Stent-assisted coiling of cerebral aneurysms with the Neuroform Atlas stent

Abstract

Objectives

The Neuroform Atlas stent (AS) is the smallest intracranial stent with an open-cell design. This study reports the first clinical experience with AS.

Methods

All intracranial aneurysms treated by stent-assisted coiling using a single AS in a single institution were retrospectively evaluated. Patient demographics, aneurysm characteristics, angles between the parent artery and stented branch, technical success, and clinical and angiographic follow-up were analyzed.

Results

Fifty-five consecutive aneurysms treated with AS-assisted coiling were included. Of these, 69.1% were located distal to the circle of Willis. Technical success rate was 100%. The mean diameters of proximal and distal parent arteries were 2.62 mm (range 1.5–4.4) and 1.8 mm (range 0.8–3.5), respectively. Except for a minor stroke in a patient who completely discontinued antiplatelet therapy on postoperative day 4, there were no clinical events with permanent sequelae, and 94.1% of patients had Raymond-Roy score of 1 or 2 aneurysmal occlusion at a mean follow-up duration of 7.9 months. Although the angle between the parent artery and the stented branch increased significantly (p < 0.001) with time, the angular change at follow-up was only 16.45 ± 11.03 degrees and was inversely correlated both with preoperative angle and the diameter of the distal parent artery (r = −0.465 and r = −0.433, respectively, p = 0.004 for both).

Conclusion

AS-assisted coiling was associated with a favorable early clinical outcome and angiographic results in this series. This stent can be used for distally located aneurysms and results in minimal alteration of the arterial anatomy.

Introduction

Stent-assisted coiling (SAC) of cerebral aneurysms has been performed for more than 20 years. Initially, coronary stents were used for SAC. In 2002, the first dedicated self-expanding stent, the Neuroform (Stryker, Kalamazoo, MI, USA) and in 2003, the Leo stent (Balt, Montmorency, France), were introduced to the market as open-cell (OC) and braided (BR) intracranial stents deliverable through a microcatheter. These were followed by the closed-cell (CC) Enterprise (Codman, Raynham, MA, USA) and Solitaire (Covidien, Irvine, CA, USA) stents. Recently, stents deliverable through microcatheters with an internal diameter of 0.0165 inches have become available. In addition, studies have been published with the miniature versions of the BR stents (Leo Baby, Balt) and the CC stents (Acclino stent; Acandis, Pforzheim, Germany). To our knowledge, this is the first report on the recently introduced small-sized Atlas (Stryker, Kalamazoo, MI, USA) stent, which was designed as an OC intracranial stent.

Materials and methods

We retrospectively evaluated all intracranial aneurysms treated by SAC with a single Atlas stent in a single medical institution between August 2015 and June 2017. Patients were excluded if the stent was used as an adjunct to another stent (such as stent placement within another stent or flow diverter to help with the device opening or to salvage a complication that occurred during the placement of these devices) or if dual stenting had been performed. The following variables were analyzed: patient demographics, aneurysm location; diameters of aneurysm dome; aneurysm neck; proximal and distal artery at the stented segment; angles between the parent artery and stented branch on pre-procedure, post-procedure and follow-up angiograms; technical success in placement of the stent; initial aneurysm occlusion rate based on the Raymond–Roy score (RRS) and clinical/angiographic follow-up. All of the angiographic measurements, including the length and angles, were obtained using the automated measurement system of the angiographic workstation (Siemens Healthcare, Erlangen, Germany), and angiograms obtained on the working projection were used for these measurements. Angular measurement was performed only for bifurcation aneurysms. If follow-up digital subtraction angiography (DSA) was not available or if the side branch was obscured because of coils on postoperative or follow-up angiograms, volume-rendered magnetic resonance angiography (MRA) images replicating the working projection were used for the angle measurements. Angular measurements were performed as described previously in the literature.¹ An example of the angular measurement on one of our patients is shown in Supplementary Figure 1.

In our institution we frequently use mini-stents for electively treated distal bifurcation aneurysms. Because of its easy navigation and deployment, we used the Atlas stent very often in such aneurysms. We also used other stents, namely the Leo Baby stent (Balt, Montmorency, France) and Enterprise stent (Codman Neuro, Codman & Shurtleff, J&J, Raynham, MA, USA) during the study period whereas flow diversion was reserved primarily for proximal aneurysms and for sidewall/fusiform aneurysms. The other stents were utilized for SAC in the following situations.

1. Bifurcation aneurysms in which we expect to have suboptimal coiling result or recanalization since wall coverage of the Atlas stent is too low to enhance neointimal coverage.

2. When an intraaneurysmal loop technique is needed to bypass the neck. We do not prefer distal anchoring for intraaneurysmal loop reduction with the Atlas stent as it non-retrievable. Additionally, in case of recurrent treatment, neck bypass for the second time is easier with BR stents owing to their tight interstices.

3. Complex bifurcation aneurysms in which flow diverters may be needed for possible recurrence. We are concerned that the thorn-like ends of open cells may catch the struts of the BR stent, and, additionally, the absence of a marker in the body of the stent makes it hard to estimate how the flow diverter will expand within the first stent.

4. T stenting where our default stent for the side branch is a BR stent.

General description of endovascular procedure

Ethical approval was obtained from our local hospital’s institutional review board, and written informed consent was obtained from all patients before the procedure. Patients were pretreated with 75 mg clopidogrel and 300 mg acetylsalicylic acid daily at least five days before the procedure. A platelet function test (VerifyNow P2Y12 assay; Accumetrics, San Diego, CA, USA) was performed to ensure a good response to clopidogrel. Patients were switched from clopidogrel to prasugrel and maintained on 10 mg per day if they were hypo-responders to clopidogrel. Under general anesthesia, anticoagulation was initiated with a bolus dose of 5000 IU of intravenous heparin after the femoral access followed by heparin infusion with the goal of maintaining the activated clotting time of approximately twice above the baseline value. A 6F guiding sheath was inserted in the target artery, and a distal access catheter was navigated into the intracranial internal carotid artery or distal V2 segment. Jailing technique was used for SAC of the aneurysms. The aneurysm sac was catheterized with an Excelsior SL-10 (Stryker, Neurovascular, Fremont, CA, USA) microcatheter and a soft-tip microguidewire. A second microcatheter was navigated across the neck of the aneurysm. After the stent deployment across the aneurysm neck, coiling was performed with detachable bare coils until total embolization was achieved or when it was no longer possible to safely place coils within the aneurysm sac.

Follow-up

Immediate post-procedure angiograms obtained at the end of the embolization procedure were evaluated for aneurysmal occlusion according to the RRS. Periprocedural and follow-up clinical data and technical or clinical complications were extracted from procedural reports and hospital records. Complications that occurred within 30 days were defined as perioperative and those occurring later were defined as delayed. After the procedure, dual-antiplatelet treatment (DAPT) was continued for six months. Although six months is a relatively long duration for DAPT when single stent-assisted coiling with laser-cut stents is utilized, we prefer to err on the side of bleeding complications rather than possible thromboembolic events during follow-up. Then, clopidogrel was discontinued and the patients were kept on lifelong 300 mg per day aspirin. Patients were called back for an early follow-up MRA at 6 to 12 weeks. A second angiographic follow-up was performed at 6 to 12 months with DSA. Then, annual MRA follow-up was performed. Follow-up angiograms were assessed for stent patency and aneurysm occlusion status (RRS).

Statistical analysis

The SPSS 20.0 (IBM Corp, Armonk, NY, USA) program was used for statistical analysis. For the comparison of angular change in percentage between groups, the Mann-–Whitney U test was used. Pearson and Spearman correlations were utilized to assess the relationship between the angulations and arterial diameters, branch angulations or categorical variables including age and sex. Statistical significance was set at p = 0.05.

Results

A total of 48 patients with 55 consecutive aneurysms, which were treated with a single Atlas stent and coiling, were included in this study. Examples of two of these cases are shown in Figures 1 and 2. Thirty patients were female and 18 were male. The average age was 51 (range 23–74) years. Aneurysm locations are listed in Supplementary Table 1. Eight of the treated aneurysms were previously ruptured, but none of these were treated within the acute period. The mean sizes of the aneurysm dome and neck size were 7.1 mm (range 2–17) and 3.8 mm (range 1.5–8.7), respectively. The mean sizes of the proximal and distal parent artery were 2.62 mm (range 1.5–4.4) and 1.8 mm (range 0.8–3.5), respectively. Thirty-eight of the 55 aneurysms (69.1%) were located distal to the circle of Willis.

Figure 1.

(a) Dual aneurysms of the left middle cerebral artery on three-dimensional (3D) images in working projection obtained two months after subarachnoid hemorrhage. (b) Another projection in 3D displays the neck of the distal aneurysm. (c) Angiogram in working projection. (d) Native image obtained during endovascular treatment shows that primarily coiling of the distal aneurysm has been performed and the proximal aneurysm is being treated with Atlas-assisted coiling. Post-procedure angiograms in working projection (e) and in 3D view (f) show some contrast material permeation within the aneurysms. (g) Total occlusion of both aneurysms and patency of the stent on 12-month control angiogram.

Figure 2.

(a) An unruptured right middle cerebral artery aneurysm is noted on three-dimensional angiogram in ipsilateral oblique Waters projection; the superimposed temporal branch was manually removed during post-processing to disclose the aneurysm neck. (b) Angiogram in the same projection. (c) Fluoroscopic capture obtained immediately after deployment of the Atlas stent. (d) Post-procedure and six-month follow-up (e) angiograms reveal complete occlusion of the aneurysm and patency of the stent.

Neuroform Atlas stents were deployed through SL-10 (Stryker), Echelon 10 (Medtronic Covidien, Irvine, CA, USA) or Vasco 10 (Balt, Montmorency, France) catheters or the Eclipse 2L (Balt) balloon. Deployment was successful for all of the treated aneurysms. We had problems with loading of the stent into the microcatheter with the initial version of the delivery sheath and we had to retrieve four stents from the hub of the microcatheter and use a second stent. However, when we intended to treat a cerebral aneurysm with an Atlas stent and coiling, the procedure was completed as intended. We did not encounter this problem with the newer version of the Atlas delivery sheath.

There were no periprocedural permanent complications or mortality related to the procedure. On delayed follow-up, one patient expired secondarily to an unrelated cause, and in one patient the follow-up magnetic resonance imaging (MRI) showed enhancing intracranial lesions consistent with a foreign body reaction without neurological deterioration; the occurrence and characteristics of these lesions were similar to those previously reported.² Otherwise, there were no procedure-related adverse events during follow-up in patients who reported to be compliant with their antiplatelet treatment (APT). Among two patients who admitted their non-adherence to APT, a clinical event occurred in one patient during follow-up. This patient never used clopidogrel or aspirin after discharge from the hospital on postoperative day 3; hence, this patient developed a small focal visual field deficit secondary to an occipital stroke two months after the procedure. The patient was again started on DAPT and currently reports no symptoms. The second patient’s DAPT was stopped for one week against medical advice for elective surgery after the treatment of the aneurysm. Anticoagulation was not started again postoperatively by the attending surgeon either. The stented upper trunk of the middle cerebral artery was incidentally noted to be obliterated during the six-month follow-up visit. The patient remained asymptomatic. There were two more cases of severe in-stent stenosis noted incidentally at the six-month follow-up visit. Both were managed medically. Otherwise, all of the remaining stents in patients presenting for the DSA follow-up were fully patent. On final imaging follow-up (51 aneurysms), at a mean follow-up duration of 7.9 months, 48 aneurysms (94.1%) showed acceptable obliteration (RRS of 1 or 2), two aneurysms were retreated and one was followed up with MRA for recurrent aneurysm filling. Three patients are doing well but refused to return for an imaging follow-up; one patient without imaging follow-up expired secondarily to an unrelated cause (accidental death). As for early DSA follow-up, at a mean follow-up duration of 8.7 months, 43 aneurysms had DSA and 36 of these were classified as RRS 1, whereas four had RRS 2 and three had RRS 3 aneurysm obliteration yielding an acceptable occlusion (RRS 1 + RRS 2) rate of 93%. DSA follow-up is pending for eight aneurysms.

Side-branch angular calculations were possible only for bifurcation aneurysms. Of the 43 bifurcation aneurysms that were eligible for angular measurements, one was omitted from the calculation because the coils obscured the visibility of the side branch both on post-procedure DSA and follow-up MRA. In the other five aneurysms, a calculation was not possible because the side branch had occluded asymptomatically by the time of follow-up in one patient and four patients who were doing well clinically did not come back for imaging follow-up. For the remaining 37 aneurysms, 32 DSA and five MRA studies were utilized for angular measurements. The median preoperative branch angulation was 80 degrees (83.05 ± 30.72). The postoperative median vascular angle was 95 degrees (94.4 ± 29.33). On follow-up, the median angle was 100 degrees (99.78 ± 27.43). Analysis of the angular difference between the preoperative versus postoperative studies and preoperative versus follow-up studies showed a statistically significant difference (p < 0.001). Similarly, the difference between postoperative versus follow-up angles showed a significant difference (p < 0.001). Although the angular change showed a statistically significant difference, the absolute difference between the mean values of these angles was only 10.69 ± 9.93, 5.91 ± 6.74, and 16.45 ± 11.03 for preoperative versus postoperative, postoperative versus follow-up, and preoperative versus follow-up studies, respectively. There was a negative correlation (r = –0.465) between the preoperative angles and the degree of angular change (the difference between the angular measurement on follow-up and the preoperative angle in each individual), which was statistically significant (p = 0.004). Similarly, the degree of angular change was negatively correlated with the diameter of the distal parent artery (r = −0.433), and this was also statistically significant (p = 0.004).

Discussion

Although thousands of SAC procedures have been performed since the introduction of this method 20 years ago,³ a randomized comparison of the different stent designs has not been published; hence, there is currently no consensus on an optimal stent design for intracranial use. There have been retrospective studies comparing Enterprise stents with either Neuroform or Solitaire or Low-profile Visualized Intraluminal Support (LVIS) stents (MicroVention-Terumo, Tustin, CA, USA),^4–9 and there was no significant difference in final clinical outcomes among patients treated with either of these stents versus the other. However, the currently available intracranial stents do differ in their physical characteristics.^10,11

Recently published studies have suggested that the differences in physical properties of the first-generation intracranial stents may be related to specific complications. The larger delivery system of the OC Neuroform stent is associated with a higher rate of technical failure.^5,6,9,12 The higher rate of recanalization^5,9,12,13 with this type of stent was attributed to the gator backing phenomenon.⁹ The CC design (i.e., Enterprise) is associated with ovalization, which may result in a higher rate of thromboembolic events.^9,12,14 The hourglass formation is specific to the BR stent (i.e., Leo and LVIS stents) and has been reported in up to 19% of patients treated with the LVIS stent. BR construction results in foreshortening³ during stent deployment, whereas after deployment, outward radial force and cell design influence the stability of the stent and result in a higher likelihood of delayed migration in CC stents.⁸ While smaller interstices and high wall coverage are associated with a low recurrence rate,^4,15 smaller cell size is also associated with limited re-entry into the aneurysm,^16,17 jeopardizing retreatment. Wall apposition is better with BR devices,^18,19 yet this is technique/operator dependent and optimal apposition does not invariably occur.^11,20 Arterial angular remodeling is also more prominent with CC stents than with others.¹ Finally, angiographic visibility is better with BR devices.³ As a result of these differences, the operator may favor one stent over another for a specific aneurysm.

Among the first-generation stents studied in vitro, the OC design has demonstrated a relatively low radial force, large cell size, average chronic outward force, good wall apposition/wall coverage and low bending stiffness.^10,11 That is, this design conforms to the vascular anatomy quite well and does not significantly alter the vascular geometry. This property most likely relates to the lower rate of thromboembolic events at the cost of a higher rate of aneurysm recurrence. We prefer the OC Atlas stent at the level of the circle of Willis and beyond, as access with a small (0.0165-inch inner diameter) microcatheter and the Atlas is straightforward and we prefer to keep the manipulation of the arterial geometry in these distal arteries at a minimum.

Our experience suggests that the characteristics of the Atlas stent are similar to its predecessor; yet, the clinical and angiographic results are better. In the multicenter prospective Safety and Efficacy of the Neuroform 3TM Stent for Aneurysm Treatment (SENAT) study, in which a single Neuroform stent was used in 95% of 107 aneurysms,²¹ the permanent morbidity related to the procedure was 1% and 12% of aneurysms were graded as residual aneurysms on follow-up. In our cohort, there were no procedure-related deaths and the only morbidity was a minor stroke related to the procedure. If we exclude the patients who did not take any APT medication after discharge, there were no deaths or permanent complications in patients adherent to APT. The rate of insufficient aneurysmal occlusion, including retreatments, was 5.8%. In contrast to the SENAT study, in which 55% of the aneurysms were side-wall internal carotid aneurysms, carotid side-wall aneurysms constituted only 9% of the aneurysms in our study and there were more middle cerebral bifurcation/anterior communicating artery aneurysms (59% in this report versus 34% in SENAT) in our cohort. The ability to treat the more challenging distal aneurysms as safely and effectively as proximal ones is a consequence of improvements in navigability and conformability of the stent. The absence of a distal wire may have been a factor in the absence of intracranial hemorrhages related to stent deployment.

The results of this series are similar to those of recently published studies on the CC or BR stents, which can be deployed through small-bore microcatheters. In three retrospective studies published on the outcome of the CC Acclino stent and a larger contemporary series regarding the smaller BR stents (Leo Baby and LVIS Jr), the rate of death or permanent neurologic complications ranged from 0% to 7% and 2.4% to 3.8%, respectively.^15,22–26 Although the rate of acceptable (RRS 1 and 2) aneurysm occlusion in our study was similar to the results reported in these manuscripts, the rate of neck remnants at final follow-up was higher in our study. One plausible explanation for this discrepancy may be the larger pore size and gator backing, which are known characteristics of the OC design.^10,11 In our series, postoperative progression of a neck remnant to complete occlusion occurred in two aneurysms, and none of the residual aneurysms progressed to neck remnants or total occlusion. It is unlikely that the Atlas stent has a considerable flow-diverting effect. Another explanation is that a portion of patients were treated with dual (X, Y, telescopic) stenting in these series, as well as with other devices during the same session (e.g., balloons and intrasaccular flow diverters in the reports on Acclino stents),^22,24 which increased the angiographic occlusion rate in these reports and precluded a pure assessment of the effects of the specific stent.

Our angiographic measurements demonstrate that the Atlas stent is associated with a statistically significant yet clinically negligible degree of angular remodeling, indirectly suggesting the conformability of the device that is not angiographically visible. Our findings suggest that remodeling occurs mostly in the perioperative period and further remodeling occurs less in the follow-up period when the Atlas stent is used. This finding is in contrast to previously reported results that indicated a considerable amount of angular modification (approximately 30 degrees, up to 74 degrees). As in those reports, the remodeling affected smaller arteries and those arising with acute angles more than the larger, proximal arteries.^27–29 Our results further suggest that the smaller side branches show angular change sooner (e.g., the end of the procedure), while the angular change in the larger branches occurs later, during follow-up. Notwithstanding the fact that such modifications are allegedly an indicator of a higher likelihood of aneurysm occlusion during follow-up,^27,30 they may also be associated with stent migrations in approximately 2% of SAC cases, especially with CC stents.²⁸ Excessive and sustained forces applied on the bifurcations of similarly sized coronary arteries are also known to result in side-branch occlusions.^31–33 The consequences of such a phenomenon in distal cerebral arteries remain unexplored and may potentially result in branch occlusion.¹ On the other hand, stents with better conformability, like the Atlas stent, are associated with a lower thromboembolic event rate.³⁴ In addition, they enable easier aneurysmal re-access during coiling and side-branch access through the large pores for dual stenting³⁵ and finally, have a very low rate of stent migration.³⁶

There were no remarkable clinical or technical events when the Atlas stent was placed in distal parent arteries less than 2.0 mm in diameter, which, per the instructions for use of the Atlas stent, is the smallest diameter indicated for treatment with this stent. The diameter of the stented distal arterial branch was as low as 0.8 mm in this cohort; an example is demonstrated in Supplementary Figure 2.

The limitations of our study include its retrospective design and the lack of long-term follow-up for some of our patients. Additionally, it would have been preferable to evaluate the geometric modification in three dimensions (3D) utilizing fusion software. As this was a retrospective study and 3D angiograms were not routinely obtained postoperatively, such data are not available for evaluation.

Conclusion

Aneurysm embolization with the OC design Neuroform Atlas stent was associated with favorable clinical and angiographic outcomes in this series. There was minimal alteration of the vascular anatomy by the stent, which suggests its conformability. However, long-term results and outcomes of Atlas-assisted coiling in the setting of acute subarachnoid hemorrhage remain to be shown.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.