binary evaluation. They used a 64-detector row CT scanner with a section thickness of 0.5 mm.9 Ozpeynirci et al reported a Cohen’s κ coefficient of 0.63 (95% CI 0 to 1) between CTA and DSA based on a binary evaluation in 16 examinations. They used a 40-detector row CT scanner with a thickness of 0.67 mm.10 In the present study, the Cohen’s κ coefficient between CTA and DSA was 0.81 based on a six-grade evaluation (BOSS) in 54 examinations, and 1.00 based on a binary evaluation. This demonstrates a very high level of concordance compared with previous reports, attributable to the use of a 160-row HR-CT scanner with a section thickness of 0.25 mm, which allows for the acquisition of more detailed images. The use of a higher-performance CT scanner enabled a diagnostic accuracy very close to that of DSA, further supporting the feasibility of HR-CTA as a reliable follow-up imaging modality for aneurysms treated with the WEB. Several studies have investigated the usefulness of magnetic resonance angiography (MRA) for follow-up of aneurysms treated with the WEB.911,15 Among these, three studies directly compared MRA with DSA,9 11 12 while the others either compared MRA with CTA or DSA,14 or focused on the utility of MRA techniques for follow-up without direct comparisons.13 15 Timsit et al reported a Cohen’s κ coefficient of 0.36 for both the comparison of DSA with contrast-enhanced MRA and 3D-time-of-flight MRA, based on a binary evaluation in 26 examinations. Both MRA techniques demonstrated low sensitivity (25%) for aneurysm detection.12 Raoult et al reported a Cohen’s κ coefficient of 0.69 between MRA and DSA based on a four-grade evaluation in 16 examinations, and 0.64 based on a binary evaluation.9 Mine et al reported a Cohen’s κ coefficient of 0.67 between MRA and DSA in 16 examinations.11 MRI has the advantage of being radiation-free. However, these reports collectively indicate that the concordance rate with DSA is generally higher for CTA than for MRI. This is further supported by a study that compared the MRI results with those of both DSA and CTA.14 Based on the current evidence, CTA appears to be a more reliable non-invasive alternative to MRA for follow-up evaluation. A notable feature of this study was the inclusion of early post-treatment follow-up, which resulted in a higher rate of aneurysm remnants (51.9%). In contrast, previous studies often conducted follow-up 1 year after treatment, leading to a lower reported rate of aneurysm remnants (6.3–15.5%).9,12 This limitation in previous studies made it uncertain whether aneurysm remnants could be reliably assessed. However, in this study, the determination of the occlusion status using binary classification showed complete concordance, supporting the reliability of HR-CTA in accurately identifying aneurysm remnants. Although HR-CTA can visualize aneurysm remnants, it tends to fail to detect flow around the neck of aneurysms treated with the WEB. This was likely due to artifacts caused by the proximal marker of the WEB. The proximal marker is made of platinum alloy, and its high metal content contributes to these artifacts. While DSA is necessary for evaluating the occlusion status accurately, CTA is sufficient for determining effective occlusion and identifying cases requiring retreatment. Therefore, the limitations of HR-CTA pose minimal clinical issues. Future directions may include the use of dual-energy CTA, which allows for tube voltage optimization and improved tissue contrast without compromising diagnostic accuracy. Additionally, photon-counting CT has the potential to provide superior spatial resolution and reduce metal artifacts. These innovations may further advance the non-invasive assessment of aneurysms treated with the WEB device. This study has some limitations. The quality of HR-CTA can be slightly affected by factors such as the timing of the scan, the patient’s physical status, and direction of the WEB. Additionally, this study was conducted at a single center with a retrospective design and limited sample size, which may have impacted the generalizability of the findings. A larger prospective study would be beneficial for validating these results. Furthermore, while the HR-CT system used in this study provides high spatial resolution, such equipment is not universally available at all institutions. This may limit the widespread applicability of our findings and highlights the need for further studies using different imaging systems. Despite these limitations, the findings of this study have significant implications for improving patient care and optimizing follow-up protocols in the endovascular treatment of intracranial aneurysms using the WEB. Conclusions Evaluation of occlusion status using HR-CTA after WEB placement demonstrated a high concordance rate with DSA. In discordant cases, HR-CTA tended to overestimate occlusion status. The assessment of aneurysm remnants was reliable, suggesting that HR-CTA could serve as a practical and less invasive alternative to DSA for follow-up evaluations. Supplementary material Acknowledgments We would like to thank radiologic technologist Akio Katagata for his assistance in CTA image processing and reconstruction. Footnotes Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Data availability free text: The imaging data that support the findings of this study have been anonymised to protect patient privacy and are not publicly available, but they can be obtained from the corresponding author upon reasonable request. Ethics approval: This study was approved by the Institutional Review Board (IRB) of Fujita Health University (approval number: HM22-128). Given the retrospective nature of the study, the IRB waived the requirement for individual informed consent and implemented an opt-out approach. Information about the study and the option to opt out was made publicly available on our institution’s website, ensuring transparency and allowing patients the opportunity to decline participation. For the procedures involving DSA and CTA, informed consent was obtained from each patient before each examination. This consent form ensured that the patients were fully informed about the nature, benefits, and potential risks associated with these imaging studies. Data availability statement Data are available upon reasonable request. References Associated Data Supplementary Materials Data Availability Statement Data are available upon reasonable request.