Clinical and angiographic outcomes of stent-assisted coiling of intracranial aneurysms

Abstract

Background

Despite the increasing use of stent-assisted coiling (SAC), data on its long-term clinical and angiographic results are limited.

Objective

The objective of this article is to assess the long-term clinical and angiographic outcomes in SAC in our single-center practice.

Methods

We conducted a retrospective analysis of intracranial aneurysms treated with detachable coils during the period 2003–2012. Patients were divided into SAC and non-SAC groups and were analyzed for aneurysm occlusion, major recurrence and clinical outcome. Logistic regression analyses identified factors associated with clinical/angiographic outcomes (p value <0.05 was statistically significant).

Results

A total of 516 procedures met inclusion criteria: Sixty-three (12.2%) patients underwent SAC, of whom 56 (89%) had an elective procedure whereas 286 (63.1%) aneurysms from the non-SAC group were ruptured. In the unruptured subcohort, baseline class I was achieved in 24 (38%, p = 0.91), and predischarge modified Rankin scale score (mRS) 0–2 was obtained in 96.4% of cases in the SAC group versus 90.4% in the non-stent group. The major recurrence was 9.5% versus 11.3% in the SAC and non-SAC group, respectively (p = 0.003). At last clinical assessment, 98.2% of the patients from the unruptured SAC group had mRS 0–2 (mean follow-up, 58 months) versus 93.6% (mean follow-up, 56 months) in the unruptured non-SAC group (p = 0.64). Periprocedural vasospasm was associated with long-term poor outcome in the unruptured SAC subcohort (p = 0.0008).

Conclusions

SAC and non-SCA techniques show comparable safety and clinical outcome. The SAC technique significantly decreases retreatment rates. Periprocedural vasospasm resulting from vessel manipulation is associated with poor outcome in SAC of unruptured aneurysms.

Introduction

Endovascular techniques are increasingly used to treat intracranial aneurysms.^1,2 Despite adjunctive techniques such as balloon remodeling, large, giant and wide-necked aneurysms remain challenging and with a significant number of incomplete occlusions. The introduction of micro-stents as an adjunct tool for coiling brought about an important paradigm shift in the treatment of intracranial aneurysms.

In 2010, a large case series reported more lethal complications in patients treated with the stent-assisted coiling (SAC) technique compared with non-SAC cases.³ A few years later, two systemic reviews concluded that SAC has more adverse events than conventional coiling.^4,5 However, a recent meta-analysis showed no significant difference in terms of complications and recurrence rate between SAC and traditional coiling techniques.⁶ In addition, this last systematic review identified that the majority of studies assessing the SAC technique have limitations such as short duration of follow-up and lack of baseline clinical data.⁶

This study will present detailed predischarge and long-term clinical and angiographic outcomes of intracranial aneurysms treated at our center by endovascular coiling using the SAC technique versus those using conventional and balloon-assisted coiling techniques. The periprocedural safety of the SAC technique as well as its efficacy with regards to recurrence, rerupture and retreatment rates were evaluated and compared to the historical non-SAC group.

Methods

All endovascular coilings of intracranial aneurysms performed between December 2003 and June 2012 were identified from our prospectively maintained database. Procedures of flow-remodeling stenting and parent vessel occlusion as well as patients harboring aneurysms arising on distal peripheral branches or associated with arteriovenous shunts were excluded. The digital subtraction (DSA) and magnetic resonance angiographic (MRA) interpretation was performed in a random fashion by four non-blinded fellowship-trained neuroradiologists following our pragmatic routine. This study was approved by the Ottawa Health Science Network Research Ethics Board.

The analysis of recurrence and clinical outcome of the entire cohort was performed according to procedure and the data were grouped into SAC versus non-SAC. We analyzed demographic, clinical and angiographic data at presentation and at follow-up. Clinical and demographic data included initial presentation, age, gender, hydrocephalus, Hunt-Hess, World Federation of Neurosurgeons (WFNS) and Fisher grades. Aneurysm data included location, shape (unilobular or multilobular), rupture status, dimensions, the total number of aneurysms present and the presence of incorporating arteries. We classified location into anterior (anterior communicating artery (AcomA), posterior communicating artery (PcomA), anterior cerebral artery (ACA), internal carotid artery (ICA), middle cerebral artery (MCA)) and posterior (basilar tip (BTip), non-basilar tip (nBTip)) circulations. The dome size was stratified into <13 mm and ≥ 13 mm, and the neck size was stratified into ≤ 4 mm and > 4 mm. Using DSA, we recorded periprocedural vasospasm, intraprocedural aneurysm perforation and thromboembolic events.

A DSA or MRA was performed immediately post-intervention and two to four months after the procedure followed by studies performed yearly thereafter. The degree of aneurysm occlusion after coiling was classified according to the Raymond-Roy classification: class I: complete/adequate occlusion; class II: residual neck/near complete/adequate occlusion; class III: residual body/incomplete/inadequate occlusion.⁷

Periprocedural vasospasm in the unruptured subcohort was any intracranial vasospasm related to vessel manipulation noted during the coiling procedure. Recurrence was defined as an interval change in class (I→II/III, or II→III), considering the immediate/predischarge angiographic study as baseline reference. Major recurrence was any change from class I/II to class III, an increase in class III remnant according to the consensus among the neuro-interventionists, cases of post-coiling rupture or rerupture, and cases receiving any type of retreatment after coiling. All cases of major recurrence were confirmed in DSA.

The primary stents used in this series included the Enterprise (Codman), Solitaire (ev3) and Neuroform (Stryker). Patients were started on dual antiplatelet medication (typically with aspirin and clopidogrel) before all elective SAC procedures. A loading dose of dual antiplatelet medication was given in cases when the stent was placed unexpectedly. Patients were continued on dual antiplatelet therapy for at least three months following the SAC procedure.

The clinical outcome was recorded using the modified Rankin scale score (mRS) at discharge (periprocedural) and at the last clinical encounter recorded in the patient’s chart. We considered lost to follow-up all patients or aneurysms that did not receive an angiographic follow-up after the hospital discharge.

Statistical analysis included nonparametric univariate and multivariable logistic regression analyses on clinical and radiological variables. Descriptive results were summarized as mean, standard deviation (SD), median, interquartiles, and ranges for continuous variables, and proportions for categorical variables. A p value of <0.05 was considered statistically significant. All analyses were conducted using Statistical Analysis Software (SAS version 9.3 for Windows). For univariate and multivariable logistic regression analyses, LOGISTIC procedure was applied in SAS. Occlusion rates and clinical outcome were compared via Fisher exact or Χ² test. Natural log transformation was applied for normalizing distribution of duration of follow-up. For the probability model, R² (proportion of variability) was also applied for measure of fit in the modeling.

Results

Baseline demographics

A total of 516 consecutive coiling procedures (441 patients, 476 aneurysms) were performed at The Ottawa Hospital during the period 2003–2012. Of these, 293 (56.8%) procedures were performed in the setting of acute subarachnoid hemorrhage. A total of 63 (12.2%) procedures (62 aneurysms) used the SAC technique (Figure 1).

Figure 1.

Flowchart of procedures, patients and aneurysms.

SAC: stent-assisted coiling.

The demographic and radiological characteristics are depicted in Tables 1 and 2. The SAC group showed a significantly lower frequency of ruptured aneurysms (63.1% versus 11.1%, odds ratio (OR) = 13.7; p < 0.0001) (Figure 2). The majority of the SAC cases were located in the posterior circulation (61.9% versus 41.3%, OR = 2.3; p = 0.0024), had a large neck (58.7% versus 18.1%, OR = 6.44; p < 0.0001) and a dome/neck (D/N) ratio <2 (71.4% versus 41.5%, OR = 3.25; p < 0.001) (Figure 3). Aneurysms associated with unruptured status (OR = 14.2; p < 0.0001)), D/N ratio <2 (OR = 3.0; p = 0.0006), presence of vessel incorporation (OR = 3.2; p = 0.0013) and in posterior circulation (OR = 2.0; p = 0.021) were more likely to be selected for SAC.

Table 1.

Demographic and radiological characteristics.

	Non-SAC	SAC	p valuea
	n = 453 (%)	n = 63 (%)
Age (year)			0.06
Mean ± SD	55.1 ± 11.7	58.1 ± 12.3
≥ 55 years	226 (49.9)	38 (60.3)	0.12
Duration from coiling to the latest follow-up (months)
Mean ± SD	44.9 ± 35.8	53.3 ± 35.8	0.03
≥ 30 months	242 (53.4)	38 (60.3)	0.30
Gender			0.45
Male	123 (27.1)	20 (31.7)
Female	330 (72.8)	43 (68.2)
Hunt-Hess grade			<0.0001
Grade 0	167 (36.8)	56 (88. 9)
Grade 1–3	213 (47)	5 (7.9)
Grade 4–5	73 (16.1)	2 (3.1)
Aneurysm location			0.0024
Anterior	266 (58.7)	24 (38.1)
Posterior	187 (41.2)	39 (61.9)
Rupture status			<0.0001
Ruptured	286 (63.1)	7 (11.1)
Unruptured	167 (36.8)	56 (88.9)

^ap values obtained by univariate logistic regression model of SAC versus non-SAC. SAC: stent-assisted coiling.

Table 2.

Aneurysms characteristics.

	Non-SAC	SAC	p valuea
	n = 453(%)	n = 63(%)
Multiple (>2) aneurysms	187 (41.2)	22 (34.9)	0.34
Unilobular shape	282 (62.2)	43 (68.2)	0.36
Dome size			0.24
< 13 mm	416 (91.8)	55 (87.3)
≥ 13 mm	37 (8.1)	8 (12.7)
Neck size			<0.0001
≤ 4 mm	371 (81.9)	26 (41.2)
> 4 mm	82 (18.1)	37 (58.7)
Dome/Neck ratio			<0.0001
≥ 2	256 (56.5)	18 (28.5)
< 2	197 (43.4)	45 (71.4)
Vasospasm	144 (31.7)	8 (12.7)	0.003
Vessel incorporation	63 (13.9)	20 (31.7)	0.0005

^ap values obtained by univariate logistic regression model of SAC versus non-SAC. SAC: stent-assisted coiling.

Figure 2.

Rupture status in the entire cohort.

SAC: stent-assisted coiling.

Figure 3.

Aneurysm location.

SAC: stent-assisted coiling.

Table 3 summarizes the indications for SAC in our patient population. As depicted, 17 (27%) cases had a recurrence/failure post-conventional coiling (15 cases) and clipping (two cases). One patient treated initially using SAC required a repeated SAC procedure. For two other cases, SAC was used after an unsuccessful conventional coiling attempt. From the 15 patients previously treated with a conventional endovascular technique, four had an unruptured status at initial presentation and none of those required a subsequent retreatment. In 45 (84.9%) SAC procedures the stent delivery preceded the coil deployment.

Table 3.

Indications for treatment in the SAC group.

Indications for treatment SAC subcohort	Procedures n (%)
Mult. An.; morphology	9 (14.3)
Acute rupture	6 (9.5)
Recurrence post-coiling	12 (19.1)
Recurrence post-clipping	1 (1.6)
Morphology	31 (49.2)a
Multiple/Other factors: Rec. post-coiling; Mult. An Failed clipping; morphology Failed coiling; morphology	4 (6.4) 1 1 2

^aMult. An.: multiple aneurysms; Rec: recurrence; SAC: stent-assisted coiling. ^aA total of 67.8% had a dome to neck (D/N) ratio ≤ 2

Follow-up duration

A total of 36 (7.9%) procedures in the non-SAC subcohort were lost to follow-up. There were two cases lost to follow-up in the SAC subcohort.

The mean clinical follow-up duration after treatment was similar for the two subcohorts (53 versus 45 ± 35.8 months for SAC and non-SAC cases, respectively). The mean clinical follow-up duration after coiling was similar for the ruptured and unruptured groups (45.4 ± 32.1 versus 55.7 ± 34.2 months, respectively) (Table 4).

Table 4.

Periprocedural vasospasm occurrence in the unruptured subgroup.

Vasospasm in unruptured aneurysms	SAC n = 56(%)	Non-SAC n = 167(%)
Periprocedural angiographic vasospasm present	4 (7.1)	15 (8.9)
Symptomatic after coiling	2 (3.5)	3 (1.8)
Treatment required during coiling procedure	2 (3.5)	6 (3.5)

SAC: stent-assisted coiling.

The distribution of the time interval elapsed between each procedure and its subsequent angiographic follow-up is shown in Figure 4.

Figure 4.

Distribution of the time interval elapsed between the treatment (SAC and non-SAC) and the last angiographic follow-up, for each procedure.

SAC: stent-assisted coiling.

Immediate angiographic occlusion, risk factors for recurrence, rates of recurrence, retreatment and rerupture

In the SAC group, 81% of cases achieved adequate baseline occlusion (class I or II) compared to 82% in the non-SAC coiling group (OR = 0.978, p = 0.9535) (Figure 5).

Figure 5.

Predischarge/baseline angiographic classification of aneurysm occlusion.

SAC: stent-assisted coiling.

Any recurrence, either minor or major, was seen in 147 (30.9% of 476 aneurysms) aneurysms of the entire cohort after a median (interquartile range (IQR)) of 33 (seven to 61) months. Major recurrence was seen in 98 (20.6%) aneurysms after six (3.5–22.5) months. Multivariable analysis of the entire cohort revealed that patients with age >55 years (OR = 2.07; p = 0.0144), male gender (OR = 2.58; p = 0.0017), posterior circulation aneurysm location (OR = 2.05; p = 0.0143), aneurysm dome size ≥13 mm (OR = 4.24; p = 0.0007), neck size >4 mm (OR = 2.39; p = 0.0069), and non-SAC procedure (OR = 2.12; p = 0.02) were more likely to have a major recurrence. There was no significant interaction among these factors and no other significant covariates were found in the multivariate analysis.

In the SAC group, major recurrence occurred in six cases (9.5% of the SAC subcohort); four unruptured and two ruptured aneurysms (three of these cases had a baseline class III remnant after conventional coiling, requiring SAC because of worsening in the class III remnant). The yearly recurrence rate was 0.28% and 0.43% in the SAC and non-SAC subcohorts, respectively.

Retreatment was required in 9.5% of the SAC versus 11.3% of the non-SAC cases (p = 0.003). The non-SAC group was 2.19 times more likely to have recurrence, retreatment or rerupture after coiling comparing to the SAC group (95% confidence interval (CI) = 1.24–3.85, log-rank test p = 0.0052) (Figure 6).

Figure 6.

The cumulative hazard of recurrence, retreatment or rerupture^a between non-SAC (red) versus the SAC group (blue) is 2.19 with 95% CI = 1.24–3.85, log-rank test p = 0.0052, censored up to 108 months (nine years).

^aThere was no rupture demonstrated among the unruptured cohort. SAC: stent-assisted coiling; CI: confidence interval.

There was no rupture demonstrated among the unruptured cohort. Three patients from the non-SAC group, all initially ruptured, presented with a rerupture, at three, 38 and 81 months after the initial treatment; the case of rerupture at 81 months had a minor recurrence (class I evolving to class II) detected at 70 months. The other two cases of rerupture were major recurrences.

Periprocedural complications and clinical outcomes

Periprocedural vasospasm was present in 15 (8.98%) non-SAC and four (7.14%) SAC procedures (Table 4, SAC subcohort).

In the unruptured cohort, a predischarge mRS of 0–2 was noted in 96.4% of cases in the SAC group versus 90.4% in the non-SAC group (p = 0.48). At last follow-up, 98.2% of the patients from the SAC group had an mRS of 0–2 (mean follow-up 58 months) versus 93.6% (mean follow-up of 56 months) in the non-SAC group (p = 0.64) (Table 5). By excluding the ruptured status cases, the presence of periprocedural vasospasm was the single factor statistically significant for long-term mRS ≥3 in the SAC group (p = 0.0008; OR = 0.147).

Table 5.

Clinical outcome (mRS status at discharge and at last clinical assessment).

mRS (mean in months)	Non-SAC	SAC	p valuea
Ruptured subcohort	n = 286(%)	n = 7(%)
mRS at discharge			0.26
0–2	185(64.6)	3(42.8)
3–5	62(21.6)	2(28.5)
6	39(13.6)	2(28.5)b
mRS at the last assessment (45.4)			0.25
0–2	197 (75.1)	3 (60)
3–5	23 (8.7)	0 (0)
6	42 (16)	2 (40)b
Unruptured subcohort	`n = 167 (%)	n = 56 (%)
mRS at discharge			0.48
0–2	151 (90)	54 (96)
3–5	14 (8.3)	2 (3.5)c
6	2 (1.2)	0 (0)
mRS at last assessment (55.7)			0.64
0–2	146 (93.5)	53 (98.1)
3–5	6 (3.8)	1 (1.8)2
6	4 (2.5)	0 (0)

^ap values obtained by Fisher exact test.

^bA 45-year-old female and a 73-year-old female with acute SAH (high WFNS and Fisher grades, hydrocephalus, EVD placement).

^cA 50-year-old male with multiple aneurysms who underwent SAC for a recurrent unruptured basilar tip aneurysm. His mRS 3 was noted after a previous non-SAC coiling of a ruptured PComA aneurysm complicated by symptomatic vasospasm.

SAC: stent-assisted coiling; mRS: modified Rankin scale; PComA; posterior communicating artery; SAH: subarachnoid hemorrhage; WFNS: World Federation of Neurosurgeons; EVD: external ventricular drainage.

In the SAC subcohort, periprocedural complication was 12.7% (eight cases) (Figure 7), represented by three patients from the unruptured SAC group who sustained a thromboembolic complication, four patients with aneurysm perforation during SAC, and one patient from the ruptured SAC group with aneurysm perforation during coiling. These eight patients had an mRS 0–2 at discharge and at long-term clinical follow-up. The mortality rate in the SAC group was represented by two cases of the ruptured SAC subcohort with high WFNS grade at admission (Figure 7).

Figure 7.

Periprocedural complications in SAC and non-SAC subcohorts according to the rupture status.

SAC: stent-assisted coiling;

The total of periprocedural neurological morbidity in the SAC subcohort was slightly higher than the non-SAC (Table 6); however, the mRS at discharge was comparable in both groups (Table 5). There were no periprocedural and long-term mortalities in the unruptured SAC subcohort (Figure 8).

Table 6.

Procedure-related complications of SAC.

Age/Sex	Location	Max. dome size (mm)	D/N ratio	Rupture status (H&H grade)	Pre-coiling mRS	Complications	Presence of VS	Discharge mRS	Total follow-up (mo)	mRS at last follow-up
64/M	SCA	8.9	2.8	R (2)	0	Perforation	Y	1	13	0
51/F	PComA	2.8	1.3	Unr	0	Perforation	N	1	67	1
47/M	BTip	6.3	0.9	Unr	0	Perforation	N	1	51	1
65/F	PComA	7.0	1.8	Unr	0	Perforation	N	1	53	0
56/M	BTip	6.3	1.0	Unr	0	Perforation; dissection access vessel	N	2	33	1
53/F	BTip	7.0	1.7	Unr	0	T/E	N	0	50	1
26/F	CavICA	4.2	0.6	Unr	0	T/E	N	1	100	0
56/F	OphICA	6.7	1.4	Unr	0	T/E	N	1	28	1

OphICA: paraophthalmic internal carotid artery; VS: vasospasm; R: ruptured; Unr: unruptured; mRS: modified Rankin scale score; T/E: thromboembolic; D/N: dome to neck; H&H: Hunt-Hess grade; M: male; F: female; SCA: superior cerebellar artery; PComA: posterior communicating artery; BTip: basilar tip; CavICA: cavernous internal carotid artery; N: no; Y: yes.

Figure 8.

Unruptured aneurysms within SAC and non-SAC subcohorts: stacked plots of mRS at discharge and at last clinic encounter (corresponding mean follow-up in months in the brackets).

SAC: stent-assisted coiling; mRS: modified Rankin scale score.

For the entire cohort, the independent factors associated with favorable long-term clinical outcome in the multivariate analysis were age <55 years (OR = 0.57; p = 0.0219), absence of any vasospasm (OR = 0.55; p = 0.0197), duration of follow-up (OR = 4.19; p < 0.0001) and WNFS grade ≤3 (OR = 10.43; p < 0.0001) at presentation.

Discussion

Endovascular coil embolization of intracranial aneurysms has been increasingly used.^1,2 The SAC technique has broadened the indications for endovascular aneurysm repair^4,6 specifically when the lesions are not amenable to traditional coiling techniques owing to their unfavorable morphology. Although the results of previous observational studies have outlined the safety involved in the use of stents, a cautious interpretation of the literature published to date is required. The majority of the studies had insufficient clinical baseline and selection criteria information and lacked comparison with other treatment groups.⁶

Our baseline degree of occlusion was similar in the SAC and non-SAC subcohorts. When comparing our baseline degree of occlusion by rupture status, we found no difference among SAC and non-SAC subcohorts. Conversely, previous SAC pooled data^3,6 leaned toward a lower immediate occlusion rate in the SAC group and a significantly higher immediate occlusion rate in ruptured aneurysms.⁸

Continuous stent-induced thrombosis facilitated by a reduction in pulsation and in-flow is a possible explanation for the lower recurrence of aneurysms treated with SAC technique.^8,9 In agreement with the findings of a recent systematic review,⁶ our study showed that SAC embolization reduced retreatment and recurrence rates compared with non-SAC controls. Our research uses a considerably longer than average clinical/angiographic follow-up duration, shedding light on a topic still to be assessed in a long-term study.^6,10

Our mRS distribution at discharge, of SAC and non-SAC, was comparable to other series.^3,11–13 The sub-analysis of our unruptured group found a slightly higher rate of periprocedural complications in the SAC mainly due to intraprocedural aneurysm perforations. However, this did not negatively impact the predischarge and long-term mRS. In addition, unlike previous reports, our ruptured aneurysms in the SAC and non-SAC subcohorts showed comparable periprocedural complication rates.⁴ Periprocedural intracranial vasospasm remained a significant predictor of poor long-term outcome when ruptured status was eliminated from the multivariate analysis. This factor, not previously reported in the literature, should raise awareness of the possible consequences of difficult access to an unruptured aneurysm. It also emphasizes the need to avoid excessive vessel manipulation.

Our few delayed ruptures happened in the non-SAC group in previously ruptured aneurysms for which an initial incomplete occlusion was achieved. The rate of delayed hemorrhage after coiling is underreported,^5,6,14 likely because of the limited number of studies with long-term follow-up data. However, it tends to be higher in studies with a larger proportion of originally ruptured aneurysms.^3,15

Approximately one-fourth of our SAC patients were treated following recurrence or complication occurring after an attempt to treat the aneurysm with a non-SAC technique. Despite the inclusion of those aneurysms we saw low mortality and morbidity rates in the SAC group. This likely emphasizes the safety of the SAC technique.

Encouragingly, no permanent complications occurred in our ruptured SAC subgroup. The morbidity rate in this subgroup was secondary to the original hemorrhage. However, the literature on this topic is contradictory. Some studies showed that SAC represents a feasible option for difficult ruptured wide-neck aneurysms.^12,13 Conversely, a systematic review depicted a higher rate of adverse events and worse clinical outcomes when SAC was used in acutely ruptured aneurysms.⁴ This study has clear drawbacks. The retrospective design, the small number of patients, the absence of independent clinical reviewers and the lack of standardized follow-up formats are factors that might have interfered with the timing or ability to accurately detect a major recurrence. Given that we report our single-center experience, extrapolation of our results should be performed carefully. However, given the inconclusive treatment guidelines for SAC we believe that the decision-making process could take into consideration the local results of the specific treatment strategy proposed.

The strengths of our study are the relative homogenous population in which a similar decision-making process was applied. The initial and follow-up determination of aneurysm occlusion grade was classified according to established three-grade criteria in order to promote inter-observer reliability. Although both MRA and DSA were used to assess the degree of obliteration, MRA proved to be well suited for detecting or excluding aneurysm recurrences that may require retreatment. No systematic review assessed the MRA use after SAC.^16–18 We used DSA in all cases where uncertain occlusion results were detected in the MRA. Furthermore, our mean clinical and radiological follow-up duration is considerably longer than that of the previous studies.^3–5,19,20

Conclusions

The SAC technique significantly decreased retreatment rates compared with non-SAC techniques without negatively affecting clinical outcome. A significant proportion of the SAC procedures happened after a failed initial conventional coiling/clipping technique. The absence of periprocedural intracranial vasospasm related to vessel manipulation remained a significant variable in the unruptured SAC subcohort. Well-designed prospective, randomized studies evaluating the clinical and radiological outcomes of SAC are warranted for further validation.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest

None declared.