Anatomic Results and Complications of Stent-Assisted Coil Embolization of Intracranial Aneurysms

Summary

The purpose of this study was to evaluate and report our anatomic results and complications associated with stent-assisted coil embolization of intracranial aneurysms using the Neuroform stent. From September 2003 to August 2007, 127 consecutive patients (ruptured 50, 39.4%; unruptured 77, 60.6%) underwent 129 stent-assisted coil embolization procedures to treat 136 aneurysms at our institution. Anatomic results at follow-up, procedure-related complications, and morbidity/mortality were retrospectively reviewed.

Stent deployment was successful in 128 out of 129 procedures (99.2%). Forty-seven patients presented with 53 procedure-related complications (37.0%, 47/127). Thromboembolic events (n=17, 13.4%) were the most common complications, followed by intraoperative rupture (n=8, 6.3%), coil herniation (n=5, 3.9%), and postoperative rupture (n=4, 3.1%). For thromboembolic events, acute intra-procedural in-stent thromboses were observed in two patients and subacute or delayed in-stent thromboses in three patients. Overall mortality rate was 16.5% (21/127) and procedure-related morbidity and mortality rates were 5.5% (7/127) and 8.7% (11/127) retrospectively. Patients with poor grade subarachnoid hemorrhage (Hunt and Hess grade IV or V; 25/127, 19.7%) exhibited 56% (14/25) overall mortality rate and 24% (6/25) procedure-related mortality rate. Immediate angiographic results showed complete occlusion in 31.7% of aneurysms, near-complete occlusion in 45.5%, and partial occlusion in 22.8%. Sixty nine patients in 70 procedures with 77 aneurysms underwent angiographic follow-up at six months or later. Mean follow-up period was 13.7 months (6 to 45 months). Complete occlusion was observed in 57 aneurysms (74.0%) and significant in-stent stenosis was not found.

Thromboembolism and intra/postoperative aneurysm ruptures were the most common complications and the main causes of procedure-related morbidity and mortality. Patients with poor grade subarachnoid hemorrhage showed poor clinical outcomes. Since most complications were induced by stent manipulation and deployment, it is mandatory to utilize these devices selectively and cautiously. While the follow-up angiographic results are promising, further studies are essential to evaluate safety, efficacy, and durability of the Neuroform stent.

Introduction

Endovascular embolization of cerebral aneur­ysms using detachable coils, first introduced by Guglielmi¹, has become a safe and effective alternative to direct surgical clipping^2-9. Endovascular treatment of wide-necked and large/giant (>10 mm) aneurysms, however, remains challenging in spite of technical advancements^10-13. Published complete occlusion rates of wide-necked and large/giant aneurysms ranges from 25 to 56.7%^11,13-18. To improve treatment outcomes in these difficult cases, new devices and techniques such as three-dimensional coils, neck-bridging devices, multiple-microcatheter techniques, and balloon-assisted neck-remodeling techniques have been utilized^19-24.

In the past, many clinicians utilized coronary stents for the endovascular treatment of wide-necked intracranial aneurysms ^25-31. The device’s physical properties as well as the tortuous intracranial vasculature limited the use of coronary stents. The Neuroform stent (Boston Scientific/Target, Fremont, CA) answered the need for an aneurysmal neck-bridging device specifically designed for intracranial use. This self-expanding stent has a high degree of flexibility, trackability and conformability which facilitates movement through the intracranial vasculature. The purpose of this study is to evaluate and report our results and complications associated with stent-assisted coil embolization of intracranial aneurysms using the Neuroform stent.

Methods

From September 2003 to August 2007, a total of 129 stent-assisted coil embolization procedures were performed to treat 136 aneurysms in 127 consecutive patients at the University of Pittsburgh Medical Center’s Presbyterian Hospital. We retrospectively reviewed the medical records, radiographic studies and endovascular procedure reports for each case. Institutional Review Board (IRB) approval and informed consent were obtained from all the patients prior to aneurysm treatment.

Thirty one men (24.4%) and 97 women (75.6%) were included in this study. Mean age was 55.6 years with a range of 21 to 84. Nineteen patients (15.0%) were over 70 years old. Fifty patients (39.4%, 50/127) presented with subarachnoid hemorrhage (SAH) from a ruptured aneurysm, resulting in 52 procedures (40.3%, 52/129) for the treatment of 54 aneur­ysms (39.7%, 54/136). Forty-seven of these patients underwent one stent deployment procedure for a single aneurysm. In the setting of SAH with multiple aneurysms, identification of the ruptured aneurysm was based on clinical, CT and angiographic findings^32-34. If multiple aneurysms were identified of similar size and shape at the same site in the face of SAH, all were regarded as ruptured and treated at the same time. Two patients, one with a middle cerebral artery (MCA) bifurcation aneurysm and a basilar apex (BA) aneurysm and the other with a posterior communicating (PCom) artery aneurysm and a BA aneurysm, underwent stent-assisted coil embolization for each aneurysm in a single session. Another patient with three anterior communicating (ACom) artery aneurysms underwent a single procedure for multiple aneurysms.

In 77 patients (60.6%, 77/127) in 77 procedu­res (59.7%, 77/129) with 82 unruptured aneur­ysms (60.3%, 82/136), 72 patients underwent one stent deployment procedure for a single aneurysm. Five patients (one patient with a BA aneurysm and a left superior cerebellar artery (SCA) aneurysm, three patients with two paraclinoid aneurysms, and one patient with a PCom artery aneurysm and a paraclinoid aneurysm) underwent one stent deployment across multiple aneurysms.

Aneurysm locations are reported in Table 1. Ninety-three aneurysms (68.4%, 93/136) were located in the anterior and 43 (31.2%, 43/136) in the posterior circulation. Fifty-four aneur­ysms (39.7%, 54/136) were ruptured and 82 (60.3%, 82/136) were unruptured. The most common location for ruptured aneurysms was the BA (n=12, 22.2%), followed by the PCom artery (n=9, 16.7%) and the MCA bifurcation (n=7, 13.0%). For unruptured aneurysms, the most common location was the paraclinoid internal carotid artery (n=32, 39.0%), followed by the PCom artery (n=12, 14.6%) and the BA (n=11, 13.4%). Overall, the most common location was the paraclinoid internal carotid artery (n=35, 25.7%), followed by the BA (n=23, 16.9%) and the PCom artery (n=21, 15.4%). One hundred and fourteen (83.8%, 114/136) of the aneurysms were treated de novo whereas 22 (16.2%, 22/136) represented previously-treated residual/recanalized lesions. Seventeen residual/recanalized aneurysms had been previously treated using coiling, four had undergone direct surgical clipping, and one had both. In the 22 residual/recanalized aneurysms, seven (31.8%) presented with SAH in this study (at the time of stent-assisted coil embolization) whereas 15 represented unruptured lesions. The most common location of the residual/recanalized aneurysm was the basilar apex (27.3%, 6/22), followed by the MCA bifurcation (18.2%, 4/22) and the paraclinoid internal carotid artery (ICA) (18.2%, 4/22). The location harboring the most residual/recanalization by volume was the MCA bifurcation (4/11, 36.4%), followed by the BA (6/23, 26.1%; Table 1).

Table 1.

Location of aneurysms.

Location	Number	Number	Number
	of total aneurysms	of ruptured aneurysms	of unruptured aneurysms
	(Number of previously	(Number of previously	(Number of previously
	treated aneurysms)	treated aneurysms)	treated aneurysms)
	n=136 (22)	n=54 (7)	n=82 (15)
ACA/distal	3 (1)	2 (1)	1
ACA/ACom	10 (1)	6	4 (1)
ACA/A1	1	0	1
MCA/bifurcation	11 (4)	7 (2)	4 (2)
MCA/M1	1	0	1
ICA/bifurcation	7 (2)	2 (1)	5 (1)
ICA/anterior choroidal	2	2	0
ICA/PCom	21 (3)	9	12 (3)
ICA/paraclinoid	35 (4)	3	32 (4)
ICA/cavernous	2	0	2
BA/apex	23 (6)	12* (3)	11 (3)
BA/SCA	6 (1)	1	5 (1)
BA/trunk	3	1	2
VB	3	3	0
V4	1	1	0
V4/PICA	7	5	2
ACA, anterior cerebral artery; ACom, anterior communicating artery; a1, A1 segment of ACA; MCA, middle cerebral artery;
M1, M1 segment of MCA; ICA, internal carotid artery; PCom, posterior communicating artery; BA, basilar artery;
SCA, superior cerebellar artery; VBJ, vertebrobasilar junction; V4, intracranial vertebral artery; PICA, posterior cerebellar artery
*; two unruptured de novo aneurysms in the face of subarachnoid hemorrhage were included.

Aneurysm geometry was evaluated by digital subtraction angiography and, if necessary, a three-dimensional reconstruction. All aneur­ysms were wide-necked, defined as a neck diameter >4 mm or aspect ratio <2. Aneurysms were measured according to their greatest diameter. Ninety-nine aneurysms (72.8%; ruptured 38, unruptured 61) were small (<10mm), 32 aneurysms (23.5%; ruptured 14, unruptured 18) were large (>10 mm, <25 mm), and five aneurysms (3.7%; ruptured 2, unruptured 3) were giant (>25 mm).

In 50 patients with SAH, initial clinical status was assessed using the Hunt and Hess grade system.35 Six patients were grade I (12%); eleven grade II (22%); eight grade III (16%); sixteen grade IV (32%); and nine grade V (18%) (Table 2). In all patients, clinical outcomes were measured using the modified Rankin scale (mRS) which was obtained on admission day, three days post-procedure, and again at 90 days (Tables 2 and 3).

Table 2.

Clinical outcomes in patients with ruptured aneurysms (n=50).

H-H		mRS 3 day							mRS 90 day
scale	n	0	1	2	3	4	5	6	0	1	2	3	4	5	6
I	6	2	2	1				1	2	1	2				1
II	11	7		1	1		1	1	8		1	1			1
III	8	1	1	1	2	1	2		2	1	2	1			2
IV	16		1	1			10	4	1	1		3	2		9
V	9						6	3			1	2	1		5
H-H, Hunt and Hess grade; mRS, modified Rankin Scale

Table 3.

Clinical outcomes in patients with unruptured aneurysms (n=77).

mRS 0 day		mRS 3 day							mRS 90 day
score	n	0	1	2	3	4	5	6	0	1	2	3	4	5	6
0	53	49	1	1	2				48	1	1	2			1
1	8		7					1		7					1
2	11			8	2	1				1	9	1
3	5				3	2						4			1
4
5
6
mRS, modified Rankin Scale

In patients with unruptured aneurysms, a dual antiplatelet regimen consisting of 325 mg of aspirin and 75 mg of clopidogrel (Plavix; Bristol-Meyers Squibb/Sanofi Pharmaceuticals, New York, NY) was administered orally for two days prior to the procedure. All procedures were performed under full heparinization with an activated clotting time of approximately 250 seconds. In patients with SAH, heparin was given after sheath placement to maintain an activated clotting time of 200-250 seconds. Fifteen milligrams (mg) of eptifibatide (Integrilin; Schering, Kenilworth, NJ) were administered intravenously once the stent was deployed and a loading dose of clopidogrel (600 mg) was given at the procedure’s conclusion via a naso- or orogastric tube. Heparin (500 units per hour IV) was continued for the next 12-15 hours in most cases. Patients continued taking 75 mg of clopidogrel daily for one month after the stenting procedure and 325 mg of aspirin for life.

All procedures were performed with patients under general anesthesia. Neurophysiologic monitoring was used for all cases, which included somatosensory evoked potentials, electroencephalography, and brain stem auditory evoked potentials. The right femoral approach was routinely used for placement of a 6F Envoy (Cordis, Miami, FL) or 7F Shuttle (Cook, Bloomington, IN) catheter. Microcatheters (Rapid Transit; Cordis, Miami, FL; Echelon; ev3, Irvine, CA) and Nexus/NXT (ev3, Irvine, CA) and/or Micrus (Micrus Co., Mountain View, CA) coils were primarily used. Cerebral angiography was performed to evaluate aneur­ysm morphology and to establish working projections. Typically, the undeployed stent was positioned across the aneurysms neck followed by placement of the microcatheter into the aneur­ysm fundus. The first coil was partially deployed into the aneurysm without detachment so as to anchor the microcatheter in place. Once the microcatheter was jailed by deploying the stent, the first coil was fully advanced and detached. The Neuroform stent was usually advanced over an exchange wire (Transcend 14/300 cm; Boston Scientific/Target, Fremont, CA; Accelerator 14/300cm, ev3, Irvine, CA) and deployed using either the provided pusher system or a coil pusher. All coils were introduced into the aneurysm by using simultaneous biplanar imaging. Coil placement proceeded until no additional coils could be placed, until angiographically complete occlusion, or until risks outweighed perceived benefits. Once the stent was deployed or shortly thereafter, 15 mg of eptifibatide was given intravenously unless contraindicated (i.e., intraoperative rupture).

In a total of 127 patients in 129 procedures with 136 aneurysms, 114 patients in 116 procedures with 123 aneurysms were treated with simultaneous stent deployment and aneurysmal coil embolization in the first session. One patient, complicated by uncontrolled intraoperative rupture, underwent Neuroform stent deployment and embolization with Onyx-18 (ev3, Irvine, CA). In one patient with complications related to the failure of Neuroform stent deployment, a Wingspan stent (Boston Scientific/Target, Fremont, CA) was deployed to pin herniated coil loops. In 11 patients in 11 procedures with 11 aneurysms, the stent was deployed in the first session and coil embolization was performed in the next session (Table 4). Seventeen patients were treated in a multi-staged fashion (Table 4). Fourteen patients underwent two sessions, two patients three sessions, and one patient four sessions.

Table 4.

Summary of cases treated in multiple stages.

	SAH	Size		Previous			Interval				Interval to
Age/Sex	(H-H,		Location	Treatment/	1st	Immediate	to Next	Next	Immediate	Remark	Follow/up	Follow/up
	Fisher)			Interval	Session	Result	Sessions	Sessions	Result		(months)	Result
				(months)			(months)
49 F	0	L	BA	0	STENT	STENT	1	2nd	PO		6	CO
			TRUNK
										Fusiform,
									2nd: STENT	19mm
40 M	0	L	PCOM	0	STENT	STENT	1, 3	2nd, 3rd	AGAIN	posttraumatic,	14	No residual
									3rd: Clipping	BTO: not
										tolerable
25 M	0	L	paraclinoid	0	STENT	PO	1	2nd	NC		9, 23	CO, CO
					+ COIL
38 F	0	S	paraclinoid	0	STENT	PO	1	2nd	NC		2, 6	CO, CO
					+ COIL
57 F	0	S	paraclinoid	0	STENT	PO	2	2nd	NC		2, 18	CO, CO
					+ COIL
43 F	0	S	ACOM	0	STENT	PO	2	2nd	NC
					+ COIL
48 F	0	S	paraclinoid	CLIPPING	STENT	PO	3	2nd	PO		5, 17	CO, CO
				149	+ COIL
30 M	0	S	MCAB	COILING	STENT	STENT	1	2nd	NC		6	CO
				14
53 F	0	L	paraclinoid	0	STENT	STENT	1	2nd	NC		7	CO
61 F	0	S	paraclinoid	COILING	STENT	STENT	1	2nd	NC		5, 11	CO
				6								Recanalized
									2nd: stent
									navigation	Deceased
65 F	0	G	ICAB	0	STENT	STENT	7 days,	2nd, 3rd	failure	POD1 in 3rd
							13 days		3rd: coiling,	session
									intraoperative
									rupture
62 F	0	S	paraclinoid	COILING	STENT	STENT	2	2nd	NC		3, 14, 26	CO, CO, CO
				11
56 F	0	L	paraclinoid	COILING	STENT	STENT	2	2nd	NC		6	Recanalized
				99, 92
55 M	Yes	G	ICAB	0	STENT	PO	4 days,	2nd, 3rd,	PO, PO, PO	Partially	6	Recanalized
	(V, 4)				+ COIL		5, 8	4th		thrombosed
										giant aneurysm
				CLIPPING
46 M	Yes	S	MCAB	5, 1	STENT	PO	6 days	2nd	STENT	small residual	8	NC
	(I, 4)			COILING					AGAIN	sac: 2mm
				4 days (PO)
51 F	0	S	PCOM	COILING	STENT	STENT	2	2nd	no treatment	decrease of
				14						aneurysm size
44 F	0	S	BA APEX	COILING	STENT	STENT	3	2nd	no treatment	decrease of
				25						aneurysm size
SAH, subarachnoid hemorrhage; H-H, Hunt and Hess grade; Fisher, Fisher grade; S, small; L, large; G, giant;
ACOM, anterior communicating artery; BA APEX, basilar artery apex; BA TRUNK, basilar artery trunk;
ICAB, internal carotid artery bifurcation; MCAB, middle cerebral artery bifurcation; PCOM, posterior communicating artery;
CO, complete occlusion; NC, near-complete occlusion; PO, partial occlusion; BTO, balloon test occlusion; POD, postoperative day

The degree of aneurysm occlusion was classified into three subgroups: complete occlusion (CO) when the sac and the neck were densely packed, near-complete occlusion (NC) when the sac was occluded but a neck remnant was suspected, and partial occlusion (PO) when loose packing or opacification of the sac or the neck persisted (Figure 1). Angiographic follow-up at six months or later was obtained in 69 patients in 70 procedures with 77 aneurysms. Mean follow-up period was 13.7 months with a range from six to 45 months.

Figure 1.

Complete occlusion: Preoperative (A) and postoperative (B) angiogram. Near-complete occlusion: Preoperative (C) and postoperative (D) angiogram. Partial occlusion: Preoperative (E) and postoperative (F) angiogram.

A.

B.

C.

D.

E.

F.

All cases in which procedure-related complications occurred were noted. Procedure-related morbidity was defined as a complication leading to neurologic sequelae which persisted for at least 90 days after the procedure and/or a modified Rankin scale score that was higher at that time as compared to the initial score. Radiologic findings, clinical presentations, treatments, and clinical outcomes were analyzed.

Results

Immediate Angiographic Results

A Neuroform stent was successfully deployed in 126 patients in 128 procedures with 135 aneurysms out of 127 patients in 129 procedures with 136 aneurysms. In a total of 129 procedures, the technical success rate was 99.2% (128/129). Simultaneous stent deployment and aneurysmal coil embolization were conducted in 114 patients in 116 procedures on 123 aneurysms in the first session. In these patients, complete aneurysm occlusion was observed in 39 aneurysms (31.7%), near-complete occlusion in 56 (45.5%), and partial occlusion in 28 (22.8%).

Procedure-Related Complications

In a total of 127 patients, 47 patients exhibited 53 procedure-related complications for an overall complication rate of 37.0% (47/127). Among the 50 patients with ruptured aneur­ysms/SAH, 23 exhibited 26 complications for a complication rate of 46.0% (23/50). In the 77 patients with unruptured aneurysms, 24 exhibited 27 complications for a complication rate of 31.2% (24/77). Thromboembolism was the most common overall complication (13.4%, 17/127), followed by intraoperative rupture (6.3%, 8/127), coil herniation (3.9%, 5/127), postoperative rupture (3.1%, 4/127), catheter-induced vasospasm (3.1%), and stent migration (3.1%). In patients with ruptured aneur­ysms/SAH, thromboembolism was the most common complication (16.0%, 8/50), followed by intraoperative rupture (14.0%, 7/50), stent migration (6.0%, 3/50), and postoperative rupture (4.0%, 2/50). In patients with unruptured aneurysms, thromboembolism was also the most common (11.7%, 9/77), followed by coil herniation (5.2%, 4/77) and catheter-induced vasospasm (3.9%, 3/77) (Table 5).

Table 5.

Procedure-related complications.

Complication	Number of cases			Incidence Rate			Morbidity (n=7) / Mortality (n=11)
							Rates from Complications
	Total	Ruptured	Unruptured	Total	Ruptured	Unruptured	Total	Ruptured	Unruptured
	(n=53)	(n=26)	(n=27)	(n/127)	(n/50)	(n/77)	(n, n/127)	(n, n/50)	(n, n/77)
Thromboembolism	17	8	9	13.4%	16.0%	11.7%	5, 3.9%	2, 4%	3, 3.9%
							1, 0.8%	1, 2%
Intraoperative	8	7	1	6.3%	14.0%	1.3%	7, 5.5%	6, 12%	1, 1.3%
rupture
Coil herniation	5	1	4	3.9%	2.0%	5.2%	1*, 0.8%	1*, 2%
through stent
Postoperative	4	2	2	3.1%	4.0%	2.6%	1, 0.8%	2, 4%	1, 1.3%
rupture							3, 2.4%		1, 1.3%
Catheter-induced	4	1	3	3.1%	2.0%	3.9%
vasospasm
Stent migration	4	3	1	3.1%	6.0%	1.3%
Iatrogenic dissection	3	1	2	2.4%	2.0%	2.6%
Coil fracture	2	1	1	1.6%	2.0%	1.3%
Inappropriate stent	2	0	2	1.6%	0	2.6%
deployment
Failure in	1	0	1	0.8%	0	1.3%
microcatheter
placement after
stenting
CCF	1	1	0	0.8%	2.0%	0
ICH	1	1	0	0.8%	2.0%	0	1, 0.8%	1, 2%
Cranial nerve palsy	1	0	1	0.8%	0	1.3%
CCF, carotid-cavernous fistula; ICH, intracerebral hemorrhage; *, one case of coil herniation led to thromboembolic event.

Clinical Outcome - Mortality

Clinical outcomes are summarized in Tables 2 and 3. In a total of 127 patients, overall mortality rate was 16.5% (21/127) and procedure-related mortality was 8.7% (11/127). In the 11 procedure-related mortalities, seven (63.6%) were due to intraoperative rupture, three (27.3%) were due to postoperative rupture, and one (9.1%) was due to thromboembolism. Table 6 summarizes the mortality cases.

Table 6.

Summary of mortality cases.

Age/	SAH	mRS	mRS	Systemic	Size	Location	Previous	Immediate	Major	Interval	Proce-
Sex	(H-H,	0	3	condition			Treatment/	angiographic	Complications	to	dure
	Fisher)	day	day				Interval	result	& Cause of Death	Death	-related
60/F	Yes	0	6		S	paraclinoid		NC	Postoperative	POD3	Yes
	(I, 2)								Rupture POD1
56/M	Yes	1	6	Hypertension	L	BA APEX		PO	Postoperative	POD1	Yes
	(I, 2)			Coronary					Rupture POD0
				Arterial Disease
				Diabetes Mellitus
80/F	Yes	1	5	Atrial Fibrillation	L	PCOM	Clipping	PO	ARDS	POD10
	(III, 3)			Hypertension			179 months		GI Bleeding
55/M	Yes	2	1	Hypertension	S	BA APEX		NC	Distal BA	POD16	Yes
	(III, 4)								& bilateral PCA
									Occlusion POD8
71/F	Yes	4	5	Hypertension	L	ACOM		CO	Metabolic	POD3
	(IV, 4)								Encephalopathy
52/F	Yes	5	5	Rebled x 2	G	paraclinoid		NC	Intraoperative	POD5	Yes
	(IV, 3)								Rupture x 2
81/F	Yes	5	5	Hypertension	S	PCOM	Coiling, 65,	CO	IICP	POD9
	(IV, 4)						53 months		ARF
44/M	Yes	5	5	Hypertension	L	BA APEX		CO	IICP	POD6
	(IV, 4)			Morbid Obesity
41/F	Yes	5	5		S	PCOM		NC	IICP	POD11
	(IV, 4)								Vasospasm
55/F	Yes	5	6	Rebled	S	PCOM		PO	Intraoperative	POD1	Yes
	(IV, 3)								Rupture
75/M	Yes	5	6	Hypertension	S	PCOM		PO	Intraoperative	POD1	Yes
	(IV, 3)								Rupture
53/M	Yes	5	6	Rebled	S	PICA		CO	Intraoperative	POD2	Yes
	(IV, 3)								Rupture
37/M	Yes	5	6	Hypertension	S	V-B		CO	Intraoperative	POD1	Yes
	(IV, 4)								Rupture
71/F	Yes	5	5	Coronary	L	ACHOA		PO	IICP	POD10
	(V, 4)			Arterial Disease					Vasospasm
60/F	Yes	5	5	Hypertension	L	DACA		NC	IICP	POD7
	(V, 4)			Intracerebral
				Hemorrhage
65/M	Yes	5	6	Rebled	S	MCAB		NC	IICP	POD2
	(V, 4)			Intracerebral
				Hemorrhage
31/M	Yes	5	6	Rebled	S	PICA		CO	IICP	POD1
	(V, 4)
68/F	Yes	5	6		S	ACOM		NC	Intraoperative	POD1	Yes
	(V, 4)								Rupture
65/F	No	0	6	Hypertension	G	ICAB		NC	Intraoperative	POD1	Yes
				Coronary					Rupture
				Arterial Disease					(in the 3rd session)
				Diabetes Mellitus
54/F	No	1	6		S	BA-SCA		NC	Postoper. Rupture	POD1	Yes
									POD0
79/F	No	3	4	Atrial Fibrillation	L	BA APEX		NC	GI & Vaginal	2 months
				Hypertension					Bleeding
				brain stem
				compression
SAH, subarachnoid hemorrhage; H-H, Hunt and Hess grade; Fisher, Fisher grade; mRS, modified Rankin Scale; S, small; L, large;
G, giant; ACHOA, anterior choroidal artery; ACOM, anterior communicating artery; BA APEX, basilar artery apex;
BA-SCA, superior cerebellar artery; DACA, distal anterior cerebral artery; ICAB, internal carotid artery bifurcation;
MCAB, middle cerebral artery bifurcation; PCA, posterior cerebral artery; PCOM, posterior communicating artery;
PICA, posterior inferior cerebellar artery; V-B, vertebrobasilar junction; CO, complete occlusion; NC, near-complete occlusion;
PO, partial occlusion; POD, postoperative day; ARDS, adult respiratory distress syndrome; GI, gastrointestinal;
IICP, increased intracranial pressure; ARF, acute renal failure.

In patients with ruptured aneurysms/SAH, overall mortality rate was 36.0% (18/50) and procedure-related mortality rate was 18.0% (9/50). Mortality rate was higher in patients whose initial clinical statuses were worse. In patients whose initial clinical statuses were poor (Hunt and Hess grade IV or V), overall mortality rate was 56.0% (14/25) and procedure-related mortality rate was 24.0% (6/25). In patients whose initial clinical statuses were good (Hunt and Hess grade I, II, or III), overall mortality rate was 16.0% (4/25) and procedure-related mortality rate was 12.0% (3/25). In patients with unruptured aneurysms, mortality rate was 3.9% (3/77) and procedure-related mortality was 2.6% (2/77).

Clinical Outcome - Morbidity

Procedure-related neurologic morbidity was observed in seven patients (5.5%, 7/127; Table 7). Three cases involved ruptured aneurysms and four involved unruptured aneurysms. In patients with ruptured aneurysms/SAH, procedure-related neurologic morbidity rate was 6.0% (3/50) and, in patients with unruptured aneurysms, 5.2% (4/77). Five patients suffered a thromboembolic event, one a postoperative rupture, and one an intracerebral hemorrhage.

Table 7.

Cases of procedure-related neurologic morbidity.

Age/	SAH	mRS	mRS	mRS	Systemic	Size	Location	Previous	Immediate	Compli-	Interval
Sex	(H-H,	0	3	90	condition			Treatment/	angiographic	cations	to
	Fisher)	day	day	day				Interval	result		complication
53/F		0	0	3	hypertension,	L	BA		NC	left PCA	2 months
					left PCA		SCA			occlusion
					stenosis, renal
					artery stenosis
63/F		0	3	3	Diabetes	L	paracli-		NC	Air	intra-
					mellitus,		noid			embolism	operative
					hypertension
47/F	Yes	0	5	3		S	BA		CO	left
	(II, 2)						APEX			occipital	<24hr
										ICH
62/F		2	4	3	dilated	G	MCAB	coiling	PO	postopera-	<24hr
					cardiomyopathy,			3 months		tive
					coronary artery					rupture
					disease, dyslipidemia
66/F	Yes	0	0		hypertension,		S	BA-	NC	subacute	1 month
	(I, 2)			2	hypercholeste-		SCA			in-stent
					rolemia					thrombosis
65/F	Yes	1	3	2	hypertension	S	BA		NC	coil	intraope-
	(III, 4)						APEX			herniation	rative,
										left PCA	<24hr
										infarction
75/M		0	3	2	coronary artery	L	ACOM		NC	both ACA	1 day
					disease, diabetes					& left
					mellitus,					parietal
					hypertension					infarction
SAH, subarachnoid hemorrhage; H-H, Hunt and Hess grade; Fisher, Fisher grade; mRS, modified Rankin Scale; S, small;
L, large; ACOM, anterior communicating artery; BA APEX, basilar artery apex; BA-SCA, superior cerebellar artery;
MCAB, middle cerebral artery bifurcation; CO, complete occlusion; NC, near-complete occlusion; PO, partial occlusion;
PCA, posterior cerebral artery; ICH, intracerebral hemorrhage; ACA, anterior cerebral artery

Follow-up Angiographic Results

In a total of 106 survivors in 108 procedures with 115 aneurysms, 69 patients (65.1%) in 70 procedures (64.8%) with 77 aneurysms (67.0%) underwent angiographic follow-up at six months or later. Mean follow-up period was 13.7 months with a range from six to 45 months. Thirty seven patients did not undergo follow-up angiography. Seventeen patients could not be located and 20 patients refused follow-up angiography for reason of old age, other medical conditions, and financial issues. On follow-up angiography, one patient exhibited 30% in-stent stenosis at six months with complete resolution on follow up at 24 months without intervention (Figure 2).

Figure 2.

A 61-year-old woman presented with an unruptured left ophthalmic artery aneurysm (A). She underwent stent-assisted coil embolization. The aneurysm was partially occluded and in-stent thrombosis was not identified on immediate postoperative angiogram (B). A follow-up angiogram at 6 months revealed insignificant (30%) in-stent stenosis (C). A follow-up angiogram at 24 months identified no in-stent stenosis and restoration of luminal integrity (D).

A.

B.

C.

D.

In a total of 77 aneurysms with angiographic follow-up, residual aneurysm was observed in 20 aneurysms (26.0%) while complete occlusion was found in 57 aneurysms (74.0%). In small aneurysms (<10 mm; n=67), 77.6% of aneurysms (n=52) showed complete occlusion at follow-up while in large aneurysms (>10 mm, <25 mm; n=9), 55.6% (n=5) showed complete occlusion (Table 8).

Table 8.

Follow-up angiographic results.

		Follow-up Results (n=77)
		Remnant or Recanalized	Complete Occlusion
		(n=20, 26.0%)	(n=57, 74.0%)
Aneurysm Size	Small (n=67)	15 (22.4%)	52 (77.6%)
	Large (n=9)	4 (44.4%)	5 (55.6%)
	Giant (n=1)	1 (100%)	0
Initial Angiographic	PO (n=17)	4 (23.5%)	13 (76.5%)
Results	NC (n=36)	9 (25.0%)	27 (75.0%)
	CO (n=24)	7 (29.2%)	17 (70.8%)
Small, <10mm; Large, 10>mm, <25mm; Large, >25mm; PO, partial occlusion; NC, near-complete occlusion; CO, complete occlusion

Thromboembolism

Thromboembolic events remained the most common complications, occurring in 17 patients (13.4%, 17/127). Ten cases occurred less than 24 hours after the procedure (58.8% of thromboembolic events, 10/17), two cases within five days (11.8%, 2/17), three cases within two weeks (17.6%, 3/17), one case at one month post procedure (5.9%, 1/17), and one case two months post procedure (5.9%).

In-stent thromboses were observed in five patients (29.4% of thromboembolic events, 5/17). Two patient exhibited intra-procedural in-stent thrombosis. One autolyzed and the other was treated with intra-arterial (IA) tissue plasminogen activator (TPA). One patient with a ruptured BA aneurysm presented eight days post procedure with in-stent occlusion combined with distal basilar and bilateral posterior cerebral artery (PCA) occlusion. Despite IA TPA and mechanical manipulation, this patient died. One patient developed symptoms from in-stent stenosis one month after the procedure concomitant with cessation of antiplatelet agents due to gastrointestinal bleeding (Figure 3). Another patient exhibited symptomatic in-stent occlusion two months after stent-assisted coil embolization.

Figure 3.

A 66-year-old woman presented with subarachnoid hemorrhage. Her initial left vertebral artery angiogram on AP view revealed a wide-necked left superior cerebellar artery aneurysm (A). She underwent stent-assisted coil embolization (B). After discharge, she was prescribed 325 mg of aspirin and 75 mg of clopidogrel. However, she stopped medication due to gastrointestinal bleeding two weeks after the procedure. One month after the procedure, she suddenly exhibited right hemiparesis and hemihypesthesia. MRI revealed cerebral infarction (C) and left posterior cerebral artery in-stent thrombosis was identified on left vertebral angiogram (D).

A.

B.

C.

D.

Six thromboembolic cases resulted in persistent neurologic morbidity and mortality (35.3% of thromboembolic cases; 4.7% of all cases). One mortality and five morbidities were identified (Tables 6 and 7). The one mortality was due to in-stent occlusion on the eighth post-procedure day. Two morbidities were due to in-stent thromboses that presented one and two months after the procedure. Coil herniation and air embolism led to persistent neurologic morbidity in two cases. In one case with an unruptured ACom aneurysm, diffuse multi-focal infarctions were identified in both anterior cerebral artery (ACA) territories even though the procedure was uneventful. Overall, in 127 patients, the morbidity and mortality rates from thromboembolic events were 3.9% (5/127) and 0.8% (1/127), respectively.

Intraoperative Aneurysm Rupture

Intraoperative aneurysm ruptures occurred in eight (6.3%) of 127 patients. Two were spontaneous (without direct device manipulation in or into the aneurysm), two occurred during microwire/microcatheter manipulations within the aneurysm, and four occurred during coil insertion. Coil-induced ruptures were caused by the first coil in two cases, the fourth in one, and the sixth in one. Seven intraoperative ruptures occurred in patients with ruptured aneurysms (14.0%, 7/50) and one in a patient with an unruptured aneurysm (1.3%, 1/77). Preoperative clinical grades of the eight patients were Hunt and Hess grade 0 in one, III in one, IV in five, and V in one. All intraoperative ruptures were managed with intravenous protamine sulfate administration, external ventricular drain insertion if necessary, and further embolization. Seven patients died (Table 6) while one had no identifiable neurologic consequence. In 127 patients, the mortality rate from intraoperative aneurysm rupture was 5.5% (7/127).

Postoperative Aneurysm Rupture

Postoperative aneurysm rupture occurred in four (3.1%) of 127 patients. Two cases involved ruptured aneurysms (Hunt and Hess grade I) and two involved unruptured aneurysms (Table 5). Immediate angiographic results were near-complete occlusion in two cases and partial occlusion in two. All cases occurred within one day of the procedure. All postoperative ruptures led to morbidity and mortality. Three postoperative ruptures resulted in death and one in persistent neurologic morbidity (Table 6 and 7). The morbidity and mortality rates from postoperative aneurysm rupture were 0.8% (1/127) and 2.4% (3/127), respectively.

Coil Herniation

Coil herniation into the parent vessel despite stent placement was observed in five patients (3.9%, 5/127). One case (0.8%, 1/127) led to a thromboembolic event and neurologic morbidity (Table 7). In one case, coil herniation was caused by coil fracture. In three patients, an additional stent was deployed with stent-in-stent technique to pin protruded coil loops against the parent artery wall.

Stent Migration

Stent migration occurred in four (3.1%) of 127 patients. Stent migration was caused by microwire manipulation, microcatheter manipulation, coil hooking onto the stent during retrieval, and delivery device retrieval. In one case, because the aneurysm was still covered by the stent, no additional stent placement was required. The stent was retrieved in two cases and migrated into the parent vessel’s extracranial portion in one case. In three cases, an additional stent was deployed to cover the aneurysm.

Iatrogenic Dissection

Iatrogenic parent artery dissection occurred in three (2.4%) of 127 patients. These involved the distal cervical portion of the internal carotid artery in two cases and extracranial portion of the vertebral artery in one case. All dissections occurred during guide catheter manipulations. In the two cases of cervical internal carotid artery dissection, blood flow was compromised, requiring stent-assisted angioplasty to restore intraluminal integrity.

Discussion

Follow-up Angiographic Results

Angiographic follow-up in our study was achieved in 69 patients at a mean period of 13.7 months. In these 69 patients harboring 77 aneurysms, 57 complete aneurysm occlusions were documented (74.0%, 57/75; mean follow-up period, 13.7 months). In other series with stent-assisted coil embolization using the Neuroform stent, Biondi et Al.¹⁴ reported 17 out of 30 complete aneurysm occlusions (56.7%; follow-up period 9 months) and Fiorella et Al.³⁶ reported progressive thrombosis in 25 out of 48 aneurysms (52.1%; follow-up period, 4.6 months). Weber et Al¹⁸ reported 15 out of 30 complete aneurysm occlusions (50%; follow-up period, six months) in their series using the Enterprise stent (Cordis, Miami, FL).

For several reasons, it is difficult to compare our angiographic results with those of other reported series. First, few papers documented delayed angiographic follow-up in stent-assisted coil embolization of intracranial aneurysms. Second, analysis of the radiographic and clinical data in our series was not performed by an independent investigator. Third, a standard definition of what constitutes complete occlusion is lacking and quantification of results is subjective³⁷. Fourth, the mean follow-up period in this series was 13.7 months, which was longer than other series. The higher complete occlusion rate in this series may be attributable to the longer follow-up period. Fifth, we did not obtain follow-up in all eligible patients for several logistical reasons.

Morbidity and Mortality

In our series, procedure-related morbidity and mortality rates were 5.5% (7/127) and 8.7% (11/127). In other series with stent-assisted coil embolization, morbidity rates ranged from 4.8 to 20% and mortality rates ranged from 0 to 8.9%^10,14,38-44. While in our series, the overall mortality and procedure-related mortality rates were relatively high (16.5%, 8.7%), this may be attributable to the greater number of SAH patients and poor grade SAH patients included in our study compared to others (patients with SAH, 39.4%, 50/127; Hunt and Hess grade IV or V, 19.7%, 25/127). The historical overall mortality rate for poor grade SAH patients undergoing any endovascular therapy is reported as ranging from 32.4 to 59%^45-47. In our series, the overall mortality rate in poor grade SAH group was 56.0% (14/25) compared to 6.9% (7/102) in unruptured and good grade SAH patients.

Thromboembolism

The most frequent complication in this series was thromboembolism (13.4%, 17/127). Historical incidence rates of thromboembolism associated with coil embolization of cerebral aneur­ysms have ranged from 1.0% to 28%^1,13,48-52. Other series looking at stent-assisted coil embolizations report incidence rates ranging from 4.8 to 21.1%^38-40,44.

With regard to in-stent thrombosis, Lee et Al.⁴² reported an acute intraprocedural in-stent thrombosis rate of 4.5% (1/22). Historical incidence rates of subacute or delayed in-stent thrombosis range from 2.4 to 5.8%^14,36,53,54. In our series, the rate of acute intra-procedural in-stent thrombosis was 1.6% (2/127) and the rate of subacute or delayed in-stent thrombosis was 2.4% (3/127).

According to Virchow’s triad, arterial thrombosis under pathologic or procedural circumstances is influenced by alterations in blood flow, endothelial thrombogenicity, and hematologic thrombogenicity (the thrombogenicity of circulating activated physiological and pathological hemostatic factors)⁵⁵. Since platelets play a key role in arterial thrombosis, antiplatelet agents have an effect on hematologic and endothelial thrombogenicity. Because cerebral angiography itself carries a risk of thromboembolism, and most thromboembolic events associated with coil embolization are caused by device manipulation^56,57, it is mandatory to modulate both hematologic and endothelial thrombogenicity before and during the procedure. Administration of antiplatelet agents, when possible, before the procedure may reduce the thromboembolic event rate. Generally, therapeutic platelet inhibition is achieved by using antiplatelet agents (aspirin, ticlopidine, and clopidogrel) for three to seven days unless bolus dosing is utilized, however longer periods may be needed to alter the endothelial surface. A combination of aspirin and clopidogrel for four weeks prior to a procedure or aspirin and ticlopidine for two weeks prior to a procedure may be more effective in preventing in-stent thrombosis^58,59.

In patients with ruptured aneurysms, pre-procedural antiplatelet agent administration is a matter of debate, but in patients with unruptured aneurysms, adequate antiplatelet agent premedication is essential. It is recommended that adequate antiplatelet therapy should be sustained at least one month after the procedure^60,61. In our series, one case of symptomatic in–stent thrombosis was associated with cessation of all antiplatelet agents due to gastrointestinal bleeding. A second case of symptomatic in-stent occlusion presented two months post procedure despite antiplatelet therapy.

While thromboembolic events associated with the Neuroform stent are problematic, they are not exclusive to this device. Balloon-assisted neck-remodeling technique can provide an alternative method for endovascular treatment of wide-necked aneurysms^62-67. However, this technique has similar problems. Van Rooij et Al⁶⁸ reported in their series of 681 patients with ruptured intracranial aneurysms that the combined morbidity and mortality rate from thromboembolic events with balloon-assisted neck-remodeling technique was 16.3% (8/49) and this technique was the only risk factor for intraoperative rupture and thromboembolic events. In our series, the combined morbidity and mortality rate for ruptured aneurysms secondary to thromboembolic events was 6.0% (3/50; two morbidities and one mortality). Unlike balloons, however, stents can permanently prevent coil protrusion into the parent artery, may redirect blood flow, may promote endothelialization at the aneurysm neck, and do not precipitate sudden intra-aneurysmal pressure changes^14,40,44,62. Stents can also secure protruded coil loops against the parent artery wall in case of protrusion from the fundus after coiling is complete.

Intraoperative and Postoperative Rupture

Intraoperative aneurysm rupture was the second most common complication in our series (6.3%, 8/127). Historical rates of intraoperative aneurysm rupture associated with coil embolization range from 1.9 to 16% for ruptured aneurysms and from 0 to 1.3% for unruptured aneurysms^{12,13,15,69-72}. In our series, rupture rates were 14.0% (7/50) and 1.3% (1/77) for ruptured and unruptured aneurysms, respectively. Aneurysm size is thought to be inversely associated with an increased incidence of intraoperative rupture. Vin˜uela et Al¹³ reported that nine of eleven intra-operatively ruptured aneur­ysms measured 4-10 mm in diameter and Sluzewski et Al⁷¹ reported that six of seven measured 2-10 mm.

In our series, six of eight ruptures involved small lesions (<10 mm). Six cases of intraoperative aneurysm rupture in this series were induced by device (wire/ca­theter/coil) manipulations and two occurred spontaneously without direct device manipulation. All spontaneous intraoperative ruptures occurred in patients with acute subarachnoid hemorrhage. Possible causes of rupture include stretching of the aneurysm wall after stent deployment, anticoagulation effect and fundal perforation.

Naidech et Al⁷³ reported that a poor Hunt and Hess grade was a strong predictor of rebleeding. In our cases of spontaneous intraoperative aneurysm rupture, all patients were Hunt and Hess grade IV. Once an aneurysm ruptured, the outcome was dismal. Seven of eight (87.5%) patients died. In poor grade SAH patients (Hunt and Hess grades IV or V), all six patients died.

Postoperative aneurysm rupture occurred in four patients (3.1%, 4/127), resulting in universally poor clinical outcomes (three died; one disabled). In our series, rupture rates were 4.0% (2/50) and 2.6% (2/77) for ruptured and unruptured aneurysms. All postoperative ruptures occurred within one day after the procedure. Two of four postoperative ruptures involved small lesions (<10 mm), one involved a large (>10 mm, <25 mm), and one involved a giant aneurysm (>25 mm). In two postoperative ruptures with ruptured aneurysms/SAH, initial clinical status was good (Hunt and Hess grade I). One lesion was small (3 mm) and relatively spherical. The other was large (12.6 mm) and irregular in shape. In two postoperative ruptures with unruptured aneurysms, one patient presented with progressively worsening severe headache and angiography revealed a small (8.6 mm), but irregular shaped aneurysm. The other patient exhibited a partially thrombosed giant aneurysm. Incomplete occlusion is thought to be associated with postoperative rupture. In our four postoperative ruptures, immediate post-coiling angiographic results showed partial occlusion in two patients and near-complete occlusion in two patients.

Conclusions

Key findings in this review of 129 stent-assisted coil embolization procedures in 127 patients with 136 aneurysms were as follows:

1) Forty seven patients presented with 53 procedure-related complications (37.0%, 47/127).

2) Overall mortality rate was 16.5% (21/127) and procedure-related morbidity and mortality rates were 5.5% (7/127) and 8.7% (11/127), retrospectively.

3) Thromboembolism (n=17, 13.4%) and intra/postoperative aneurysm ruptures (n=8/4, 6.3%/3.1%) were the most common complications and the main causes of procedure-related morbidity and mortality.

4) Patients presenting with a clinically poor grade SAH had dismal clinical outcomes (overall mortality rate, 56%; procedure-related mortality rate, 24%).

5) Since most complications were induced by device manipulation, it is mandatory to handle devices cautiously.

6) Follow-up angiographic results in 65.1% of eligible patients showed complete occlusion in 74.0%.

7) Further studies are essential to evaluate the safety, efficacy, and durability of the Neuroform stent.