Background Preliminary short-term results for stent-assisted coil embolization (SACE) using woven/braided stents have been promising. However, evidence supporting mid- to long-term efficacy and durability is lacking.
Objective To report the long-term results for the durability of elective intracranial aneurysms treated with woven stents.
Materials and methods Between May 2012 and May 2015, 98 consecutive patients with 103 aneurysms underwent elective woven SACE across three Australian neurovascular centres. All patients had immediate, 6- and 18-month clinical and radiological follow-up. Radiological assessment was performed with modified Raymond–Roy occlusion scores based on angiography results, while clinical assessment was based on the modified Rankin Scale.
Results Six-month follow-up was available in 100 aneurysms, and an 18-month follow-up in 97 aneurysms. Total occlusion rates of 82% were achieved at inception, 82% at 6 months, and 90% at 18 months. Satisfactory occlusion with small neck remnants was present in 17% at inception, 16% at 6 months, and 9% at 18 months. Good neurological outcomes were achieved in 95% at 18 months. Intraprocedural thromboembolic events were recorded in 3% and delayed events in 1% (all in patients taking clopidogrel). Aneurysm recurrence occurred in one patient (1%). Technical complications occurred in 5%. The total complication rate was 10%.
Conclusions Woven SACE is safe, efficacious, and durable at long-term 18-month follow-up, with very low recurrence and re-treatment rates. Preliminary results appear better than those for traditional laser-cut stents.
Statistics from Altmetric.com
Since the International Subarachnoid Aneurysm Trial,1 endovascular coiling has become the mainstay of treatment of cerebral aneurysms. However, treatment of wide-necked aneurysms remains challenging, given the risk of coil prolapse into the parent vessel. Stent-assisted coil embolization (SACE) was developed as one strategy to reduce this risk, by providing a permanent scaffold to hold coils within the aneurysm cavity.2 With SACE, higher long-term occlusion rates, and improved packing density compared with coil occlusion alone have been reported.3 4 However, most publications relate to laser-cut stents. In recent years, woven stents have been introduced, which differ from laser-cut stents in that they are braided from nitinol wires, which move independently of each other. This construction theoretically allows greater vessel conformity by the stent,5 facilitating better treatment of aneurysms located on curved vessel segments.6 7 It has also been proposed that woven stents may have a flow-diverting effect due to their greater metal surface coverage compared with their laser-cut counterparts. However, the safety and efficacy of woven SACE has not been confirmed by research. To address this, we present our immediate and long-term outcomes of aneurysms treated with woven intracranial stent-assisted coiling.
Between May 2012 and May 2015, 98 consecutive patients with 103 aneurysms underwent woven SACE across three Australian neurovascular referral centres. All patients treated for unruptured saccular cerebral aneurysms were included. Patients who presented with acute subarachnoid hemorrhage, or treated for aneurysms with a dissecting or fusiform morphology, were excluded. All patients were discussed in a multicentre neurovascular multidisciplinary team meeting, where both open and endovascular approaches were considered and consensus about treatment reached. Patients were then given the opportunity to give informed consent after being presented with all possible options, including no intervention, endovascular or open treatment. This is a retrospective analysis of a prospectively collected study cohort, and this study was ethically approved by the Australian National Health and Medical Research Council.
All patients were pre-medicated with dual antiplatelet medication. Between May 2012 and February 2014 inclusive, a single loading dose of 300 mg aspirin and 75 mg clopidogrel was given 7 days before their procedure, followed by 100 mg aspirin and 75 mg clopidogrel daily. From March 2014 onwards, clopidogrel was replaced with prasugrel, with a 30 mg loading dose 7 days before the procedure. This was followed by 5 mg prasugrel daily. Patients continued with their dual antiplatelet regimen until 6 months after the procedure, when clopidogrel or prasugrel was stopped after satisfactory evaluation with MRI. Aspirin was continued indefinitely. Compliance with pre-medication was verified by phone calls, and a routine pre-admission clinic appointment 2–5 days before the procedure.
Systemic heparinization was given during the procedure to reach a doubling of activated clotting time throughout the operation. This was followed by 24 hours of IV heparin infusion to maintain an activated partial thromboplastin time of 50–90 s.
All procedures were performed under general anesthesia. A 6- to 8-French guiding catheter was introduced through a femoral sheath into the carotid and/or vertebral artery. Evaluation and final measurements of the target vessel and aneurysm were achieved using both planar digital subtraction and three-dimensional rotational angiography. The choice of woven stents (Leo and Leo Baby (Balt Extrusion, Montmorency, France); Low-profile Visualized Intraluminal Support—LVIS and LVIS Jr (Microvention, Tustin, California, USA)) was based on operator preference and experience. The majority of stents were deployed through a 0.017-inch SL-10 microcatheter (Boston Scientific, Marlborough, Massachusetts, USA), with a minority using the 0.017-inch Headway 17 microcatheter (Microvention, Tustin, California, USA). The coiling microcatheter was routinely introduced into the aneurysm sac before stent deployment and ‘jailed’. All aneurysms were coiled as densely as possible at time of treatment. Figures 1–4 show examples of aneurysms treated in our series.
Demographic data, including age, gender, aneurysm data (including location, morphology, and size), stent devices used, and construct type were recorded in the study database. All intraoperative and/or postoperative complications were documented on the patients’ operation report. Follow-up notes were meticulously recorded. Filling of the aneurysmal sac was assessed by the senior author, and quantified according to the Raymond–Roy Occlusion Scale (RROS).
Patients were clinically assessed by the treating interventional neuroradiologist immediately after the procedure at 2–4 hours, at day 1 after the procedure, and at the 3-month clinic visit. Functional status was quantified at discharge, at the 3-month follow-up, and at the 18-month follow-up angiography visit, according to the modified Rankin scale (mRS).
Cranial CT and CT angiography were performed routinely on day 1 postprocedure to check for any evidence of thromboembolic sequelae and stent patency. Unless problems were identified warranting earlier imaging, the first follow-up was performed using 3T MRI with time-of-flight angiography, and/or gadolinium-enhanced MR angiography (MRA) at 6 months to determine aneurysmal filling, in-stent stenosis, or thrombosis. Six-month angiography was performed in patients with contraindications to MRI. The second imaging follow-up was obtained at 18 months with DSA and accompanying clinic review. For patients who were unsuitable for DSA, or physically or financially unable to undergo DSA, MRI/MRA was carried out instead. Patients were counseled on possible treatments for any aneurysm recurrences or remnants.
A total of 98 consecutive patients with 103 aneurysms were treated with SACE. Fifty-one patients were treated with clopidogrel and aspirin, and 52 with prasugrel and aspirin. At 6 months, 100 aneurysms were followed up with MRA (97%). Of the three aneurysms not followed up, two were in patients who had died and one was in a patient requiring high-level care, all unrelated to the aneurysm treatment. At 18 months, 97 aneurysms were followed up (94%). Our study has ethics approval to continue the follow-up of this cohort for 5 years after treatment. Of the three further patients without follow-up, one had age-related dementia and was receiving high-level care. Two others declined to return for follow-up, but were completely asymptomatic. Of these 97 aneurysms, 85 aneurysms (88%) had an 18-month DSA follow-up, and 12 (12%) had MRA follow-up. The patient cohort characteristics, aneurysm features, and treatment devices used are given in table 1. Aneurysm location, segregated into total results, and location, are shown in table 2. Of the 103 treated aneurysms, 73 were located in the anterior circulation (71%), and 30 in the posterior circulation (29%). All but two aneurysms were previously untreated; of these, one had a post-clipping anterior communicating aneurysmal remnant. The second was re-treated for coil compaction following treatment for a posterior communicating aneurysm. For both aneurysms, no complications were recorded intraprocedurally or postprocedurally, or at 18 months.
Satisfactory clinical aneurysm occlusion was achieved in 98% of the treated aneurysms at inception (n=101/103), deemed as aneurysms graded as RROS 1 or 2. Of these, 82% (n=84/103) were graded as RROS 1, with complete occlusion, and 17% (n=17/103) as RROS 2, with a small aneurysmal neck remnant. One aneurysm remnant each was graded as an RROS 3a or 3b at inception. At 6 months' imaging follow-up, 98% of the aneurysms were occluded (n=98/100). Of these, 82% (n=82/100) were graded as RROS 1, and 16% (n=16/100) as RROS 2. No aneurysm recurrence or new remnants were found at 6 months' follow-up. At long-term 18-month follow-up, 99% of aneurysms were deemed to be satisfactorily occluded (n=96/97), with 90% (n=87/97) graded as RROS 1, and 9% (9/97) as RROS 2. One case of aneurysm recurrence (1%) secondary to coil compaction was encountered in a partly coiled anterior communicating aneurysm treated with a complex Y-stenting configuration. This was scored as a RROS 3a with an 18-month DSA (RROS 3b immediately after the procedure at inception). This patient refused elective repacking despite coil compaction. This aneurysm later ruptured, and the patient died 2 years after the initial treatment (mRS score 6). This was the only death encountered in this series. In our series, there was no other aneurysm- or procedure-related mortality, no other aneurysms requiring re-treatment, and no difference in occlusion rates for different aneurysm locations. A summary of these total occlusion results is given in table 2.
Procedure-related thromboembolic events occurred in three patients (3%), all during the period when our department was prescribing oral clopidogrel. All three cases involved perioperative platelet aggregation, which was recognized, and treated with abciximab infusion intraoperatively. One patient developed transient hemiparesis, which recovered postoperatively with continued abciximab infusion over 12 hours. On discharge, this patient continued to have residual right upper limb mild weakness, which fully recovered. On 18-month follow-up, a mRS score of 0 was recorded. The second patient similarly developed transient hypobulia and left hemiparesis postprocedure, which fully recovered with intra-arterial abciximab. There was no neurological deficit, with a long-term mRS score of 0. This patient was later found to have been non-compliant with her clopidogrel before her procedure. The third patient developed a small left hemispheric embolic event postprocedure, but no demonstrated stent thrombosis, resulting in right-handed weakness, which responded well to rehabilitation and occupational therapy. This procedure was technically difficult, and the only patient in our series who underwent treatment in two stages, initially with a Y-stent construct, followed by coiling 4 months later, secondary to technical difficulties opening one of the Y-construct limbs. Six-month follow-up MRI results demonstrated a 5 mm ischemic focus within the deep white matter of the left frontal lobe. A long-term mRS score of 2 was recorded.
One delayed stent thrombosis was encountered (1%), also in a patient taking clopidogrel. This patient had an anterior communicating artery aneurysm, and no ischemic foci were present on the 6-month MRI scan. The 18-month DSA scan demonstrated occlusion of the proximal limb of the stent in the A1 segment, with good distal A2 flow due to collateralization, and the patient was asymptomatic (mRS score 0). Since the substitution of prasugrel for clopidogrel, no further cases of thromboembolic events have occurred. Additionally, we recorded no complications with acute or delayed localized or systemic hemorrhages with the use of prasugrel.
Technical complications occurred in five instances (5%). In one patient, on deployment, the Leo Baby stent migrated slightly, requiring the addition of a Solitaire stent as an anchor. There was no resultant neurological deficit, and the patient had an mRS score of 0 at 18 months. In a second patient, friction between the stent delivery catheter and the jailed intra-aneurysmal catheter resulted in sudden forward movement of the aneurysmal microcatheter through the aneurysm dome. This patient had an mRS score of 3 at the 18-month follow-up (mRS score 0 before the procedure). The third patient had an asymptomatic wire perforation and low-volume subarachnoid hemorrhage during placement of the second stent of a T-stent construct, requiring balloon inflation for hemostasis. No resultant neurological deficit postprocedure or at long-term follow-up was seen and an 18-month mRS score of 0 was recorded. In a fourth patient, the distal end of a Leo Baby stent migrated into a lateral mesencephalic branch of the first part of the posterior cerebral artery during stent deployment. Repositioning of the device was not possible owing to a significant risk of major vessel rupture under dual antiplatelet cover. The resultant midbrain stroke was confirmed by postoperative MRI, and a long-term mRS score of 3 recorded (mRS 0 before the procedure). No other intraoperative or delayed stent-related complications were encountered. The fifth patient, previously described, required a second-stage coiling procedure (mRS score 2).
Higher long-term occlusion rates and improved packing density compared with coil occlusion alone are reported benefits of SACE.3 4 There is a suggestion in a recent meta-analysis that using stents promotes progressive aneurysm thrombosis (37.5%) compared with coiling alone (19.4%, P<0.00001), and produces a lower recurrence rate (16.2%, SACE vs 34.4%, coil occlusion; P<0.00001).8 However, all first-generation stents for aneurysm repair were laser-cut, and differ fundamentally from the newer woven stent devices, which are less well characterized.2 5 This includes their inability to be re-sheathed and repositioned when partially delivered, their low radial force, and difficulties with visualization with only proximal and distal markers.9
Some authors have noted the greater advantages of woven stents in comparison with laser-cut stents. This includes their closed-cell design with a compliant cell size, enabling better wall apposition, and scaffolding compared with traditional laser-cut stents.10 Their ability for re-sheathing and repositioning up to 95% of their length for Leo/Leo-Baby stents, and 80% in LVIS/LVIS Jr stents provides another additional advantage over most laser-cut stents.7 11 It has been proposed that woven stents have better flow-diverting properties owing to their greater metal surface coverage than their laser-cut counterparts, which theoretically contribute to aneurysmal occlusion. However, they have also been criticized for higher incidences of thromboembolic complications,12 which also theoretically stem from the larger surface of foreign body exposed to the bloodstream.
Analysis of the safety and efficacy of woven SACE has been promising for short-term follow-up, but mid- to long-term results have not been fully explored. Of published articles that reported clinical results for woven SACE, several discussed perioperative experiences and results only, with no short or long-term follow-up data,13–15 while a few other studies had follow-up of ≤6 months.6 7 10 11 16 17 The majority of these short-term follow-up results indicated generally positive clinical outcomes, although radiological results were mixed, with one study reporting an 86.7% in-stent stenosis rate at 6 months.16 Of the studies that reported some follow-up results of >6 months, a number did not separate them into short-term results of ≤6 months and longer-term follow-up results, making the results difficult to interpret.9 18–24 Another study reported long-term outcomes of SACE, but only 3% with woven SACE, and did not report outcomes separately for this group.25 Only three studies in the literature reported 12-month follow-up results for woven SACE. The first was for 26 aneurysms. However, these results included aneurysms treated with coil embolization alone, with no separation of results into SACE and coil embolization groups.26 Another reported 43 aneurysms, with generally positive results, but only 67% had a 12-month angiographic follow-up, and a further 28% had a 6-month angiographic follow-up, and the results were not separated into mid- and short-term.27 Four studies also reported results of electively treated aneurysms together with ruptured aneurysms, thereby creating a bias in the results because two patient populations with markedly different underlying prognoses were treated.9 18 20 21
Our initial experience to date with woven SACE demonstrates a 1% total recurrence rate at both the 6 and 18 month follow-up in 103 aneurysms. Additionally, we achieved satisfactory occlusion rates of 98% at inception, 98% at 6 months, and 99% at 18 months, with patients scoring RROS 1 or 2. Even taking total occlusion rates (RROS 1) into account, our success rates were 82% immediately postprocedure, 82% at 6 months, and 90% at 18 months. These results are comparable to those reported in other studies for woven SACE follow-up, noting immediate complete aneurysmal occlusion rates in 73–85%, and 82–90% at a follow-up of approximately 6 months.7 11 ,18 Feng’s study on 97 aneurysms treated with LVIS stents was an outlier, achieving complete occlusion immediately postoperatively in only 28.8%, neck remnant in 40.2%, and at an 8.1-month mean follow-up, 84.2% with complete occlusion, neck remnant in 11.8%, and residual sac in 4% of patients were achieved. It is uncertain why poorer occlusion rates were achieved at inception. Recurrence rate was reported as 0%, although only 78% had follow-up results.24 Aydin et al reported recurrence in 6.5% of patients, with 5.2% requiring re-treatment by 6 months' follow-up.18 Alghamdi et al reported recanalizations in 5% of patients with 6- or 12 month follow-up angiographic results.27 Our occlusion rates and recurrences are better than those reported in the literature for laser-cut stents. A meta-analysis of 2174 patients reported immediate occlusion rates of only 57.6%,8 and a 16.2% recurrence rate recorded during a similar follow-up period to that in our study.4 8 Another similar meta-analysis of 696 patients, predominantly patients treated with laser-cut SACE, noted a 45% complete initial occlusion rate, 61% at various short-term follow-ups, and 13% of these requiring re-treatment due to recurrences.28 Additionally, Lopes et al studied the outcomes of 410 patients treated with laser-cut SACE, and reported a complete occlusion rate of 43.3% at inception, which increased to 72.9% at 6 months.29 These differences in occlusion rates between woven and laser-cut SACE are highlighted by Ge et al, who compared 96 aneurysms treated with LVIS stents, and 112 treated with Enterprise stents. At inception, 97.1% occlusion rates were achieved in the woven SACE group, and 88.4% within the laser-cut SACE cohort (P=0.034). However, interpretation is limited due to lack of follow-up results, with only 39.1% of patients receiving angiographic follow-up, some with only 1-month results. Follow-up occlusion rates were also not reported.23
Our study provides the first and firmest evidence to date that woven SACE is safe, efficacious, and durable at an 18-month follow-up, with results appearing better than those obtained with laser-cut stents. We report an overall complication rate of 10% in our total cohort to date, including immediately postoperatively, and at the 18-month follow-up. Permanent neurological deficit occurred in 4%. The overall technical complication rate with no clinical morbidity was 6%. It is interesting to note that the four thromboembolic events that occurred were all during use of clopidogrel, rather than prasugrel, as the second antiplatelet agent. This is supported by preliminary evidence that prasugrel is better than clopidogrel. Sedat et al compared two cohorts, one treated with aspirin plus clopidogrel, and the other with aspirin plus prasugrel. Although the number of thromboembolic events did not differ statistically, clopidogrel appeared to be associated with three times more serious statistically significant thromboembolic events, such as complete stent thrombosis at 30 days30. Overall, our complication rate compares favourably with rates reported in the literature, with meta-analysis compiling total complication rates of 10.9–17.6%, with a 6.8–9.1% mortality rate, thromboembolic complication rate of 4.2%, and permanent complication rate of 3.8–5.6% recorded8.
Our series has several limitations. Our cohort is of a mixed retrospective and prospective nature, with imaging data of patients treated before 2014 retrospectively reviewed, while clinical and imaging data after 2014 were prospectively evaluated. Future prospective multicentre data collection and independent analysis would be valuable. Another limitation of this study is its non-randomized, unblended nature. For a truly accurate assessment of safety, efficacy, and durability of woven SACE versus laser-cut SACE, a randomized controlled trial comparing the two classes of device would be required. Additionally, larger-scale multinational collaboration and longer-term follow-up would help to verify our 18-month experience.
Our study, which represents the largest consecutive patient series treated with woven SACE to date, has shown that this treatment is a safe, efficacious, and durable means of aneurysm repair. Results are durable up to 18 months. Our results also suggest that aneurysms with woven SACE are more likely to be completely occluded at the time of treatment, and less likely to recur. Further prospective large-scale multicentre data and comparison of laser-cut and woven stents are needed to verify these findings.
Contributors All authors made substantial contributions to the design of the study, analysis and interpretation of data for the work. All authors drafted and/or revised it critically, and provided final approval of the version to be published; and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Competing interests None declared.
Patient consent Obtained.
Ethics approval National Health and Medical Research Council, Australian Research Council.
Provenance and peer review Not commissioned; externally peer reviewed.
Correction notice Since this paper was first published online the author affiliations have been updated.
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.