Stent-assisted coiling of ruptured wide-necked intracranial aneurysms: A single-center experience of 218 consecutive patients
Background: Stent-assisted coiling (SAC) in the treatment of ruptured intracranial aneurysms (RIAs) remains controversial.
Keywords: Ruptured; stent-assisted coiling; subarachnoid hemorrhage; wide-necked aneurysm
The International Subarachnoid Aneurysm Trial (ISAT) showed that endovascular coiling of intracranial aneurysms could achieve improved outcomes with lesser morbidity and mortality rates when compared with surgical clipping.  However, endovascular treatment of wide-necked or broad-based intracranial aneurysms have remained a technically challenging procedure. Incomplete occlusion achieved by conventional endovascular techniques often leads to subsequent continued aneurysmal growth. Consequently, such treatments are prone to re-bleeds and retreatments.  Recently, the introduction of intracranial stents as an adjunct to endovascular coiling has widened the applicability of stent-assisted coiling (SAC) to wide-necked aneurysms. ,, These stents could potentially allow reconstruction of the aneurysm neck and hence may permit a more complete occlusion of the aneurysmal sac.
However, stents always have a significant risk of thromboembolism and subsequent ischemic complications. Previously published data has shown a complication rate of 8.5 to 25% in the setting of ruptured aneurysms. ,,,,,,,,, To circumvent this potentially fatal problem, these patients are often administered and maintained on dual antiplatelet therapy or anticoagulation drugs. In the setting of subarachnoid hemorrhage (SAH), these drugs significantly heighten the risk of hemorrhagic complications, particularly when some other intervention, most commonly external ventricular drainage (EVD), has to be performed.  Hence, majority of the existing literature regarding SAC of wide-necked aneurysms has largely remained confined to the patients with unruptured aneurysms and SAC in the setting of aneurysmal SAH remains controversial. ,,,,,,,,, To the best of our knowledge, only one study has evaluated the independent factors affecting the risks of SAC for RIAs. In that study, in 36 (50%) of the 72 patients with acutely ruptured intracranial wide-necked saccular aneurysms, SAC was performed in the setting of EVD. Utilizing multivariate logistic regression analysis, the authors concluded that the use of EVD was the only independent risk factor for periprocedural complications. 
With the above background, we performed this retrospective, single-center study to analyze the feasibility and efficacy of SAC in ruptured wide-necked aneurysms in both the acute and sub-acute settings. At the same time, a multivariate analysis was performed to find out the independent factors that were significantly related to the periprocedural complications arising as a result of these interventions.
Patient and aneurysm characteristics
Between January 2011 and December 2014, 641 consecutive patients with ruptured aneurysms were treated using endovascular techniques at our institution. Among them, 218 patients (34%) with saccular aneurysms were treated using stents. All procedures were performed by the senior author (ML), with more than 10 years of experience in intervention neuroradiology at our hospital. The exclusion criteria were the following: (1) SAC was used as a stent salvage technique after a procedure-related complication (n = 4), (2) ruptured dissecting aneurysms were present (n = 17), (3) blister-like aneurysms were present (n = 5), and (4) multiple aneurysms were present (n = 15).
With Institutional Medical Ethics Committee approval and patient consent, the clinical presentation, angiographic features and the management of these 218 patients were reviewed retrospectively. Their mean age was 57.8 ± 10.6 years (range from 34 to 82 years) and majority of the patients (160; 73.4%) were female. The Hunt and Hess grade (H and H) distribution was as follows: Grade I in 154 patients (70.7%), II in 21 patients (9.6%), III in 31 patients (14.2%), IV in 11 patients (5.0%), and V in 1 patient (0.5%). Clinically, good-grade patients were defined as those with initial Hunt and Hess grades I to III; while, poor-grade patients were defined as those having H and H grade IV or V.
Conventional as well as three-dimensional (3D) angiography were used to evaluate all the patients, particularly noting features like aneurysm location, the maximum diameter of the aneurysm at the dome and neck; and the dome-to-neck ratio (DNR). Aneurysmal neck >4 mm in diameter or those with a DNR <2 were defined as wide-necked.
Treatment decisions were made after a joint discussion between the treating parties involved (the neurosurgeons and the interventional radiologist). Points considered in the process were the patient's clinical status and feasibility of performing either surgery or coiling. Surgical clipping was preferred for anterior circulation aneurysms, especially if the patient had a large parenchymal hematoma. Endovascular procedure was usually preferred over surgical clipping in older patients and those with a poor neurological status. If the patient had the potential need for an EVD, a SAC procedure was avoided and coiling without the protection of a stent (balloon-assisted coiling or double-catheter technique) was considered. In our case series, patients with a clinically poor grade (n = 2) or with acute hydrocephalus underwent an EVD (n = 4) before the embolization to reduce the risk of catheter-related hemorrhagic complications due to antiplatelet therapy.
Intraprocedural heparinization was achieved by adding 3000 IU of heparin to 500 mL of 0.9% (wt/vol) sodium chloride solution which was administered through the guiding catheter during the procedure. , A loading dose of 300 mg aspirin and 225 mg clopidogrel was given at least 4 h before the procedure. After the procedure, the patients were maintained on daily clopidogrel (75 mg) and aspirin (100 mg) for 6 weeks, followed by aspirin alone (100 mg daily) for another 6 months. During the procedure, thromboembolism complications were managed with 0.5 mg loading dose of glycoprotein (GP) IIb/IIIa antagonist (tirofiban) injected intra-arterially.
All procedures were performed under general anesthesia using an unilateral femoral access. The Enterprise stent (Codman Neurovascular, Miami Lakes, FL, USA), Solitaire stent (Solitaire AB neurovascular remodeling device, eV3, Inc., Irvine, CA, USA), and the low-profile Visualized Intraluminal Support Junior (LVIS Jr; MicroVention-Terumo, Tustin, CA) stent were used. The SAC procedure was done utilizing a "jailing" technique in which a coiling microcatheter was introduced into the aneurysm sac, and the first coil was inserted without detachment. Following this, a stent was deployed, and further coils were introduced subsequently. According to the size and shape of the aneurysm, appropriate coils were selected. We inserted the coils within the aneurysm sac as densely as possible until we achieved a complete angiographic occlusion, or it was not possible to introduce coils any further. After the procedure, a three dimensional angiography was obtained to assess the result.
Thromboembolic complications were defined as angiographic evidence of thrombosis during or at the end of SAC, or the development of new ischemic manifestations that were evident on computed tomography (CT), or magnetic resonance imaging (MRI) scans done subsequently. Leakage of contrast material during the procedure or newly developed hemorrhage evident on CT or MRI studies within 30 days of the procedure were included as hemorrhagic complications. Hemorrhages that occurred along the tract of the ventricular catheter were considered to be EVD related and not as a hemorrhagic complication of the SAC. Delayed ischemic stroke was defined as a transient ischemic attack or a frank stroke with evidence of infarction on diffusion-weighted images occurring in the vascular territory of the stented artery beyond 1 month of the stent-assisted coil placement.
Clinical and angiographic follow-up
Two independent investigators (YL, XL) assessed the initial and follow-up occlusion information and categorized the results according to the 3-point Raymond Scale (RS1: Complete obliteration of the aneurysm and neck; RS2: Neck remnant without contrast filling the aneurysm sac; and, RS3: Contrast filling the aneurysm sac.  Follow-up angiograms included either the digital subtraction angiography (DSA) or computed tomographic angiography. We defined aneurysm recanalization when a totally occluded aneurysm showed a partial recurrence of the neck and/or the sac, or if a subtotally occluded aneurysm had an increasing size of the neck remnant or residual aneurysm at follow-up. Clinical outcomes were obtained using the Glasgow Outcome Scale (GOS), the relevant information being collected over telephonic interview with the patients. A favorable outcome was defined as a GOS score of 4 or 5 (moderate disability or better) while a poor outcome was defined as a GOS score of 2 or 3 (severe disability or vegetative state). A GOS score of 1 meant death.
SPSS version 19.0 (SPSS, Inc., Chicago, IL, USA) was used for all statistical analyses. Normally distributed continuous data were summarized as mean ± standard deviation and categorical data as frequency and percentage. Multivariate logistic regression analysis was then performed on age, sex, hypertension, daughter blebs, Hunt-Hess grade, aneurysm location, aneurysm size, timing of endovascular embolization, types of stent and requirement of EVD, to analyze the significance of these known risk factors on periprocedural complications. A P value of < 0.05 was considered statistically significant.
Patient demographics and aneurysm characteristics
Out of the 218 aneurysms in 218 patients, 200 were in the anterior circulation while only 18 had involved the posterior circulation. The locations of the anterior circulation aneurysms included internal carotid artery (ICA)-posterior communicating artery (n = 138), ICA-ophthalmic artery (n = 34), ICA-anterior choroidal artery (n = 1), ICA-carotid terminus (n = 5), anterior communicating artery (n = 14), middle cerebral artery (n = 7), and A1 segment of the anterior cerebral artery (n = 1). The locations of the posterior circulation aneurysms consisted of basilar artery (n = 10), posterior cerebral artery (n = 2), superior cerebellar artery (n = 4), anterior inferior cerebellar artery (n = 1), and posterior inferior cerebellar artery (n = 1).
EVD was performed in 6 patients out of which 2 patients required an EVD before the SAC; while, 4 patients underwent EVD subsequent to the embolization. Decompressive craniectomy was performed in 1 patient with a subdural hemorrhage that occurred 1 month after the endovascular procedure.
Initial and follow-up digital subtraction angiography
The initial DSA after the procedure showed RS1 occlusion in 182 aneurysms (83.5%), and RS2 occlusion in 36 aneurysms (16.5%). Conventional angiographic follow-up that extended longer than 6 months was available in 178 of 210 patients (84.8%) during a mean follow-up of 19.5 ± 11.9 months (range from 7 to 50 months). All the patients remained stable without aneurysmal regrowth. This included 158 patients with complete occlusion and 20 with near-complete occlusion. None of the aneurysms exhibited residual filling of the sac. Conventional angiographic follow-up that extended longer than 12 months was available in 130 out of 210 patients (61.9%), the mean followup being 23.4 ± 11.7 months (range from 12 to 49 months). It showed 117 complete and 13 near complete occlusions. There was no significant in-stent stenosis (defined as >50% stenosis) or aneurysm recurrence seen on follow-up angiography.
Procedure-related complications occurred in 33 patients (15.1%) leading to death in 4 (1.8%) patients and a permanent morbidity in 9 (4.1%) patients.
Thromboembolic complications occurred in 17 patients (7.8%) including an in-stent thrombosis in 4, stent-unrelated thrombosis in 7, and new symptomatic infarcts in 6 patients. Eleven patients with thromboembolic complications including in-stent thrombosis and stent-unrelated thrombosis underwent intra-arterial thrombolysis in which a glycoprotein IIb/IIIa antagonist (tirofiban) was used. Nine of these patients experienced successful recanalization, and 6 patients still had morbid complications.
Hemorrhagic complications occurred in 16 patients (7.6%) of which 4 patients were symptomatic. Hemorrhage, revealed by the leakage of contrast agent during the procedure was detected in 13 patients, and all these patients were managed with rapid and dense packing of the aneurysm. Two patients suffered from persistent hemiparisis due to the hematoma. Two patients died of the aneurysmal re-bleeding during the procedure.
Another 3 patients experienced intraparenchymal hemorrhage, out of which 2 patients suffered an EVD-related bleeding at day 3 and day 13 after the procedure, respectively, and both patients were on dual antiplatelet therapy at the time of hemorrhage. The other patient developed subdural hemorrhage 29 days after the procedure and was treated by hematoma evacuation and decompressive craniectomy. The patient recovered uneventfully.
Delayed ischemic complications occurred in 3 patients (0.12%) at a mean follow-up of 11 months. Three patients suffered from intermittent limb numbness after discontinuation of the antiplatelet therapy.
Univariate analysis showed no risk factors for the periprocedural thromboembolic complications [Table 1]. Multivariate logistic regression analysis showed that a history of hypertension (odds ratio [OR] 4.899, 95% confidence interval [CI] 1.266-18.951; P = 0.021), and presence of a daughter bleb (OR 12.165, 95% CI 3.247-45.577; P = 0.0001) were the significant risk factors for periprocedural hemorrhagic complications [Table 2].
Of the 218 patients, 8 patients died resulting in a mortality rate of 4.0% in our series. Four of them died of causes unrelated to the presence of the aneurysm. These included pulmonary embolism in 1 patient and chest infection in 1 patient in the immediate postprecedure period, head trauma in 1 patient after a year of the endovascular procedure, and acute kidney failure in 1 patient (who had no prior history of renal disease) 20 days after the endovascular procedure. Two patients died of aneurysmal rerupture during the procedure, and 2 died of EVD-related hemorrhage. Of the remaining 210 patients, 201 (95.7%) had a favorable GOS score of 4 or 5 during the follow-up period (29.1 ± 16.2 months, range from 7 to 53 months). Nine patients (4.3%) had an unfavorable outcome.
Of the 17 patients with thromboembolic complications, midterm follow-up was available for all the patients. Eight patients had a favorable GOS score. The other 9 patients suffered from an unfavorable outcome and had persistent neurological deficits. Of the 16 patients with hemorrhagic complications, midterm follow-up was available for the surviving 12 patients and all of the latter patients had a favorable GOS score of 4 or 5.
In the treatment of wide-necked aneurysms, remodeling techniques such as balloon-assisted coiling  or SAC are known to prevent coil protrusion. ,, The clinical and anatomic results in 'the Treatment of Ruptured Intracranial Aneurysms' study have ascertained that in the treatment of ruptured aneurysms, the remodeling technique was as safe as the conventional coil embolization and had higher rates of adequate postoperative aneurysmal occlusion.  Several recent single institution series have reported initial occlusion rates ranging from 53% to 98.1% with stent assistance. ,,,,, The initial complete occlusion rate in our series was 83.5%. With the assistance of stents, proper reconstruction of the neck and sufficient occlusion of the aneurysmal sac could be achieved. During the follow-up period, no patient suffered from aneurysm-related rebleeding.
However, the use of a stent in the setting of ruptured aneurysms remains controversial since the deployment of intracranial stents necessitates administration of dual antiplatelet therapy to minimize thromboembolic complications.  After reviewing the literature, the complication rates reported in the previous studies on the SAC treatment of ruptured wide-necked aneurysms in the acute period have ranged from 8.5% to 25%. ,,,,,,,, In our study, thromboembolism was the most common procedure-related adverse event (7.8%). The thromboembolic complication rate in the acute period (<3 days) in our study was 6.7%, which was similar to two earlier reports. , In 2012, Amenta et al.,  reported a retrospective study of 65 patients with acutely (<3 days) ruptured wide-necked aneurysms treated with SAC. The intraoperative in-stent thrombosis was found in 5 patients (7.7%) and 3 (4.6%) patients had a fatal hemorrhage. Patients were treated intraoperatively with 600 mg of clopidogrel and maintained on daily doses of 75 mg of clopidogrel and 81 mg of aspirin in the postoperative period. In 2011, Bodily et al.,  reviewed the literature regarding stent-assisted coil embolization of 339 acutely (<7 days) ruptured aneurysms and found a clinically significant thromboembolic event in 16 (6%) of them. Clinically significant intracranial hemorrhagic complications occurred in 27 (8%) patients. No patients in their previous studies had received antiplatelet therapy or heparin before the procedure. They were administered dual antiplatelet therapy after the procedure. In 326 (96%) of 339 cases, dual antiplatelet therapy, usually aspirin and clopidogrel, was administered after the procedure. Recently, preprocedural loading doses of clopidogrel and aspirin have been applied in several studies, and yet the incidence of thromboembolic complications have remained rather high (11.2% to 20%). ,,, A recent study by Chung et al., in 2014  reported a series of 72 patients with wide-necked aneurysms who underwent SAC within 72h of the ictus. Before the procedure, loading doses of clopidogrel (300-600 mg) and aspirin (200-400 mg) were given via the nasogastric tube. The procedure-related thromboembolic and hemorrhagic complication rates were 12.5% and 6.9%, respectively. After the procedure, patients were prescribed 75 mg clopidogrel daily for 3 months and 100 mg aspirin daily for a minimum of 12 months. The authors noted that the placement of an EVD catheter was the only independent risk factor for periprocedural complications (the incidence of periprocedural complications was 30.6% in patients who underwent an EVD). In their study, 36 of the 72 patients received EVD primarily because of the higher proportion of poor grade SAH in their series (29.2%). The authors thus concluded that microsurgical clipping or endovascular treatment with another technique (multiple microcatheters or balloon-assisted technique) may be a more appropriate option as the first-line treatment of wide necked RIA, especially in patients requiring an EVD. We agree with these conclusions drawn by previous authors. In our study, only 12 (5.5%) of the 218 patients manifested in the clinical status Hunt-Hess grade IV-V. In addition, in our case series, EVD was performed in 6 patients, and 2 patients (33.3%) suffered from EVD-related hemorrhage. In our hospital, we strictly limited the indications for the SAC treatment of patients with a ruptured aneurysm in such a way that those who were in need of an EVD procedure either underwent a double-catheter technique/balloon-assisted coiling or surgical clipping (if the alternative endovascular strategies were not feasible).
Previous data has shown that acute SAH was a hypercoagulable state  and the likelihood of thrombosis was high. Colwell et al.,  found that thrombin production was significantly increased in SAH. In our study, the complication rate was not different in the acute or the sub-acute period. Also, there was no significant difference between the anterior and the posterior circulation aneurysms with respect to the thromboembolic complications, which is in agreement with the findings of Nilsson et al. 
In our study, all coil perforations occurred in small aneurysms (measuring less than 10 mm), which was similar to the experience in the previous studies. , In our study, the presence of hypertension and daughter bleb were found to be the risk factors for periprocedural hemorrhagic complications. As far as the history of hypertension is concerned, the blood pressure must be intensively managed in the normal range during the procedure. We believe that the aneurysm wall in patients with a history of hypertension might be more fragile and may be more prone to rupture. In aneurysms with daughter blebs, our experience suggested that using smaller coils with careful manipulation might help. Avoiding the selection of oversized coils or stiffer coils may relieve the pressure against the aneurysm wall thus preventing the inadvertent hemorrhagic complications.
Although performance of EVD before the coil embolization and the initiation of dual antiplatelet therapy has been suggested, such a strategy may not avoid delayed hemorrhagic events. A meta-analysis of 1790 ventriculostomies, with regard to the hemorrhagic risks related to EVD, has revealed a hemorrhage rate of 5.7%, with a clinically significant hemorrhage rate of 0.61%.  Tumialan et al.,  have noted the elevated risk of hemorrhage associated with dual antiplatelet therapy in patients with EVD. In our study, EVD was performed in 6 patients. Unfortunately, 2 patients suffered from EVD-related hemorrhage. In one patient each, with placement of a preembolization and a postembolization EVD, respectively, had EVD-related bleeding occurred due to the dual antiplatelet therapy after the procedure.
Specific guidelines regarding the management of patients who developed intracranial hemorrhage while on dual antiplatelet medication is not available.  Whether or not to withdraw the antiplatelet drugs remains a controversial issue. Kim et al.,  reported that 2 of their 4 patients who stopped dual antiplatelet therapy after hemorrhage developed occlusion of the stent site and suffered new infarction in the corresponding arterial territory. During the postprocedural period, one of our patients had intracranial hemorrhage while on dual antiplatelet therapy. We discontinued clopidogrel and continued aspirin, and fortunately the patient did not experience stent-related thromboembolic complications.
In patients who have undergone a SAC, the reported regrowth rate in the wide-necked ruptured aneurysms ranges from 0% to 13.8%. ,,, In 2003, Raymond et al.,  reported a regrowth in 76 of 191 aneurysms (39.8%) that were treated acutely after rupture with the traditional detachable coils without the assistance of a stent. They concluded that the duration of follow-up was also an important predictor of aneurysm recurrence. In 2012, Jia et al.,  reported recanalization of 10 of the 119 aneurysms (8.4%) during the mean follow-up of 8 months after the first procedure. In 2015, Song et al.,  reported on the long-time efficacy in treating 53 wide-necked aneurysms with coil embolization (coiling alone in 45; SAC in 8) and suggested that an angiography at follow up should be performed after at least 36 months. In our study, an angiographic follow-up exceeding 36 months was available for 30 of the 178 patients and none of the patients showed any angiographic recurrence.
First, this is a retrospective study without randomization and the effects of antiplatelet agents on thromboembolic as well as hemorrhagic complications have not been compared. Second, the individual response to each antiplatelet agent was not evaluated. Third, the follow-up was relatively short as a longer angiographic follow-up is needed to look for future recanalization.
SAC is an effective treatment option for selected ruptured wide-necked aneurysms, especially in patients who do not require a subsequent EVD. The hemorrhagic complications that may occur following the procedure have a significant association with the presence of hypertension and daughter blebs.
Financial support and sponsorship
This article is supported by the Commission of Beijing Municipal Science and Technology, municipal clinical special application study, the special fund project (No.Z14110000211441); National Natural Science Foundation (81171078 and 81471166); Beijing Talents Training Project (Category D), Beijing Hygiene System High-level Hygienic Technical Personnel Training Program and the Talents Program of Beijing Tiantan Hospital (Hospital Backbone Program).
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2]