Introduction

Wide-necked bifurcation aneurysms present significant challenges for endovascular treatment due to their broad neck and associated daughter branches. Various strategies, including double-balloon remodeling, dual stent-assisted coiling, flow diverters, neck bridging, and intrasaccular devices, have been developed to address these challenges [1]. Y-stent-assisted coiling (Y-SAC) was first introduced in 2004 by Chow et al. [2] ,and despite all advancements, still remains as an effective and durable treatment option for wide-necked bifurcation aneurysms.

Between 2004 and 2018, the majority of reported series on Y-SAC utilized stents that were delivered through microcatheters with an internal diameter of either 0.027 or 0.021 inches. Catheterizing the sharply angled and narrow side branches of bifurcations and navigating these large-profile catheters through the struts of a fully deployed first stent is challenging. A high rate of periprocedural complications and technical complications, including stent dislocation or deformation, have been reported [3,4,5,6]. However, the newer version of intracranial self-expandable stents has a low-profile and they can be delivered using microcatheters specifically designed for coiling, with internal diameters of 0.0165 and 0.017 inches [7,8,9]. Their miniaturized design enables more straightforward catheterization of sharply angled bifurcation branches and smoother catheterization through the stent struts. Many centers have published their experiences with Y-SAC using these newer stents, reporting relatively high technical success and favorable safety and efficacy outcomes [7,8,9,10,11,12,13].

To our knowledge, there is no comprehensive study systematically assessing the influence of these newly introduced low-profile stents on the safety and efficacy of Y-SAC. Herein, we present a systematic review and meta-analysis that evaluates the angiographic and clinical outcomes of Y-SAC, using these low-profile stents for the treatment of wide-necked bifurcation aneurysms.

Materials & methods

Literature search

We conducted a systematic literature search using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [14] guidelines across PubMed, Embase, Cochrane Library, and Web of Science databases from inception to September 29, 2023. Our search terms included “y-stent”, “y-configuration”, “y-“, “dual”, “double”, “stent”, “stent-assisted”, “y-stent-assisted”, “low-profile”, “neuroform atlas”, “lvis jr”, “acclino”, “Lvis”, “low-profile visualized intraluminal support”, “leo baby”, “accero”, “wide-necked”, “wide-neck”, “wide”, “neck”, “bifurcation”, “aneurysm”, “intracranial”, “brain”, and “cerebral”. We used “AND” and “OR” Boolean operators to maximize the search sensitivity. Our full search strategy is provided in Supplementary Table 1. We used the Zotero reference manager for de-duplication of the studies. Following this, the initial study triage was done based on titles and abstracts using the ‘Rayyan’ web application for systematic reviews [15].

Eligibility criteria

The low-profile self-expandable intracranial stents are compressed to a very small diameter for delivery through ultra-fine 0.0165-inch and 0.017-inch microcatheters, which are specifically used for coil delivery. A summary of low-profile stents used in the Y-SAC is provided in Table 1. Our inclusion criteria encompassed studies in the English language reporting on series of ≥ 5 patients with intracranial wide-necked bifurcation aneurysms treated with Y-SAC using low-profile stents, which included both angiographic and clinical outcomes. Exclusion criteria were stated in the supplementary material 1. Two independent authors screened articles in their entirety to determine eligibility for inclusion (A.G. and M.Y.F.).

Table 1 Overview of low-profile stents used in the Y-stent-assisted coiling technique
Full size table

Angiographic and neurological assessment

Aneurysms were categorized as small (≤ 10 mm), large (10–25 mm), and giant (> 25 mm). Aneurysms with a neck diameter ≥ 4 mm or a dome-to-neck ratio < 2 mm were defined as wide-necked. The Raymond-Roy occlusion classification (RR) was used to assess the degree of aneurysm occlusion [16]. RR class 1 was defined as complete occlusion, RR class 2 as neck residue, and RR class 3 as sac residue. RR classes 1 and 2 were defined as adequate occlusion. In the studies where RR or modified RR was not used, complete occlusion was identified only when the term “complete occlusion” was used. Terms such as “near complete occlusion” or “neck remnant/residue” that signify only neck filling were considered as adequate occlusion along with complete occlusion. Terms such as “sac residue” or “sac/body filling” were identified as inadequate occlusion. The immediate degree of aneurysm occlusion was determined by angiographic evaluation at the conclusion of the procedure. Follow-up occlusion degrees were assessed from either MRA or DSA follow-ups. If a study reported multiple angiographic follow-up results, those with the longest duration and DSA follow-ups were prioritized. Recanalization was identified if any deterioration in the RR class was reported (e.g., from 1 to 2/3 or 2 to 3) or if terms like “recanalization” or “aneurysm growth” were used. Significant chronic in-stent stenosis at follow-up was identified by the presence of in-stent stenosis of ≥ 50% or if terms such as “significant stenosis or intimal hyperplasia” were explicitly mentioned.

Complications that occurred during the operation were classified as intraoperative, those that occurred within 1-month post-procedure were classified as early postoperative, and those that arose after 1 month were classified as delayed. Procedure-related complications encompassed intraoperative, early postoperative, and delayed complications. An acute in-stent thrombus was identified if any thrombus occurred in the newly deployed stent(s) and obstructed blood flow, either partially or totally, during the intraoperative or early postoperative period. Procedure-related stroke referred to an ischemic stroke that was directly attributable to the procedure, whether it occurred as an intraoperative, early postoperative, or delayed complication. Clinical outcomes were assessed with the modified Rankin Scale (mRS); scores of 0–2, or terms like “no morbidity”, “good recovery” or “asymptomatic” were indicative of good clinical outcomes (absence of any procedure-related morbidity or death). Procedure-related morbidity was defined as a clinical outcome of mRS 3–5 due to procedure-related complications. Procedure-related mortality was defined as the death of a patient from procedure-related complications. For cases presenting with rupture, any initial rupture-related morbidities or death were excluded from these pooled calculations.

Outcomes and data extraction

For each study, we extracted data on the number of patients, number of aneurysms, sex, mean age, aneurysm rupture status, mean aneurysm diameter, mean neck size, and durations of clinical and angiographic follow-up. The antiaggregation protocol for each study was noted for unruptured cases, and ruptured cases (if any). Additionally, we obtained details on technical success, type of stents used, in-stent stenosis, procedure-related complications, procedure-related stroke, procedure-related morbidity and mortality rates, and immediate and follow-up angiographic and clinical outcomes, as well as recanalization and retreatment rates.

The primary outcomes of this study included immediate and follow-up aneurysm occlusion, aneurysm recanalization, retreatment, and procedure-related complications. In addition, although classified as either intraoperative or early postoperative complication, acute in-stent thrombus was also included as a separate primary outcome. Secondary outcomes included technical success, procedure-related morbidity, procedure-related mortality, procedure-related stroke, and chronic in-stent stenosis at follow-up. Outcomes based on rupture status or size were not differentiated, as most studies did not stratify results by rupture status or mean diameter.

Study risk of bias assessment

We utilized a modified version of the Newcastle-Ottawa Scale to evaluate the risk of bias in the included studies, based on guidance from Zhao et al. [17] In line with the criteria set by Granja et al. [18], we designated studies that had satisfactory angiographic and clinical follow-ups, along with clear outcome reports, as having a “low risk of bias.” Studies with unsatisfactory follow-ups were labeled as “medium risk of bias.” Those lacking both satisfactory follow-ups and clear outcome reports were categorized as “high risk of bias.”

Statistical analysis

In each cohort, we determined the cumulative incidence (event rate) and its associated 95% confidence interval (CI) for every outcome. We assessed data heterogeneity across the studies using the Higgins index (I2), noting that I2 values of ≥ 50% indicate high heterogeneity. A p-value of less than 0.05 (p < 0.05) was considered statistically significant. We employed the DerSimonian and Laird random-effects model. We additionally conducted subanalyses of the studies that used solely laser-cut stents and studies that used solely braided stents. A two-proportion z-test was applied to compare the percentages of these subanalyses and determine p-values. We conducted the meta-analysis and generated forest plots using the OpenMeta[Analyst] software (http://www.cebm.brown.edu/openmeta/).

Results

Study selection and characteristics

The literature search in stated databases yielded 1221 articles. After removing duplicates and assessing titles and abstracts for relevance, we excluded 1090 articles. We further reviewed 131 articles in full-text. Out of these, 112 were excluded, primarily because they pertained to single stent-assisted coiling series or did not provide adequate data on the Y-SAC subgroup (Supplementary Table 2). In the end, 19 studies [7,8,9,10,11,12,13, 19,20,21,22,23,24,25,26,27,28,29,30], representing 507 patients with 509 aneurysms, were included in our final analysis. Of these, 11 had a low risk of bias [7,8,9,10,11, 19,20,21, 26, 28, 30], six had a medium risk of bias [12, 13, 23, 25, 27, 29], and two had a high risk of bias [22, 24] (Supplementary Table 3). The search flow is presented in Fig. 1.

Fig. 1
figure 1

PRISMA flow diagram

Full size image

The mean age of patients was 58.6 ± 1.4 years, and the majority were female (66.2%). Among the 509 assessed aneurysms, 38 presented as ruptured, 471 as unruptured, and 53 experienced recurrence post-treatment—52 following coiling and one following clipping. The largest study consisted of 111 patients with 111 aneurysms, while the smallest had 6 patients with 6 aneurysms. The most common aneurysm locations were the middle cerebral artery (MCA) bifurcation (196, 38.5%), basilar apex (153, 30%), and anterior communicating artery (Acom) (134, 26.3%). Y-stenting in a kissing configuration was reported in only one study, [25] where all stents used were Acclino (Acandis, Pforzheim, Germany). All other aneurysms in the included studies were treated with crossing Y-configuration. In all studies, treatment was conducted in a single stage; no study reported a staged Y-SAC.

Out of the 509 cases, only three were treated with two different stents. In these cases, the first stent was a Neuroform Atlas (Stryker Neurovascular, Fremont, CA, USA), while the second stent was an LVIS (MicroVention, Tustin, California) in two cases, and an Enterprise (Codman Neurovascular, Raynham, Massachusetts, USA) in one. Despite the second stents being non-low-profile, these cases were retained in the overall analysis since individual outcomes were not reported. All other included aneurysms were treated with the same low-profile stents. Two Neuroform Atlas stents were used in 206 cases, two Leo Baby stents (Balt Extrusion, Montmorency, France) in 155 cases, and two Lvis junior (MicroVention, Aliso Viejo, CA, USA), stents in 121 cases. A summary of the included studies is provided in Table 2. All studies reported dual antiplatelet therapy, most commonly consisting of 75 mg clopidogrel and 100–300 mg aspirin, started from 2 days to 2 weeks pre-treatment. For ruptured cases treated in the acute setting, either dual antiplatelets were loaded pre-treatment, or a glycoprotein IIb/IIIa inhibitor or cangrelor infusion was administered before stent deployment. All studies reported the continuation of dual antiplatelet regimen post-treatment ranging from 3 months to 12 months. The full antiaggregation protocol for all studies is detailed in Supplementary Table 4.

Table 2 Summary of the included studies
Full size table

Angiographic and clinical outcomes

Immediate adequate aneurysm occlusion was reported by 17 studies, with pooled rates being 90% (95% CI, 86–94.0%). Immediate complete occlusion was reported by 16 studies, with a rate of 65% (95% CI, 53–77%). Follow-up angiographies were available for 443 aneurysms. The mean angiographic follow-up duration was 15.6 ± 1.9 months (range, 4.4–43.5 months). The rates for follow-up adequate aneurysm occlusion and complete occlusion were 98% (95% CI, 97–99%) and 89% (95% CI, 85–93%), respectively. Aneurysm recanalization rates were 3% (95% CI, 2–5%), and aneurysm retreatment rates were 1.7% (95% CI, 0.7–3.1%). All reported cases of recanalization or retreatment are detailed in Supplementary Table 5.

Pooled rates for intraoperative and early postoperative complications were 6% (95% CI, 3-9%) and 3% (95% CI, 1.5-4.6%), respectively. The acute in-stent thrombus was the most common specified complication during these periods, developing in 18 cases with a pooled rate of 2% (95% CI, 0.9–3.3%). Only 6 cases of delayed complications were reported, with a pooled rate of 1% (95% CI, 0–2%). However, the majority of procedure-related complications resulted in good clinical outcomes. After a mean clinical follow-up of 15 ± 2.4 months, the rate of good clinical outcomes was 98% (95% CI, 97–99%). All reported procedure-related complications, their management (if any), and outcomes during follow-up are detailed in Supplementary Table 6.

A high rate of technical success was reported, with pooled rates across the studies being 99% (95% CI, 98–100%) (Fig. 2). During the procedure, early postoperative, and delayed periods, a total of 27 cases of procedure-related strokes occurred. Due to procedure-related complications, mortality occurred in 2 patients and morbidity in 9 patients. Overall, procedure-related morbidity and procedure-related mortality rates were 1.3% (95% CI, 0.4–2.3%) and 0.4% (95% CI, 0–1.4%), respectively. All reported procedure-related stroke cases, mortality, and morbidity cases are also detailed in Supplementary Table 6. After a mean clinical follow-up of 15 ± 2.4 months, the rate of good clinical outcomes was 98% (95% CI, 97–99%). Follow-up DSA revealed chronic in-stent stenosis in 10 cases, with a pooled rate of 1% (95% CI, 0–2.2%). The results of primary and secondary outcomes are summarized in Table 3.

Fig. 2
figure 2

Forest plot of technical success across included studies

Full size image
Table 3 Results of the meta-analysis
Full size table

When comparing the results of studies that solely used laser-cut stents with those that solely used braided stents, follow-up adequate occlusion was reported significantly higher in the braided stent group at 99% (95% CI, 98-100%) compared to the laser-cut group at 96% (95% CI, 93–98%), with a p-value of 0.002. Follow-up adequate occlusion rates for the laser-cut and braided stent groups is shown in Fig. 3. Acute in-stent thrombus was reported significantly more often in the braided stent group than in the laser-cut stent group, while chronic in-stent stenosis and procedure-related morbidity were observed more frequently in the laser-cut stent group (Table 4).

Table 4 Subanalysis by the structural design of stents
Full size table

Heterogeneity assessment

The majority of outcomes demonstrated low or no heterogeneity. For immediate angiographic results, considerable heterogeneity was observed, with I2 values exceeding 50%. Intraoperative complications also showed heterogeneity, with an I2 value of 63%. In subanalysis, heterogeneity in intraoperative complications was observed only in the laser-cut stents group.

Fig. 3
figure 3

Follow-up adequate occlusion rates for the laser-cut (a) and braided (b) stent groups

Full size image

Discussion

Our meta-analysis, which included 509 bifurcation aneurysms treated with Y-SAC using low-profile stents, presented several important findings. Y-SAC is a technically complex procedure that necessitates catheterization of both bifurcation branches and the aneurysm sac. In our review, 95% of the treated aneurysms were managed using the crossing Y-configuration. This approach requires catheterization through the struts of the fully deployed first stent, followed by the deployment of the second stent through the struts of the first. Especially, for braided stents, concerns have arisen regarding catheterization through struts and the adequate expansion of the second stent at the intersection point [3, 8]. However, the overall technical success rate in our study was 99%, and it did not differ based on the structural or cell design of the stent (Table 4; Fig. 2). In total, approximately 90% of aneurysms achieved complete occlusion, and 98% had adequate occlusion at the last angiographic follow-up. Recanalization rates stood at 3%, and only 1.7% of the aneurysms required reintervention. Procedure-related mortality rates were below 1%. A good clinical outcome was observed in 98% of patients, and procedure-related morbidity developed in 1.3%.

In this study, we defined complete occlusion only if studies reported “complete occlusion” or “RR grade 1”. This was done to better assess and compare the efficacy of this technique with other available treatment options. However, two previous Y-SAC meta-analyses by Cagnazzo et al. [31] and Granja et al. [18], which included almost 100% non-low profile stents, described angiographic outcomes as complete/near complete occlusion. In our study, this was classified as adequate occlusion (RR 1 or 2). In the study by Cagnazzo et al. [31], the immediate and follow-up adequate occlusion rates were 82% and 95%,respectiely. In the study by Granja et al. [18], the follow-up adequate occlusion rate was 91%, and the procedure-related morbidity and mortality rates were 4% and 2%, respectively. Good clinical outcomes from these studies were reported at 94% and 92% [18, 31]. When comparing our findings to these studies, our safety and efficacy outcomes seem superior. These important findings suggest that low-profile stents enhance the efficacy of Y-SAC, along with technical success rates and safety outcomes.

There may be several advantages of low-profile stents in Y-SAC, regardless of their structural or cell design: (1) Low-profile stents can be utilized with microcatheters that have an internal diameter of 0.0165 inches [7,8,9]. Navigating through the acutely angled, small-sized side branches of bifurcations is easier and safer than with 0.021-inch microcatheters [7, 8]. (2) Another advantage of 0.0165-inch microcatheters is the absence of a step-off with the 0.014-inch microwire, simplifying navigation through stent struts [27]. This becomes more challenging with 0.021 or 0.027-inch microcatheters due to the ledge effect between the 0.014-inch microwire and the microcatheter [13]. Easier trans-strut navigation reduces the need for exchanges with the 300 cm microwire, minimizing technical complications. (3) Additionally, low-profile stents can be safely deployed in parent vessels with diameters less than 2 mm, offering benefits, especially when bifurcation branches are notably small [7, 21].

Low-profile stents with a laser-cut structural design, primarily the Neuroform Atlas, have been recommended and prioritized due to hybrid cell design, which offers an advantage in crossing Y-SAC [9, 10, 19, 21]. The central portion of this stent features an open-cell design, facilitating the easier catheterization of the second bifurcation branch and aiding the expansion of the second stent. However, our additional subanalyses of the studies that used solely laser-cut stents, and those that used solely braided stents, showed that braided stents, Leo Baby and Lvis Jr., were not inferior to laser-cut stents in terms of technical success. Procedure-related complications were reported at similar rates. Despite their smaller cell size, these braided stents feature sliding struts that allow for cell expansion and facilitate catheterization [7, 8, 13]. Using the technique of pushing the delivery catheter with the stent wire during the deployment of the second stent ensures optimal expansion at the intersection point [8, 13]. Moreover, braided stents are visible throughout their length and are resheathable up to almost 90% of their length, enabling more controlled deployment [7, 12]. Their braided design and enhanced metal coverage contribute to a flow diversion effect [7]. In our study, for the braided stent group, follow-up adequate occlusion rates were significantly higher. On the other hand, the immediate and follow-up complete occlusion rates were higher in the laser-cut stent group. These findings suggest that irrespective of the structural or cell design of the stents, the Y-configuration creates a sufficient flow diversion effect, contributing to progressive aneurysm occlusion [32,33,34]. Although initial angiographic outcomes were highly heterogeneous, this heterogeneity significantly reduced in follow-up angiographic results. In Y-SAC technique, coiling can be performed either by jailing the coil microcatheter before stent deployment or after the deployment of the stents using a coil-through approach. Some authors recommend using a single microcatheter for the entire process [27], while others suggest using two microcatheters for coiling after deploying the second stent [7]. Furthermore, many operators count on the progressive occlusion of the aneurysm by the flow-diversion effect, so they do not strive for tight coiling [33]. These approaches result in heterogenity in initial angiographic outcomes. In addition, Y-SAC can be performed without the need to cross the struts of the first stent by using a kissing Y-configuration [25, 28]. Although this provides a technically easier alternative, kissing Y-SAC has been associated with a higher risk of in-stent stenosis at follow-up [25, 35]. In our study, 10 cases of in-stent stenosis were reported, and 6 of these were associated with kissing Y-SAC. Parallel deployment of two stents in the parent artery may lead to insufficient deployment, which could subsequently result in-stent stenosis [25].

Recent prospective studies on the WEB treatment of wide-necked bifurcation aneurysms reported a complication rate of under 1% and long-term adequate aneurysm occlusion rates of 77.9% and 87.2% [36, 37]. However, aneurysms outside the 3–10 mm range or those with complex morphology, such as asymmetry between the aneurysm and bifurcation plane, poly-lobed shape, or incorporation of a bifurcation branch into the aneurysm base, may pose challenges to WEB treatment [12]. Despite a very safe profile, recanalization-related retreatment rates of 15.5% are still clinically noteworthy [36]. In our study, follow-up adequate occlusion rates were 98%, recanalization rates were 3%, and retreatment was 1.7%. Endovascular alternatives recently introduced for wide-necked bifurcation aneurysms include neck bridging devices such as PulseRider and pCONus. These devices provide scaffolding at the aneurysm neck to facilitate safe coiling. However, the lack of intrasaccular or intravascular flow diversion effects raises questions about their long-term durability. Additionally, recent meta-analyses indicate a long-term complete occlusion rate of 60% and a retreatment rate of 14% for pCONus [38], while PulseRider has a six-month complete occlusion rate of 64% [39]. Furthermore, a novel neck bridging device eCLIPs has an additional flow diversion effect; however, its application is more complex and may not be suitable for aneurysms with a neck size larger than 6 mm or those located at sites other than the basilar tip and internal carotid bifurcation [40]. In the current study, 64.4% of the included aneurysms were located at the MCA bifurcation and Acom locations, where using eCLIPS was not feasible.

Another technique for the treatment of wide-necked bifurcation aneurysms is flow diversion. However, the two main concerns are the insufficient neck coverage and the occlusion risk of covered branches [1]. In a meta-analysis, 244 MCA aneurysms underwent flow diverter treatment, with 76.3% of them located at the bifurcation. The study found an adequate occlusion rate of 78.7% after a median follow-up of 12 months. The rate of treatment-related complications was 20.7%, with nearly 10% of the covered branches occluded during follow-up [41]. While using two stents can make the procedure challenging, preserving both bifurcation branches—rather than covering one of them—is an advantage of Y-SAC. In our study, the overall rate of procedure-related complications was 10%, and there was only one case of asymptomatic branch occlusion during follow-up. Wide-necked bifurcation aneurysms, with branches frequently arising from or adjacent to the aneurysm neck, pose challenges for single-stent-assisted coiling. Complete occlusion rates for wide-necked bifurcation aneurysms treated with single stent-assisted coiling have been reported to range from 30.6 to 70.3% in follow-up studies [42, 43].  As the aneurysm size increases and the neck widens the success rate of complete occlusion with single stent-assisted coiling decreases, necessitating the Y-SAC [42].

This systematic review has several limitations. Except for one, all included studies were retrospective, and most were conducted at single institutions. Furthermore, most of the studies were small and lacked a comparison group. Angiographic outcomes were assessed by core laboratory adjudication in one study and by independent investigators in two others; the remaining studies either conducted self-adjudication or did not specify the independence of their assessments. Thus, potential observer bias and selection bias may impact the presented results. Variability in the definitions of morbidity, as well as in the categorization and definitions of complications across studies, presents a limitation; however, we have sought to mitigate this by detailing all reported complications and their outcomes. Subgroup analyses based on aneurysm size, rupture status, location, and patient comorbidities were not possible due to insufficient data. Nevertheless, this systematic review and meta-analysis could serve as valuable data for comparing the results of Y-SAC with this newer version of stents.

Conclusion

This systematic review and meta-analysis demonstrated that using low-profile stents for Y-SAC in treating wide-neck bifurcation aneurysms, under dual antiplatelet therapy, is effective and safe. The technique achieved a 99% technical success rate, a 98% rate of follow-up adequate aneurysm occlusion, and a procedure-related mortality rate of only 0.4%. Safety and efficacy outcomes are better when compared with older versions of stents. Notably, comparable follow-up angiographic outcomes were achieved with both braided and laser-cut low-profile stents.