Introduction The endovascular stent-assisted coiling approach for the treatment of cerebral aneurysms is evolving rapidly with the availability of new stent devices. It remains unknown how each type of stent affects the safety and efficacy of the stent-coiling procedure.
Methods This study compared the outcomes of endovascular coiling of cerebral aneurysms using Neuroform (NEU), Enterprise (EP), and Low-profile Visualized Intraluminal Support (LVIS) stents. Patient characteristics, treatment details and angiographic results using the Raymond–Roy grade scale (RRGS), and procedural complications were analyzed in our study.
Results Our study included 659 patients with 670 cerebral aneurysms treated with stent-assisted coiling (NEU, n=182; EP, n=158; LVIS, n=330) that were retrospectively collected from six academic centers. Patient characteristics included mean age 56.3±12.1 years old, female prevalence 73.9%, and aneurysm rupture on initial presentation of 18.8%. We found differences in complete occlusion on baseline imaging, defined as RRGS I, among the three stents: LVIS 64.4%, 210/326; NEU 56.2%, 95/169; EP 47.6%, 68/143; P=0.008. The difference of complete occlusion on 10.5 months (mean) and 8 months (median) angiographic follow-up remained significant: LVIS 84%, 251/299; NEU 78%, 117/150; EP 67%, 83/123; P=0.004. There were 7% (47/670) intra-procedural complications and 11.5% (73/632) post-procedural-related complications in our cohort. Furthermore, procedure-related complications were higher in the braided-stents vs laser-cut, P=0.002.
Conclusions There was a great variability in techniques and choice of stent type for stent-assisted coiling among the participating centers. The type of stent was associated with immediate and long-term angiographic outcomes. Randomized prospective trials comparing the different types of stents are warranted.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
A variety of endovascular approaches currently exist for the treatment of intracranial aneurysms with unfavorable anatomy precluding coiling alone, such as balloon-assisted coiling, stent-assisted coiling, or the use of flow diversion devices. The use of endovascular coiling began with the results of a large randomized control trial with the preliminary use of self-expanding stents shortly after.1–3 The devices are theoretically advantageous in blood vessels with difficult anatomy or wide-necked aneurysms. Additional devices become available to neurointerventionalist for the treatment of brain aneurysms with continued innovations in minimally invasive technology.
Stent-assisted coiling remains a widely popular option for the treatment of aneurysm, according to a recent survey of 211 neurointerventionalists.4 The two types of intracranial stents used for the treatment of aneurysms with stent-assisted coiling are laser-cut and braided stents.5 Most of the literature describes operator experience with a single stent system. Only a limited number of single-center studies have described comparative clinical performance of laser-cut versus braided stents.6–8 These studies have shown conflicting results when describing the occlusion, recanalization, and complication rates.
The goal of the present work was to compare the radiographic and clinical outcomes among patients treated with a stent-assisted coil device including Neuroform (NEU; Stryker, Kalamazoo, Michigan, USA), Enterprise (EP; Codman & Shurtleff, Raynham, Massachusetts, USA), or Low-profile Visualized Intraluminal Support stent (LVIS/LVIS Jr; MicroVention, Tustin, California, USA).
The study was approved by the local institutional review board at each participating center for retrospective data collection and review. Consecutive cases of intracranial aneurysms treated with the stent-assisted coiling using NEU, EP, or LVIS/LVIS Jr. between May 2009 and May 2018 were included in this study. Cases of aneurysms treated with more than one type of stents (NEU, EP, LVIS) used for coiling of the same lesion (n=24) or other devices (n=32) were excluded. Included and excluded data from our cohort of patients are available by request to the corresponding author. This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Treatment of aneurysms was based on operator preference, including timing of the procedure, choice and number of devices, administration and dose of antiplatelet agents, monitoring of antiplatelet responses, and post-procedure care. For elective cases, a typical antiplatelet regimen consisted of dual-antiplatelet therapy for the first 6 months, followed by long-term use of a single agent. A variety of treatment approaches including the use of IIb/IIIa agents and different loading doses of antiplatelet agents were used in cases of ruptured aneurysms. Baseline and post-procedural imaging studies including CT, MRI, and digital subtraction angiography were reviewed by local investigators at each participating site and were not adjudicated by a central core laboratory. The following data were collected: patient demographics, aneurysm characteristics, clinical presentation, procedural details and complication, and follow-up results of radiographic studies and functional outcomes. An aneurysm was considered wide neck if the size was greater than 4 mm or if the dome-to-neck ratio was less than 2. The degree of aneurysm occlusion was measured radiographically using the Raymond–Roy Grade Scale (RRGS; I-complete occlusion, II-neck remnant, III-residual sac).9 Clinical outcome was measured using the modified Rankin scale (mRS). mRS 0–2 was considered a good clinical outcome.
Statistical analyses included Fisher’s exact test, one-way ANOVA, χ2,, and multivariate analysis were performed using SPSS (IBM SPSS Statistics for Windows, v.25, IBM Corp., Armonk, N.Y., USA) and GraphPad Prism (version 7.99, GraphPad Software, La Jolla, California, USA). For statistically analyses, P<0.05 was considered statistically significant.
There was a total of 670 intracranial aneurysms treated with a stent-assisted coiling approach using one of the three stents (NEU, n=182; EP, n=158; LVIS/LVIS Jr., n=330). Aneurysm characteristics across stent groups are summarized in table 1. Mean age in the study was 56.3±12.1 years and 73.9% were female. In our study, 18.8% of patients presented with ruptured aneurysms. 72.7% were anterior circulation lesions, 58.0% were medium-sized aneurysm (5–15 mm), and 88.8% of treated aneurysms were considered wide neck using the definition of neck width >4 mm or dome to neck ratio <2. Aneurysm by anatomical side (right, left, midline) were evenly distributed between the three groups, 34%, 31%, and 34%, respectively. Stent configurations included a single device (87.7%), Y-configuration (9.7%), and telescope (2.6%) across the total cohort. Stent deployment techniques were classified according to the following categories: ‘jailing’ of the coiling microcatheter by the stent (39%); ‘coil-through’ by advancing the coiling microcatheter through the stent struts after stent deployment (38%); ‘coil-stent’ by coiling first and then deploying the stent at the end (9.8%); and ‘balloon-stent’ with balloon-assisted coiling attempted first followed by stenting (13.2%).
Including all patients undergoing stent-assisted coiling for aneurysm treatment, the type of stent was significantly associated with time of procedure (P<0.0001): the LVIS/LVIS Jr. was used at a greater proportion in most recent years compared with the NEU and EP stent in our cohort (see online supplementary table 2A–G). There was no association with procedural-related complication or post-procedural complication and time, P=0.059 and P=0.192, respectively. Furthermore, a larger proportion of complete occlusion based on RRGS baseline (P=0.023) and angiographic follow-up (P=0.001) were achieved in most recent procedural dates compared with earlier procedural dates. There was a significant association between type of stent and procedural data (year) in our unruptured (P<0.001) and ruptured (P<0.001) cohorts, see online supplementary tables 3A–E and 4A–E, respectively. A larger number of centers used the LVIS/LVIS Jr. stents for assisted coiling in more recent years compared with earlier.
The overall distribution of RRGS index at the end of the embolization procedure (baseline imaging) and on follow-up angiography according to stent type for the entire study cohort is shown in figure 1. The LVIS group was associated with the highest rate of complete occlusion on baseline imaging (RRGS 1: 64.4%; 210/326) compared with the other two stent groups (NEU 56.2%, 95/169; EP 47.6%, 68/143; P=0.008). A statistically significant difference of complete occlusion was also found on angiographic follow-up (LVIS/LVIS Jr. 83.9%, 251/299; NEU 78%, 117/150; EP 67.5%; 83/123; P=0.004). RRGS distributions on baseline and follow-up imaging according to stent type for unruptured and ruptured aneurysms are summarized in tables 2 and 3, respectively.
The relationship between patient characteristics, procedural complications, and outcomes based on stent type for patients with unruptured aneurysms (n=543) is summarized in table 2. There was no association between age, size of aneurysm, or multiple aneurysms (P=0.239, 0.643 and 0.758, respectively). The LVIS group was associated with higher proportion of use in anterior circulation aneurysms (79%) compared with other groups (NEU 67% and EP 72%, P=0.017). Descriptions of stent configuration and deployment techniques are presented in table 2. There was no significant difference among rates of angiographic occlusion on baseline imaging (P=0.198). However, there was a significant difference in RRGS indexes and stent type on follow-up (P=0.042). There was also a significant association between procedural-related complications and stent type (intraprocedural rupture occurred in 0.6% NEU, 0.8% EP, 3.5% LVIS/LVIS Jr. cases; stent thrombosis in 1.3% NEU, 0% EP, 3.5% LVIS/LVIS Jr.; stroke in 4.4% NEU, 0% EP, 0.8% LVIS/LVIS Jr. group; (P=0.002). There was no association in post-procedural complication rates of aneurysm recurrence, stroke, or stent stenosis based on stent type, P=0.815. Clinical outcomes in unruptured aneurysm patients was not analyzed due to limited use across all stent groups.
The relationship between patient characteristics, procedural complications, and outcomes based on stent type for patients with ruptured aneurysms (n=126) is summarized in table 3. There was no association between age, sex, size of aneurysm, wide neck anatomy, or multiple aneurysms status compared with each stent group in patients with ruptured aneurysms, P=0.565, 0.323, 0.350 and 0.237 respectively. There was a higher proportion of LVIS/LVIS Jr. stent use in anterior circulation aneurysms (81%) compared with NEU 46% and EP 50%, P<0.0001. The ruptured aneurysm group showed a significant association in anatomical location and stent type:midline aneurysms were treated with LVIS/LVIS Jr. in 53% compared with NEU in 28% and EP in 16% of cases, P=0.004.
There was a significant association between the baseline angiographic RRGS index and stent type (RRGS 1: LVIS 80%, NEU 52%, EP 42%, P<0.0001). Similarly, on angiographic follow-up, LVIS/LVIS Jr. group showed the highest compete occlusion rate (RRGS 1: LVIS 86%, NEU 63%, EP 58%, P=0.018). There was no association in the ruptured aneurysm cohort in intra- or post-procedural complications based on stent type, P=0.119 and 0.163, respectively). Clinical outcomes in ruptured aneurysm patients were unable to be calculated due to limitation in use across all stent groups within our ruptured aneurysm cohort.
The relationship between the RRGS index at baseline and follow-up based on all patient characteristics and complication rates is summarized in table 4. The median and mean of time to angiographic follow-up were 8 and 10.5 months, respectively, with an IQR of 6–12 months. The degree of occlusion at baseline angiography was significantly associated with age (P=0.002), location by circulation (P=0.002), aneurysm size (P=0.009), ruptured status (P=0.013), multiple aneurysm status (P=0.049), lesion side (P=0.007), configuration of stent (P=0.007), stent-coiling technique (p=0.001), and stent type (P=0.008). At follow-up, the degree of occlusion based on the RRGS index was significantly associated with aneurysm size (P<0.0001), multiple aneurysm status (P=0.002), wide neck anatomy (P=0.029), configuration (P=0.007), stent type (P=0.004), and occurrence of post-procedural complications (P<0.0001).
Across the nine dependent variables (age, type of stent, multiple aneurysm, rupture status, lesion location, aneurysm size, lesion side, configuration of stent, and stent deployment) the ANAOVA found statistically significant differences compared with angiographic baseline, all except configuration of stent remained significant in the multivariate analysis (see supplementary table 5A).
Across the six dependent variables (aneurysm size, multiple aneurysm, wide-neck anatomy, configuration, type of stent, and post-procedural complications) that ANOVA found statistically significant, the variables that remained independent predictors after multivariate adjustment of angiographic outcome at follow-up were aneurysm size, multiple aneurysm, and post-procedural complications (see supplementary table 5B).
Six centers participated in the final analysis. The mean and median number of patients included from each center was 117 and 86 (IQR 46–189). There was a significant association between an individual center and type of stent used in the procedure, P<0.001, (see online supplementary table 1A–C), approach to stent-assisted coiling (P<0.001) and stent configuration (P<0.001). The LVIS/LVIS Jr. stent was most used in three centers and NEU most used in two centers. Centers were significantly different among angiographic outcomes at baseline (P<0.001) and follow-up (P<0.001). Procedural-related and post-procedural complications were significantly different among centers in the unruptured aneurysm cohort (P=0.006 and Pp=0.02, respectively). However, procedural and post-procedural complications rates were not associated with center in the ruptured aneurysm cohort (P=0.338 and P=0.332, respectively).
The main finding of our study is that stent type affects the rate of angiographic occlusion immediately after the embolization procedure (baseline imaging), as well as at angiographic follow-up. Based on sub-analysis of our cohort by rupture status on presentation, the significant association of complete occlusion achieved with LVIS as demonstrated on baseline and follow-up imaging may be primarily driven by the high rate of complete occlusion in the ruptured aneurysms cohort. A similar difference in angiographic occlusion depending on stent type, with a highest degree of complete occlusion (RRGS I) using LVIS/LVIS Jr. devices was reported in a single-center study of 229 patients treated with three different stent types – laser-cut open cell design (NEU), laser-cut closed cell design (EP), and braided (LVIS/LVIS Jr) stent types.7
Several factors appear to be associated with the degree of angiographic occlusion in our cohort. A younger age, anterior circulation aneurysm, smaller aneurysm size at longest length, and the use of a braided device (LVIS/LVIS Jr. stent) appear to result in higher rates of complete occlusion compared with other groups at both baseline and follow-up imaging. LVIS/LVIS Jr. stents due to their braided design provide a substantially higher metal surface area (18%–23%) than laser-cut NEU and EP stents (7%–10%). Unfortunately, the findings of our study did not allow to determine whether superior angiographic occlusion results seen in the LVIS/LVIS Jr. group were secondary to higher metal coverage and thus potentially some additional flow diversion effect of braided stents or simply an increased packing density of the coil mass in conjunction with LVIS/LVIS Jr. Feng et al8 in a series of 142 patients with a total of 161 aneurysms found no significant differences in occlusion rates between stent-assisted coiling using EP and LVIS stent types. The authors did not address the rates of occlusion specifically for patients with ruptured versus unruptured aneurysms, presenting results for the entire cohort of patients.
Our study did not include the most recent generation of NEU stent family Neuroform Atlas. A single-center study by Gross et al compared the outcomes of patients treated with LVIS Jr. versus Neuroform Atlas stents. Higher rates of complete occlusion were seen in the laser-cut (Atlas) stent group than in the LVIS Jr.-treated patients. However, a small sample size (64 patients total) and single-center representation may limit the generalizability of these findings.
We also found substantial differences in the use of various types of stents for stent-assisted coiling, approaches to stent-assisted coiling, (‘jailing’, ‘coil-through’, ‘coil-stent’, and ‘balloon-coil’) as well as in immediate and follow-up angiographic outcomes among the centers included in our analysis. In our series, there were five US-based and one international center.
Our observations suggest that confirmatory studies are warranted when -single-center results become available reporting the safety and efficacy of neurointerventional devices. We also believe our data provides evidence that registries or, ideally, randomized trials designed to directly compare the performance of adjunct devices are needed to help guide the neurointerventionalist when planning for cerebral aneurysm embolization. A recent study by Turk et al10 that evaluated 28 North American and 15 international sites showed geographic differences in the treatment of aneurysms, also arguing that differences in practice patterns must be accounted for when comparing outcomes in patients undergoing complex neuroendovascular interventions such as intracranial aneurysm embolization.
There are several limitations to our analysis. One of the biggest limitations is that imaging results were not adjudicated by a central core laboratory, which may result in inaccurate interpretation of angiographic data in our cohort. In a recent study of bifurcation aneurysms treated by endovascular techniques where angiographic self-reporting versus core laboratory-adjudicated results were analyzed, on-site operator assessment was found to assign more favorable grades of aneurysm occlusion, indicating the importance of core laboratory—adjudication for accurate data reporting.11 Also, various imaging modalities (catheter angiography, as well as magnetic resonance or CT angiography) were used, potentially reducing the accurate assessment of angiographic occlusion rates at follow-up. Second, due to its multicenter retrospective data design, the results did not account for variations in perioperative treatment protocols and could also contribute to missing data points.
The type of stent used for stent-assisted coiling affects immediate and long-term angiographic outcomes. There are differences in the use of various types of stents for stent-assisted coiling, approaches to stent-assisted coiling, as well as in immediate and follow-up angiographic outcomes among the centers included in our analysis. Randomized prospective trials comparing the different types of stents are warranted.
Contributors MM, AT: study concept and design. MM, AT, and CTP wrote the manuscript. MM, AT, and CTP: statistical analysis. All authors participated in data collection and analysis, edited the manuscript, and approved the final version.
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.
Competing interests Mokin: reports consultant: Cerebrotech, Canon Medical, Imperative Care; stock options – Serenity Medical, Inc. Fiorella: reports grants, non-financial support, and other from Penumbra; non-financial support and other from Cerenovous and Medtronic; grants and non-financial support from Siemens and Microvention; and stock in Vascular Simulations and Neurogami. Rai: reports other from Penumbra; and non-financial support from Microvention and Stryker. Mocco: reports other from Penumbra, Cerebrotech, Rebound Therapeutics, TSP, Lazarus Effect, Medina, Pulsar Vascular, and Blockade. DeLeacy: reports other and non-financial support from Penumbra; and non-financial support from Cerenovus and Siemens. Turner: consultant for Penumbra, Medtronic, Microvention, Codman, Pulsar Vascular, Q’Apel, Rebound Medical, and Blockade Medical. Guilherme Dabus is a consultant for Medtronic, Microvention, and Penumbra and a shareholder in Surpass/Stryker, InNeuroCo, and Medina/Medtronic. Chaudry: reports other, grants, and non-financial support from Penumbra and Pulsar Vascular; grants and non-financial support from Medtronic, Microvention, and Codman; non-financial support and other from Blockade and non-financial support from Siemens. Turk: reports other, grants, and non-financial support from Penumbra and Pulsar Vascular; grants and non-financial support from Codman, Microvention, and Medtronic; non-financial support and other from Blockade; and non-financial support from Siemens.
Provenance and peer review Not commissioned; externally peer reviewed.
Patient consent for publication Not required.