Procedural Outcome Following Stent-Assisted Coiling for Wide-Necked Aneurysms Using Three Different Stent Models: A Single-Center Experience

Previous case series have described the safety and efficacy of different stent models for stent-assisted aneurysm coiling (SAC), but comparative analyses of procedural results are limited. This study investigates the procedural outcome and safety of three different stent models (Atlas™, LEO+™ (Baby) and Enterprise™) in the setting of elective SAC treated at a tertiary neuro-endovascular center. We retrospectively reviewed all consecutively treated patients that received endovascular SAC for intracranial aneurysms between 1 July 2013 and 31 March 2020, excluding all emergency angiographies for acute subarachnoid hemorrhage. The primary procedural outcome was the occlusion rate evaluated with the Raymond–Roy occlusion classification (RROC) assessed on digital subtraction angiography (DSA) at 6- and 12-month follow-up. Safety assessment included periprocedural adverse events (i.e., symptomatic ischemic complications, symptomatic intracerebral hemorrhage, iatrogenic perforation, dissection, or aneurysm rupture and in-stent thrombosis) and in-house mortality. Uni- and multivariable logistic regression analyses were performed to identify patient baseline and aneurysm characteristics that were associated with complete aneurysm obliteration at follow-up. A total of 156 patients undergoing endovascular treatment via SAC met the inclusion criteria. The median age was 62 years (IQR, 55–71), and 73.7% (115) of patients were female. At first follow-up (6-month) and last available follow-up (12 and 18 months), complete aneurysm occlusion was observed in 78.3% (90) and 76.9% (102) of patients, respectively. There were no differences regarding the occlusion rates stratified by stent model. Multivariable logistic analysis revealed increasing dome/neck ratio (adjusted odds ratio (aOR), 0.26.; 95% CI, 0.11–0.64; p = 0.003), increasing neck size (aOR, 0.70; 95% CI, 0.51–0.96; p = 0.027), and female sex (aOR, 4.37; 95% CI, 1.68–11.36; p = 0.002) as independently associated with treatment success. This study showed comparable rates of complete long-term aneurysm obliteration and safety following SAC for intracranial aneurysm with three different stent-models highlighting the procedural feasibility of this treatment strategy with currently available stent-models. Increased neck size and a higher dome/neck ratio were independent variables associated with less frequent complete aneurysm obliteration.


Introduction
Stent-assisted coiling (SAC) is a popular treatment option for both small and complex intracranial aneurysms, as it facilitates coil packing of wide-necked aneurysms that are at risk of coil dislocation and herniation into the parent target vessel, causing potentially periprocedural or delayed ischemic complications [1,2]. Furthermore, SAC allows higher densities of endo-saccular coil embolization and, as a result, higher occlusion and lower recurrence rates due to improved aneurysm thrombosis in comparison to coiling alone, which is not suitable in the majority of wide-necked aneurysm [3].
In the evolution of intracranial stents approved for assisted-coiling, various types of stents have become commercially available and along with them different procedural advantages [4].
We compared and analyzed the procedural results of three different three self-expanding stent models and hypothesized that all three types of stents were efficient for assisted-coiling of wide-necked aneurysms, resulting in high rates of complete aneurysm obliteration at the time of follow-up.

Study Population
We retrospectively analyzed data from all consecutive patients with intracranial aneurysms endovascularly treated between 1 July 2013 and 31 March 2020 at our comprehensive stroke center. All patients receiving SAC were selected and analyzed. The angiographic results of three self-expanding stents (LEO+(Baby)™, Balt, Montmorency, France; Neuroform Atlas™, Stryker Neurovascular, Fremont, CA, USA; En-terprise™, Codman Neurovascular, Raynham, MA, USA) were analyzed together with patient baseline characteristics and anatomical aneurysm details treated consecutively at our comprehensive stroke center.
The main inclusion criteria for all cases were (1) the diagnosis of an intracranial aneurysm within the anterior or posterior circulation and (2) endovascular treatment in an elective setting (3) via SAC using the LEO+, Enterprise, or Atlas stent models, regardless of previous treatments (i.e., endovascular solely coiling or surgical clipping). There were no exclusion criteria regarding the choice of navigation and delivery catheters. All emergency angiographies were excluded, including patients with acute subarachnoid hemorrhage (SAH) due to intracranial aneurysm rupture.
For all patients meeting the inclusion criteria, baseline and procedural characteristics were collected from medical records and imaging studies. All data were recorded in accordance with the local ethical review board, and informed consent was waived after review (ethics committee of the chamber of Physicians, Bremen, Germany) due to the retrospective study design using fully anonymized data only.

Antiplatelet and Anticoagulation Therapy
All patients, from 7 days to a minimum of 3 days before the procedure patients, received dual antiplatelet therapy primarily with Acetylsalicylic acid (100 mg daily) and Clopidogrel (75 mg daily). All patients were tested using Multiplate ® test or in vitro platelet function for adequate response to anti-platelet therapy, and based on the test results, the medication was adjusted. For non-responders, Clopidogrel was changed to Ticagrelor. At the beginning of the intervention, weight-adapted Heparin was administered intra-arterially. After the procedure, all patients were kept lifelong on daily Acetylsalicylic acid and daily Clopidogrel for a minimum of six month. In cases with signs of in-stent stenosis, the patient was kept on dual antiplatelet therapy for another 3 months until the next follow-up.

Endovascular Procedure
All procedures were carried out under general anesthesia using a biplane angiography system (Philips Allura Xper FD20/15, Koninklijke Philips N.V., Amsterdam, The Netherlands). Cerebral vessel access was established using a 6F 088 Neuron MAX Long Sheath 90/4 Straight (Penumbra, Inc., Alameda, CA, USA). Subsequently, a three-dimensional rotational angiography was performed to plan the procedure. In all cases that involved placement of a LEO+(Baby) or Enterprise stent, jailing technique was used: First, a microcatheter system was navigated in the aneurysm; second, another microcatheter system was navigated distal to the aneurysm, and the stent was then partially unsheathed and the aneurysm was coiled. After the aneurysms were fully packed with coils, the stent was completely unsheathed. In the cases in which the Atlas stent was used, the flow diverter was first deployed, and then a microwire/microcatheter system was navigated through the stent into the aneurysm for subsequent coiling.

Stent Models
The use of each stent model depended on the commercial availability and the preference of the interventionalist in charge of the procedure.
The Neuroform Atlas™ (Stryker Neurovascular, Fremont, CA, USA) is a laser cut, nitinol, self-expanding stent system consisting of a hybrid design with closed cells proximally and open cells distally. Accordingly, the proximal part facilitates microcatheter re-crossing, and the distal part allows for better anchoring and stability within the distal landing zone of the artery, as well as vessel wall apposition. The stent includes radiopaque markers at the proximal and distal parts to guarantee visualization under fluoroscopy. For the stent delivery it is mandatory to use the Excelsior SL-10 0.016 microcatheter (Stryker, Fremont, CA, USA), and stents are available in diameters and lengths of 3.0 to 4.5 mm and 15 to 30 mm, respectively [5][6][7].
The LEO+ (Baby) ™ (Balt, Montmorency, France) is a nitinol self-expanding stent system with a woven closed-cell design. The stent contains two helical platinum wires that radiopaque for the entire length of the stent, allowing full visualization of both length and diameter under fluoroscopy. This feature facilitates identification of the stent position and deployment. The LEO+ (Baby)™ can be placed through every 0.017" microcatheter lumen, and stent sizes are available in diameters and lengths of 2.0 to 5.5 mm and 12 to 75 mm, respectively [6,8,9].
The Enterprise™ (Codman Neurovascular, Raynham, MA, USA) is a self-expandable nitinol stent with a closed-cell design. The ending of the stent contains four radiopaque markers on each side. The delivery system contains a radiopaque marker in the midportion of the stent for better visualization during the deployment process. For the stent delivery it is mandatory to use the 0.021" Prowler Plus microcatheter (Codman Neurovascular, Raynham, MA, USA). The Enterprise™ stent is available in diameters and length of 4.5 mm and 14 to 37 mm, respectively [6,10].

Aneurysm Details and Procedural Outcome
Aneurysm details were analyzed including the exact location, size, dome/neck ratio, and pre-/post-target arterial vessel diameters in the aneurysm area. All aneurysm details were stratified by stent models. The angiographic outcome was evaluated postinterventionally, and at the time of follow-up, angiography by digital subtraction angiography (DSA) using the Raymond-Roy occlusion classification (RROC) [11]. In this scheme, Class I is defined as complete obliteration, Class II as residual neck, and Class III as residual aneurysm (5). RROC I was defined as the primary angiographic outcome endpoint, and all occlusion rates were compared by utilized stent models. Patient baseline, aneurysm, and procedural characteristics were analyzed for significant associations with complete aneurysm occlusion (RROC 1) at last 6 month (first follow-up) and last available follow-up, including 12-and 18-month follow-ups.
Periprocedural and long-term complications were defined as ischemic events without being clinically symptomatic (any signs of ischemia on follow-up imaging) or ischemic events leading new relevant disabling neurological deficits, occurrence of SAH on postprocedural follow-up imaging or intracerebral hemorrhage (ICH) with regard to clinical symptoms (sICH, defined as an bleeding event that most likely caused neurological deterioration). Furthermore, the occurrence of in-stent thrombosis, intima hyperplasia or stenosis, or thrombosis periprocedurally or at the time of follow-up imaging, including computed tomography angiography and magnetic resonance imaging.

Statistical Methods
Standard descriptive statistics were applied for all data endpoints. Univariable distribution of metric variables was described with median and interquartile range (IQR). Categorical variables were compared with the Chi-squared or Fishers exact test was performed. The Mann-Whitney U test (non-normally distributed data) was used to compare continuous variables. Uni-and multivariable logistic regression analysis was performed to identify variables (patient baseline characteristics and aneurysm details) associated with complete aneurysm occlusion (RROC 1). Variables included in the model were selected based on available literature and clinical relevance [12][13][14][15][16]. Odds ratios (OR) and adjusted OR (aOR) are presented with 95% confidence intervals. p values < 0.05 were considered as statistically significant. All statistical analyses were performed in with SPSS version 27.0 (IBM Corporation, Armonk, NY, USA) and Stata 17.0 (StataMP, StataCorp, College Station, TX, USA).

Discussion
Our study on the impact of different stent models on angiographic outcome following SAC for wide-necked aneurysms revealed several findings: (1) a similar high rate of complete obliteration on long-term follow-up can be expected when performing

Discussion
Our study on the impact of different stent models on angiographic outcome following SAC for wide-necked aneurysms revealed several findings: (1) a similar high rate of complete obliteration on long-term follow-up can be expected when performing SAC with different stent models (Atlas™, Enterprise™, LEO+™); (2) periprocedural complication rates were relatively low when compared previous studies on SAC; and (3) increased neck size and a higher dome/neck ratio were independent variables associated with less frequent complete aneurysm obliteration.
Wide-necked aneurysms are, in some cases, untreatable or difficult to treat with standard coiling [17]. For these cases, endovascular therapy strategies include flow diverters with and without additional coiling, balloon or temporary stent-assisted (i.e., Comaneci) coiling, and coiling in assistance with intracranial stents. In the latter, the stent covers the neck, stabilizing the coil mass inside the aneurysmal sac and avoids coil herniation into the parent artery. Since the first report of self-expandable stents for intracranial use, multiple advances in stent designs have expanded the commercial availability of different types of intracranial stents for SAC with different properties and, thus, technical advantages and disadvantages [18].
In our case series, we compared three self-expandable nitinol stents, including the Atlas™, with a lase-cut hybrid closed/open-cell design, the Enterprise™, with a laser-cut closed-cell design, and the LEO+(Baby)™, with a woven close-cell design. Additionally, open-cell stent designs have the advantage of improved wall apposition in tortious vessels but the disadvantage that re-sheathing is impossible and the increased cell size, especially in the vertex region of curved vessels, provokes the risk of coil dislocation and subsequent prolapse. Closed-cell stent designs, on the other hand, contain the risk of poor wall apposition and kinking tortious vessel curves. We observed a relatively high rate of complete aneurysm obliteration (RROC I) of 78% in the whole study cohort, which is in line with previous studies investigating SACs ranging from 54-81% with different stent models [4,5,[19][20][21]. In our study the rate did not differ between different stent models but was numerically higher in the closed-cell LEO+™ stent group. This finding may be explained by the higher mesh density of the LEO+™ compared with the other stents; however, a bias due to the higher frequency of use of this stent at our department combined with greater experience, which is known to be critical for better outcomes in neuroendovascular procedures, cannot be excluded [22]. Additionally, Pumar et al. observed slightly lower rates (73%) of complete occlusion at long-term follow-up [23]. Multivariable logistic regression analysis showed that an increasing dome/neck ratio and increased neck size were independently associated with incomplete aneurysm obliteration. At first glance, these results seem contradictory, but they must be interpreted in the context of wide-necked aneurysms with relatively increasing neck and aneurysm sizes. Previous studies corroborated that that especially large and wide-necked aneurysms tend to show incomplete aneurysm obliteration at the time of follow-up. Interestingly, even in our cohort that only consisted of wide-necked aneurysms, these findings were stable and may be considered a priori as negative prognostic predictors for complete occlusion in patients treated with SAC for wide-necked aneurysms. Furthermore, female sex was independently associated with aneurysm obliteration, which is most likely biased by the fact that aneurysm cohorts are likely to include higher numbers of women than men because female sex is a common risk factor for developing intracranial aneurysms and rupture [24][25][26].
The endovascular procedure of SAC has been proven to be effective in the past but remains a technical challenging intervention that is associated with increased risks for ischemic events and intracranial hemorrhage due to the necessity of dual antiplatelet therapy prior and after the procedure. Additionally, balloon-assisted coiling may have an advantage regarding the control of medical risk factors for periprocedural hemorrhage but increases the weight of ischemic risk factors. Additionally, the rates of ischemic events and intracranial hemorrhage after SAC in the literature differ around 5% and 2%, respectively [19]. In our case series, we observed low rates of hemorrhagic and ischemic events, highlighting the general safety of the procedure. However, one patient died after a symptomatic bleeding event, and one patient experienced new and disabling deficits leading to a permanent hemiparesis. Nevertheless, these periprocedural complications are marking a devastating clinical course in an elective treatment setting. Important longterm complications after SAC are in-stent intima hyperplasia and stenosis. Our results of an 6% in-stent stenosis rate corroborated previous studies that reported on in-stent thrombosis rate of 10.2% and did observe differences in the risk of developing in-stent stenosis between different stent models such as Atlas and Enterprise [27][28][29]. However, this delayed complication underlines the importance of follow-up imaging not only for detecting aneurysm recurrence but also for cases with in-stent stenosis that possibly need to be re-treated [30].

Study Limitations
This study has all limitations that are associated with a retrospective single-center study, including missing data (e.g., follow-up imaging) in a certain number of cases and unequal numbers of patients in each stent group. Additionally, occlusion rates were not assessed by an independent core laboratory; however, the treatment results were evaluated by an independent and experienced analyst that did not take part in the procedure. Followup data on functional and clinical outcome were not available.

Conclusions
This study showed comparable rates of complete long-term aneurysm occlusion rates and following SAC for intracranial aneurysm with three different stent-models. Increased neck size and a higher dome/neck ratio were independent variables associated with less frequent complete aneurysm obliteration. Periprocedural complications were comparatively low highlighting the procedural feasibility and safety of this endovascular treatment strategy with currently available stent models.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board (or Ethics Committee) of Ethikkommission der Ärztekammer Bremen. Patient consent was waived due to the retrospective fully anonymized study design.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.