Introduction

Endovascular coiling has been established as the standard treatment modality for intracranial aneurysms1,2. Although advancements in microcatheter and coil technologies have enabled less traumatic intra-arterial access and improved aneurysm occlusion3,4, aneurysm recurrence remains a significant limitation, particularly in cases of wide-necked aneurysms5. Consequently, endoluminal flow diverters, stent-like devices placed across the aneurysm neck, have been introduced to divert blood flow away from aneurysms6,7. These flow diverter devices promote interface remodeling between the aneurysm and the parent artery and emphasize the importance of achieving a stable structure across the aneurysm neck for aneurysm treatment, even in wide-necked cases8,9. However, the clinical utility of flow diverters is limited, as their technical characteristics make them difficult to apply to bifurcation aneurysms, and most cases require long-term dual antiplatelet medication10,11.

As a solution to these limitations, intrasaccular flow disruptor (IFD) devices have recently been developed for placement entirely within the aneurysmal cavity11,12. These devices function as stand-alone therapy and aim to create flow disruption within the aneurysm through a mechanism similar to that of an endoluminal flow diverter8. IFDs have been designed in various shapes, including spherical, hemispherical, and umbrella-shaped, and eliminate the need for dual antiplatelet therapy in the treatment of large, wide-necked, and bifurcation aneurysms11,13,14. Although IFD devices have shown promise in aneurysm treatment, the available shapes and types of the device remain limited.

Among the recently introduced IFDs, the Woven EndoBridge (WEB; MicroVention; Aliso Viejo, CA, USA) is the most extensively studied device in both preclinical and clinical settings, and has been shown to achieve complete deployment within the aneurysmal cavity, providing neointimal coverage across the aneurysmal neck15,16. This device features a flat or slightly concave base design, which helps reduce the risk of protrusion into the parent artery when covering the aneurysm neck. However, in many wide-neck bifurcation aneurysms, the bifurcation region often presents a relatively large space between the parent artery and branch vessels, which may limit ideal neck coverage with this base design17. The Contour Neurovascular System (Contour; Ceres Neuromuscular, Fremont, CA, USA) features a reversed umbrella shape that covers the aneurysmal neck to restrict inflow into the aneurysm sac at the neck level18. Due to its distinctive design, it multi-functions as both a flow diverter and IFDs, and is also intended to reconstruct the natural bifurcation of the artery19. A structural advantage of this device is its relatively low dependence on precise sizing compared with other devices, allowing for greater flexibility in size selection20. However, suboptimal deployment may result in inadequate apposition to the aneurysmal wall, potentially compromising efficacy. In addition, limited usability in acute settings further compounds these challenges19. Considering the advantages and limitations of existing devices, a novel FD (NFD) was designed to incorporate key beneficial features. Like the WEB, it exerts radial expansile force, providing stability during deployment, while also adopting the size selection simplicity of the Contour. Additionally, it was developed to reconstruct the aneurysm neck in alignment with the bifurcation anatomy.

This study aimed to evaluate the safety, performance, and efficacy of the NFD in a rabbit elastase-induced aneurysm model.

Methods

Preparation of the NFD

The NFD was designed by combining the intrasaccular shape of the WEB device with the conical configuration of the Contour, which crosses the aneurysmal neck. The NFD embolization device was constructed using a double-layer braided structure composed of 144 interwoven nitinol wires, each with a diameter of 30 μm (HansBioMed, Seoul, Korea). Radiopaque platinum markers were positioned at both proximal and distal ends to enhance visualization during deployment. The distal portion of the NFD was designed with a conical shape, featuring a proximal cone angle of approximately 15°, intended to minimize protrusion into the parent artery, particularly in bifurcation aneurysms. Upon deployment, the device conforms closely to the base of the aneurysm, covering the aneurysm neck while expanding fully to occupy the aneurysmal cavity. Devices were manufactured in two sizes: 5 mm (width) × 2.5 mm (height) and 6 mm × 3 mm.

The NFD was delivered using a commercially available 4.2 Fr intermediate catheter (Asahi Fubuki; Asahi Intecc, Aichi, Japan) to ensure precise distal navigation (Fig. 1a). Deployment into the aneurysmal cavity was achieved using a controlled push–pull technique (Fig. 1b, c). In addition, NFD is designed to be non-retrievable device once deployed, resulting proper positioning prior to final detachment. Detachment of the device was mechanically accomplished by rotating the delivery handle counterclockwise, following operator confirmation of satisfactory positioning.

Fig. 1
figure 1

Technical details and three-dimensional modeling of the aneurysm and novel flow disruptor (NFD) for simulation analysis. (a) NFD loaded into a 4.2 Fr intermediate catheter using a manual crimping technique. (b) Partially deployed NFD. (c) Fully deployed NFD exhibiting a conical shape with a proximal cone angle of approximately 15°, functioning as both a flow diverter and disruptor. (d) Computational modeling of a bifurcation cerebral aneurysm and NFD. (e) Geometry and simulated deployment configuration of the NFD. (f) Defined regions (A1–A5) for velocity measurement within the aneurysm.

Full size image

Computed fluid dynamics (CFD) simulation of blood flow velocity

To evaluate hemodynamic characteristics at the aneurysm neck, CFD simulations were performed using COMSOL Multiphysics® v5.6 (COMSOL, Inc., Burlington, MA, USA). CFD analyses were conducted to compare intra-aneurysmal blood flow velocities before and after NFD deployment. The flow-diverting and flow-disrupting effects of the NFD were assessed through numerical modeling of changes in blood flow velocity within the aneurysm. The material properties used for blood and arterial walls in the simulation are listed in Table 1. For CFD simulations, a bifurcation aneurysm model was employed, and the NFD device was specifically designed to achieve both flow diversion and flow disruption simultaneously (Fig. 1d).

Table 1 Material properties used in the computed fluid dynamics simulation model21,22.
Full size table

For the numerical modeling of the NFD, morphological characteristics were measured using optical microscopy based on the fully deployed configuration, as shown in Fig. 1c. Key design parameters, such as pore size between braided nitinol wires, wire angles, and curvature diameters, were measured and used to construct a simplified, representative structural model of the NFD for CFD simulation. The bifurcation model was designed based on the morphological characteristics and dimensional ratios of a patient-specific aneurysm obtained from computed tomography (CT) imaging. Because the NFD closely apposes the aneurysm wall after deployment, it was modeled with the same elliptical outer curvature as the fully deployed and geometrically simplified NFD structure to replicate its post-deployment behavior within the aneurysm. As illustrated in Fig. 1e, the NFD was deployed to fully appose the aneurysm neck within the bifurcation aneurysm model, maintaining an insertion angle of 15° within the aneurysm. Blood flow velocities were measured at three key aneurysm locations (neck, maximum diameter, and dome), and five intra-aneurysmal reference points (A1–A5), enabling comprehensive evaluation of local hemodynamic changes induced by the device (Fig. 1f)23.

Aneurysm model creation

Elastase-induced aneurysms were created in 21 New Zealand White rabbits, as previously described24,25. Briefly, after identification and surgical exposure of the right carotid artery (RCA), a 5 Fr arterial sheath (Terumo Corporation, Tokyo, Japan) and a trimmed 5 Fr guiding catheter (Envoy; Cordis Corporation, Miami Lakes, FL, USA) connected to a rotating hemostasis valve were introduced. Subsequently, a 3 Fr Fogarty balloon catheter (Edwards Lifesciences, Irvine, CA, USA) was inserted to occlude the RCA stump. A 1:1 mixture of porcine elastase solution and contrast medium was incubated within the arterial segment for 20 min (Supplementary Fig. 1a-c). Following incubation, the elastase mixture was aspirated, the RCA lumen was thoroughly flushed with saline, and the RCA was ligated proximally. After the procedure, animals received appropriate analgesics and antibiotics and were housed under standardized environmental conditions.

Animal experiments

Following aneurysm maturation for at least 28 days, animals underwent NFD implantation procedures. Under general anesthesia with a mixture of 50 mg/kg zolazepam, 50 mg/kg tiletamine (Zoletil 50; Virbac, Carros, France), and 10 mg/kg xylazine (Rompun; Bayer HealthCare, Leverkusen, Germany), the right common femoral artery was cannulated using a 6 Fr arterial sheath (Radifocus® Introducer II; Terumo Corporation, Tokyo, Japan), and intravenous heparin (100 U/kg; JW Pharmaceutical Corporation, Seoul, Korea) was administered. Subsequently, a 6 Fr guiding catheter (Envoy; Cordis Corporation, Miami Lakes, FL, USA) was introduced over a 0.032-inch hydrophilic guidewire (Radifocus® Guidewire M; Terumo Corporation, Tokyo, Japan) and positioned within the brachiocephalic artery. Subsequently, digital subtraction angiography (DSA) was performed to measure aneurysm dimensions, including neck diameter, width, and height.

The size of the NFD device was selected according to measured aneurysm dimensions, specifically based on the maximum aneurysm width and height. The NFD was delivered through a 4.2 Fr intermediate catheter (Asahi Fubuki; Asahi Intecc Co., Ltd., Aichi, Japan) and deployed precisely into the aneurysm cavity under fluoroscopic guidance. After satisfactory device positioning and mechanical detachment by rotating the delivery handle counterclockwise, the intermediate catheter was removed. Immediate post-implantation DSA images were acquired via the guiding catheter (Supplementary Fig. 1d-f).

Subsequently, the femoral artery was ligated using 4 − 0 black silk sutures. No antiplatelet medication was administered before or after device implantation. Animal body weights were monitored weekly until the designated endpoint.

Animals were randomly assigned to three follow-up groups: 1-month (n = 7), 2-month (n = 7), and 3-month (n = 7) post-implantation. For follow-up DSA, the left femoral artery was cannulated with a 5 Fr arterial sheath (Radifocus® Introducer II; Terumo Corporation, Tokyo, Japan) under general anesthesia. Angiographic evaluations were conducted using a 4 Fr angiography catheter (Jung Sung Medical, Ansung, Korea) positioned in the brachiocephalic artery. After completion of follow-up imaging, the rabbits were euthanized according to the predetermined schedule by intravenous administration of potassium chloride (JW Pharmaceutical Corporation, Seoul, Korea).

Angiographic analysis

Intra-aneurysmal flow disruption observed immediately post-implantation within 30 min was classified into four grades: Grade 1, persistent flow or mild disruption blood flow within the aneurysm; Grade 2, moderate disruption blood flow within the aneurysm; Grade 3, near-complete disruption of blood flow within the aneurysm; and Grade 4, complete disruption of blood flow within the aneurysm8,26. Follow-up angiographic aneurysm occlusion was assessed using a 3-point grading scale: Grade 1, incomplete occlusion (< 90%); Grade 2, near-complete occlusion (> 90%); and Grade 3, complete occlusion6,8. Each follow-up DSA image was compared with the immediate post-embolization angiogram to evaluate evidence of device migration from the aneurysmal cavity into the parent artery.

Histopathology

After euthanizing the animals, the aneurysm and a portion of the parent artery were harvested and fixed in 10% formaldehyde for 48 h. Long-axis sectioning of the aneurysmal neck was performed on a coronal plane containing the aneurysm and parent artery, and the specimens were sliced at 500-µm intervals. Specimens were then sequentially dehydrated using different concentrations of alcohol. Subsequently, samples were embedded in a resin block using glycol methacrylate (Technovit 7200® VLC; Heraus Kulzer GMBH, Wertheim, Germany) infiltration. These specimens were further sectioned at 10-µm and stained with hematoxylin and eosin for histologic evaluation.

All histologic analyses were conducted using a digital slide scanner, and measurements at the aneurysm neck were obtained using a digital microscope viewer. Final histologic findings represent the consensus of three observers who were blinded to the experimental groups.

Approval for animal experiments

This study was approved by the Institutional Animal Care and Use Committee (IACUC No. 2023-KE-0480) and conducted in accordance with the guidelines of US National Institute of Health for the humane handling of laboratory animals. The study was performed in compliance with the ARRIVE guidelines.

Results

Simulated blood flow velocity

Velocity streamline distributions at peak velocity, derived from CFD simulations comparing scenarios with and without NFD placement, are presented in Fig. 2. Time-dependent blood flow velocities measured at five specific points (A1–A5) around the aneurysm neck are also shown. Changes in maximum blood flow velocity following NFD deployment are summarized in Table 2. After NFD placement, significant reductions in blood flow velocity were observed compared with baseline: specifically, velocity decreased by 96.3% at the aneurysm neck, 99.9% at the point of maximum aneurysm diameter, and ceased completely (100%) at the dome. The conical design of the NFD led to symmetrical velocity reductions near the aneurysm neck wall (points A1 and A5, and points A2 and A4), resulting in balanced flow diversion around the aneurysm neck.

Fig. 2
figure 2

Velocity streamline distributions at the time of peak velocity. (a) Velocity streamline before (yellow dotted line) and after placement of novel flow disruptor (NFD) (red dotted line) in a bifurcation cerebral aneurysm model. Blood flow velocities at five points (A1-A5) on the aneurysm neck (b) before and (c) after NFD placement. Note. w/o device: without device, NFD: novel flow disruptor.

Full size image
Table 2 Comparison of the maximum blood flow velocity changes within the cerebral aneurysm.
Full size table

Procedural outcomes

Elastase-induced aneurysm creation was technically successful in all rabbits (100%), with no procedure-related complications observed during the aneurysm maturation period. Additionally, aneurysm formation was successful in all rabbits (100%). All 21 aneurysms treated with the NFD devices were successfully implanted, and all animals survived until the end of the study at the corresponding time point without any adverse events or neurologic deficits. Body weight decreased during the first week post-procedure but gradually recovered thereafter (Supplementary Fig. 2). Aneurysm dimensions and corresponding device sizes are summarized in Table 3. The mean (± standard deviation [SD]) aneurysm width, height, and neck were 3.52 ± 0.56 mm, 7.27 ± 2.25 mm, and 3.75 ± 0.80 mm, respectively. Based on the average aneurysm height and width, nine 5 × 2.5 mm NFDs and twelve 6 × 3 mm NFDs were used for aneurysm treatment.

Table 3 Summary of serial cases.
Full size table

Angiographic findings

Aneurysm occlusion grades immediately after NFD implantation and at follow-up periods of 1, 2, and 3 months are summarized in Table 4. Representative angiographic images are shown in Fig. 3.

Table 4 Aneurysm occlusion grade summary at corresponding time points.
Full size table
Fig. 3
figure 3

Novel flow disruptor (NFD) placement into rabbit elastase-induced aneurysm model and representative follow-up angiographic and respective histological images. (a ) Digital subtraction angiography (DSA) after NFD detachment showing immediate flow disruption and complete occlusion on follow-up. (b) Histology image showing loose connective tissue within the aneurysm cavity and neointima layers across the neck level. (c) DSA after NFD detachment showing moderate flow disruption and near complete occlusion on follow-up. (d) Histology image revealing concave connective tissue across the aneurysm neck, with partial thrombus and a neck remnant observed on follow-up angiography in the (c). (e) DSA after NFD detachment showing mild flow disruption and incomplete occlusion on follow-up. (f) Histology image showing minimal connective tissue in the aneurysm cavity, with no neointimal layer at the neck.

Full size image

Immediately after implantation (within 30 min), angiographic evaluations demonstrated near-complete or complete disruption (Grades 3 or 4) in 11 out of 21 aneurysms (52%). Moderate disruption (Grade 2) was observed in six aneurysms (29%), and mild disruption (Grade 1) in four aneurysms (19%). Additionally, protrusion of the NFD device into the parent artery was noted in five cases (23%) immediately after placement.

A total of 21 aneurysms were assigned to three groups (n = 7 each) for cross-sectional follow-up at 1-, 2-, or 3-months post-implantation. In the 1-month group, complete occlusion (Grade 3) was observed in two cases (29%), near-occlusion (Grade 2) in three cases (42%), and incomplete occlusion (Grade 1) in two cases (29%). The two-month group demonstrated complete occlusion in two cases (29%), near-occlusion in four cases (57%), and incomplete occlusion in one case (14%). In the three-month group, complete occlusion was observed in two cases (29%), near-occlusion in three cases (42%), and incomplete occlusion in two cases (29%).

Device-related observations at follow-up were summarized collectively (n = 21): device protrusion increased to 10 cases (48%), migration of the NFD occurred in two cases (9%), and deformation was noted in four cases (19%). Importantly, no arterial abnormalities, such as thrombosis, dissection, luminal narrowing, or embolization, were detected at immediate or follow-up angiography.

Histopathological findings

Histological images are presented in Fig. 4. Histological findings revealed myofibroblast infiltration, organized thrombus, and organized loose connective tissue in the domes of six (29%) completely occluded aneurysms. In these cases, a thin neointima layer with endothelial cells lining the aneurysmal neck was identified. Furthermore, the entire aneurysm cavity was filled with organized loose connective tissue and thrombus. In 10 aneurysms (48%) with near-complete occlusion, a mixture of organized and unorganized thrombus along with loose connective tissue was observed within the aneurysm dome. The neointima layer was identified over part of the neck, whereas the open area at the neck remnant exhibited a concave shape toward the aneurysmal dome. Furthermore, the entire aneurysm cavity was filled with mixed organized/unorganized thrombus and loose connective tissue. Five aneurysms (24%) demonstrated incomplete occlusion, characterized by a large concave neck remnant containing predominantly unorganized thrombus and loose connective tissue. Due to the intentional design of the NFD, protrusion was noted in five aneurysms (23%) immediately after placement, increasing to 10 aneurysms (48%) at follow-up, contributing to incomplete aneurysm occlusion. Among the 10 protruding cases, six (60%) aneurysms were incompletely occluded, with the proximal tip of the NFD floating in the space between the parent artery and aneurysm. Meanwhile, near complete occlusion was achieved in four (40%) cases where the proximal tip protruded toward the wall of the parent artery and was embedded in the arterial wall. In these cases, a neointima layer was observed extending from the embedded tip of the NFD to the concave portion at the neck remnant.

Fig. 4
figure 4

Representative hematoxylin and eosin (H&E)-stained images of aneurysm treatment with novel flow disruptor (NFD) protrusion. (a) H&E images showing a near-complete aneurysm within the aneurysm cavity (black dotted line) and at the neck level (red dotted line). (b) The aneurysmal cavity is filled with loose connective tissue (magnification: × 10.0). (c) The proximal tip of NFD protrudes toward the parent artery and is embedded in the arterial wall (magnification: × 3.0). A concave shape (star) is formed toward the aneurysmal dome. (d) H&E images showing an incomplete aneurysm within the aneurysm cavity (black dotted line) and at the neck level (red dotted line). (e) The aneurysmal cavity is filled with loose connective tissue (magnification: × 10.0). (f) The proximal tip of the NFD is floating (arrowheads) in the parent artery (magnification: × 3.0). A macroconcave shape (star) is formed toward the aneurysmal dome.

Full size image

Discussion

In this study, the NFD was designed with a conical shape to be deployed within the aneurysmal cavity, covering both the aneurysm neck and the enlarged space between the parent artery and bifurcations. This distinctive design was expected to function as both a flow diverter and an IFD, enabling more ideal anatomical reconstruction of wide-necked bifurcation lesions. To analyze the flow diversion and disruption characteristics of the NFD, CFD simulations were performed to compare the intra-aneurysmal blood flow velocity before and after device placement. Owing to its combined flow-diverting and flow-disrupting properties, the NFD achieved reductions in flow velocity of up to 96.8%, 99.9%, and 100% at the aneurysm neck, maximum diameter, and dome, respectively, and resulted in a more balanced overall flow distribution, particularly at the aneurysm neck. Due to the intentional protrusion of the device, early post-implantation angiography revealed near-complete or complete flow disruption (Grades 3 or 4) in 52% (11/21) of aneurysms, relatively lower than rates reported in previous preclinical studies6,8. However, follow-up angiography demonstrated a high rate of near-complete or complete aneurysm occlusion in 16 cases (76%). In these cases, histologic analysis revealed treatment effects within the aneurysmal cavity and along the aneurysmal neck in most cases, typically accompanied by complete or near-complete cessation of blood flow. In the remaining five cases (24%), the proximal end of the NFD was observed to float within the parent artery, without evidence of re-endothelialization, and exhibited a significant reduction in disorganized thrombus and loose connective tissue in the aneurysmal lumen and dome. These findings suggest that the appropriate placement of the IFD, ensuring intrasaccular positioning and interaction with the parent artery through intentional protrusion, may facilitate re-endothelialization in aneurysm treatment.

Currently available intra-aneurysmal IFDs vary considerably in shape and structural properties, which influence their clinical applicability and outcomes. The WEB and LUNA spherical devices are designed as single-piece structures to serve as alternatives to coil embolization, particularly for bifurcation anatomy and vascular beds containing eloquent penetrating arteries, while also offering a compression-resistant construct6,8,27. The Contour shares similar applicability with the WEB for treating wide-necked bifurcation aneurysms;19 however, its disc-like configuration results in a non-self-compressible, ‘stent-like’ coil-retainer effect. Furthermore, its shape, which does not completely fill the aneurysm sac, may be advantageous in managing large, partially thrombosed aneurysms18,28,29. The newly proposed NFD in this study features a conical configuration designed to support the aneurysm neck, characterized by a deeper, dome-conforming mesh volume that extends into the aneurysmal sac, unlike the Contour. This geometric depth is expected to promote more extensive intra-aneurysmal flow disruption. Furthermore, the NFD’s proximal portion comprises a tightly braided mesh structure designed to ensure robust coverage of the aneurysm neck plane, a contrast to the Contour device’s looser, centrally open framework. In follow-up angiography and histological assessment of neck coverage, the NFD system showed comparable performance to the WEB device (Supplementary Table 1). Although the precise metal coverage ratio (MCR) of the NFD was not quantitatively assessed in this study, its fabrication specifications closely resemble those of the WEB device. The WEB device’s neck plane MCR has been reported in the range of 35–45%, depending on deployment configuration and aneurysm morphology30,31. Given the denser braid pattern observed in some NFD deployment scenarios, the effective MCR at the neck surface may approach or exceed that of the WEB, suggesting potential for improved inflow disruption and subsequent thrombosis, particularly in wide-neck bifurcation aneurysms. In terms of reducing intra-aneurysmal flow velocity of the CFD simulation study, the NFD demonstrated an effective decrease in blood flow velocity along the aneurysm neck wall while maintaining a uniform flow disruption effect. The NFD’s morphological characteristics, as evaluated in simulation studies, effectively integrate the advantages of both flow disruption and diversion. Therefore, the NFD design can effectively reduce intra-aneurysmal flow velocity after insertion, supporting its potential to promote aneurysm occlusion and indicating the conceptual soundness of the design. Although the hemodynamic advantages of the NFD have been clearly demonstrated in computational simulations, in vivo results may be influenced by biological factors such as endothelialization and IFD–tissue interactions. Further validation by physical performance assessment and clinical research is required to substantiate the observed flow velocity reduction ratios.

A mixture of organized or disorganized thrombus, along with loose connective tissue, was observed within the aneurysmal sac. NFDs that were fully inserted into the sac demonstrated superior occlusion rates compared with those protruding into the aneurysmal neck. Therefore, rather than focusing solely on the design of the IFD, ensuring that the IFD is fully inserted during deployment, without protrusion from the sac, appears to be an important factor for successful aneurysm treatment. This finding is consistent with previous studies emphasizing the importance of minimizing protrusion32,33. Remodeling of endothelial cells that form along the aneurysm neck is considered one of the essential mechanisms for proper aneurysm healing16,34. Neointimal cells formed through interactions between the occlusive device and parent artery are regarded as the primary cause of aneurysmal occlusion, ultimately leading to endothelialization of the aneurysmal neck over the device surface35,36. Thus, promoting aneurysm neck coverage with endothelial cell growth is a potential therapeutic target for long-term protection against aneurysm recurrence37. In the present study, the intentional design of the NFD led to protrusion in 10 cases; however, not all of these cases resulted in incomplete aneurysm occlusion. In four cases (40%), when the proximal end of the NFD protruded toward the parent artery wall, a neointimal layer was observed extending from the embedded end of the NFD to the concave portion of the neck remnant. In a few cases, re-endothelialization of adjacent vessels was also observed, attributed to interaction between the protruding NFD tip and the parent artery. As previously noted, the presence or absence of IFD protrusion is an important factor in aneurysm treatment. However, treatment outcomes varied significantly depending on whether the protruding IFD floated within the aneurysmal neck cavity or was anchored to and interacted with the parent artery. Although intentional protrusion was part of the NFD design, protrusion into the parent artery negatively impacted occlusion rates unless the device was securely embedded into the arterial wall. Future device iterations should prioritize minimizing protrusion and enhancing stable wall apposition to consistently achieve effective aneurysm occlusion. Moreover, neoendothelialization along the IFD at the aneurysmal neck periphery, originating from the parent artery wall, has been shown to definitively exclude the aneurysm from the native circulation38,39. Notably, aneurysms exhibiting suboptimal disruption at early time points (e.g., immediately after implantation) did not exhibit progressive thrombosis or neointimal maturation at 1, 2, and 3 months. This suggests that the initial hemodynamic response to the device determines long-term occlusion outcomes in the rabbit elastase-induced model. Therefore, early poor occlusion may serve as a predictor of treatment failure, regardless of the follow-up period. The present study highlights the central role of neointimal cell formation along the aneurysmal neck, facilitated by interaction between the IFD and parent artery.

In this study, the NFD frequently exhibited deformation or shape modification, a phenomenon analogous to the WEB-shape modification reported in commercialized WEB devices40,41. This has been observed in approximately 60% of cases in both preclinical and clinical studies. Ding et al. reported a significant correlation between changes in device width and aneurysm healing and fibrosis within the first month following aneurysm repair42. Similarly, Han et al. observed that when aneurysm height significantly exceeds width, excessive lateral compression of the device is inevitable within the aneurysm cavity43. Taken together, these factors may contribute to significant morphological deformation of the NFD device, potentially affecting its intended placement. Additionally, variations in the degree of lateral compression may result from differences between the external environment of subarachnoid aneurysms within the human skull and the surrounding conditions of rabbit aneurysms, which must be carefully considered. The design of the device significantly influences its mechanical properties, with the edges typically generating the greatest radial forces, whereas the waist remains more vulnerable to external compression. Under conditions of extreme lateral compression, the WEB SL device often deforms into a characteristic “hourglass” shape43. Given the conical shape of the NFD used in this study, its waist may be particularly vulnerable to external compression. This susceptibility could contribute to the relatively lower occlusion rates observed and should be carefully considered in the design and development of future intra-aneurysmal IFD.

This study has several limitations. First, the NFD prototypes were manufactured in only two sizes, which may have limited the ability to achieve optimal aneurysm neck coverage and wall apposition across all test subjects. Second, the elastase-induced aneurysm model used in this study consistently produces elongated aneurysm sacs, which differ morphologically from typical clinical wide-necked bifurcation aneurysms that are generally broader at the neck and shorter in height. In addition, when applied to aneurysms with limited luminal space and a nearly spherical shape, as commonly seen in clinical practice, there may be a risk of parent artery stenosis/ischemic complication due to the intentional protrusion of the proximal tip into the parent artery, and a separate risk of aneurysm wall perforation from the distal marker tip. These morphological differences significantly limit the generalizability of the findings to diverse clinical aneurysm anatomies, emphasizing the need for further validation and optimization of the NFD protrusion design in more clinically relevant aneurysm models. Third, the delivery system used in this study (4.2 Fr intermediate catheter) represents a significant limitation, as it does not reflect typical clinical practice where smaller microcatheters are routinely employed. Although the relatively larger intermediate catheter facilitates device insertion, it may not accurately represent the real-world procedural challenges, where device delivery through smaller microcatheters may lead to increased friction, reduced maneuverability, and greater procedural complexity44. Although successful deployment in this experimental setup demonstrates initial feasibility, refinement of the delivery system to ensure compatibility with standard microcatheters is essential for improving clinical applicability and procedural success in future studies. Finally, due to technical constraints in performing repeated invasive angiographic examinations in the same animal, only cross-sectional follow-up imaging at 1, 2, and 3 months was feasible, with different animals evaluated at each time point. Although this approach allowed initial evaluation of both short- and intermediate-term outcomes, the absence of longitudinal follow-up within the same animals limits the ability to precisely assess temporal changes in aneurysm healing and device behavior. Future studies employing non-invasive longitudinal imaging modalities or refined long-term experimental designs allowing repeated evaluations in the same subjects would provide more robust insights into the dynamic healing processes associated with the NFD.

In conclusion, the newly developed NFD was designed to function simultaneously as a flow diverter and IFD, incorporating advantageous features from existing devices. Promising aneurysm occlusion outcomes were achieved when the device was adequately positioned, particularly when the protruding tip was effectively embedded into the arterial wall, thereby promoting neointimal formation. However, the intended design feature, a conical base allowing slight protrusion, did not consistently demonstrate superior effectiveness, likely due to the elongated morphology and anatomical limitations of the rabbit elastase-induced aneurysm model. Although these preliminary findings are encouraging, further studies incorporating improved delivery systems, diverse aneurysm morphologies, and longitudinal imaging follow-up within individual subjects are required to definitively validate the clinical potential and optimize the design of this NFD device.