Research Human Clinical Studies

Transcription

1 TOPIC Hilal Kanaan, MD Brian Jankowitz, MD Aitziber Aleu, MD ; and Department of Neurology, Universitat Autonoma de Barcelona, Barcelona, Spain Dean Kostov, MD Ridwan Lin, MD Department of Neurology, Kimberly Lee, RN, BSN Narendra Panipitiya, BS Yakov Gologorsky, MD Mount Sinai School of Medicine, New York, New York Emir Sandhu Lauren Rissman Elizabeth Crago, RN, MSN Yue-Fang Chang, PhD Seong-Rim Kim, MD ; and Holy Family Hospital, Catholic University of Korea Seoul, Korea Tudor Jovin, MD Department of Neurology, Michael Horowitz, MD Reprint requests: Michael Horowitz, MD, 200 Lothrop St, Ste B400, Pittsburgh, PA [email protected] Received, July 14, Accepted, July 19, Copyright ª 2010 by the Congress of Neurological Surgeons In-Stent Thrombosis and Stenosis After Neck- Remodeling Device-Assisted Coil Embolization of Intracranial Aneurysms BACKGROUND: Intrinsic thrombosis and stenosis are complications associated with the use of neck-remodeling devices in the treatment of intracranial aneurysms. OBJECTIVE: To examine the technical and anatomic factors that predict short- and longterm stent patency. METHODS: We undertook a retrospective review of 161 patients who underwent coil embolization of 168 ruptured and unruptured aneurysms assisted by the use of a neckremodeling device. One hundred twenty-seven patients had catheter-based angiographic follow-up to evaluate 133 stent-coil constructs (mean, 15.4 months; median, 12.7 months). The technique of microcatheter jailing was used in a majority of patients; nonstandard stent configurations were also used. RESULTS: Clinical follow-up for all patients who had catheter-based angiograms demonstrated that among 133 stent constructs, a total of 9 (6.8%) had an in-stent event: 6 acute or subacute thrombosis (4.5%) and 3 delayed stenosis or occlusion (2.3%). Seven of these constructs were associated with a symptomatic event (5.3%). A significantly higher rate of in-stent events was seen with the use of constructs to treat anterior communicating artery aneurysms. When all patients are considered, including those who did not receive catheter-based follow-up imaging, 2 of 168 procedures (1.2%) resulted in the death of a patient, and procedural morbidity was 14.9%. CONCLUSION: From these results and those in the published literature, in-stent complication rates are low in carefully selected patients. The use of dual antiplatelet therapy, sensitivity assays, and glycoprotein IIb/IIIa inhibitors may decrease the rate of acute and chronic in-stent complications. KEY WORDS: Aneurysm, Coil, In-stent stenosis, Neck-remodeling device, Stent, Stent-assisted coiling, Stent-coil Neurosurgery 67: , 2010 DOI: /NEU.0b013e3181f8d194 Stent-assisted coil embolization of intracranial aneurysms has greatly expanded the types of lesions that can be treated endovascularly. 1 Although coil embolization has been demonstrated to be safe and effective, 2,3 questions remain about the morbidity associated with neck-remodeling device (NRD) placement in terms of short- and long-term patency. ABBREVIATIONS: AComm, anterior communicating artery; GP, glycoprotein; NRD, neck-remodeling device; SAH, subarachnoid hemorrhage RESEARCH HUMAN CLINICAL STUDIES Research Human Clinical Studies Intracranial stenting was first developed to treat flow-limiting atherosclerotic lesions, and devicerelated restenosis and acute thrombosis echoed the interventional cardiology experience. Although acute thrombus formation and stenosis may be dependent on factors intrinsic to all stents, the pathophysiological underpinnings of thrombosis and restenosis of stented atherosclerotic vessels may not apply to otherwise healthy vessels that are stented with low-radialforce devices to reconstruct wide-necked aneurysms. The following study is a retrospective report regarding in-stent thrombosis and stenosis related to the use of NRDs. NEUROSURGERY VOLUME 67 NUMBER 6 DECEMBER

2 KANAAN ET AL METHODS Patients We performed a retrospective review of all neurointerventional cases in which a stent was deployed in the treatment of intracranial aneurysms between September 2003 and March All interventions and standard follow-up at 6, 12, and 24 months were performed with conventional cerebral arteriography. As part of an ongoing data collection previously published in 2008, patients have been treated with 168 stent constructs. To be included in the final analysis of anatomic outcomes, patients needed to have had a stent deployed in the context of aneurysm treatment, no atherosclerotic disease, and at least 1 angiographic follow-up. When 2 stents were used to treat 1 aneurysm, this was considered a single stent-coil construct. Five patients had multiple stents placed. Four of these patients received 2 stents each for 2 different aneurysms, and the fifth patient had 3 stent constructs placed on separate occasions for the treatment of a wide-necked complex giant aneurysm. Thirty-four patients with 35 stent constructs could not be evaluated. Fourteen of 34 were excluded because their stent-coiling was performed within the last 6 months and thus were not eligible for routine follow-up. Eleven of 34 patients were lost to follow-up. One of 34 patients had a craniotomy with clip sacrifice of the middle cerebral artery proximal to the stent-coil construct before repeat angiography. Eight of 34 patients died before repeat angiography. Of these patients who died, 1 died during elective coiling of intraoperative rupture, 1 died of sepsis after a complicated 60-day hospitalization after aneurysmal rupture and stentcoiling, 3 died after progression of poor-grade subarachnoid hemorrhage (SAH), and 3 died after rehemorrhage of a recently treated ruptured aneurysm. Perioperative rehemorrhage was defined as evidence of new SAH on computed tomography without evidence of perforation or intraprocedural contrast extravasation. In total, the anatomic results of 127 patients with 133 stent constructs were analyzed. Technique The technique of stent-assisted coiling was described previously by our group. 4 All procedures were performed under general anesthesia. Neurophysiological monitoring was used for all cases, which included somatosensory evoked potentials, electroencephalography, and brainstem auditory evoked potentials for posterior circulation aneurysms. The right femoral approach was routinely used for placement of a 6F Envoy (Cordis, Miami, Florida) or 7F Shuttle (Cook, Bloomington, Indiana) guide catheter. Microcatheters (Rapid Transit, Cordis Echelon, ev3, Irvine, California) and Nexus/NXT (ev3) and/or Micrus (Micrus Co, Mountain View, California) coils were primarily used. Cerebral angiography was performed to evaluate aneurysm morphology and to establish working projections. Coils were introduced into the aneurysm by use of simultaneous biplanar imaging. Coil placement proceeded until no additional coils could be placed, until angiographically complete occlusion was achieved, or until risks outweighed perceived benefits. In patients with unruptured aneurysms in which stent placement was anticipated, a dual antiplatelet regimen consisting of 325 mg aspirin and 75 mg clopidogrel (Plavix, Bristol-Meyers Squibb/Sanofi Pharmaceuticals, New York, New York) was administered orally for 5 days before the procedure. When the decision to stent was made intraprocedurally or with patients who presented with SAH, 325 mg aspirin and 600 mg Plavix were administered through an orogastric or nasogastric tube during the procedure after stent deployment. During stent positioning and placement, kinking or bending of vessels was avoided. Only after several coils had entered the aneurysm would full heparinization take place with a goal activated clotting time of approximately 250 seconds. Eptifibatide 15 mg (Integrilin, Schering, Kenilworth, New Jersey) was administered intravenously once the stent was deployed unless contraindicated (ie, intraoperative rupture). Heparin (500 U/h IV) was continued for 12 to 15 hours in most cases. Patients continued taking 75 mg clopidogrel daily for 1 month after the stenting procedure and 325 mg aspirin for life. Three types of NRDs were used: the Neuroform (Boston Scientific, Natick, Massachusetts), Enterprise (Cordis, Warren, New Jersey), and Wingspan (Boston Scientific) stents. Typically, the undeployed stent was positioned across the aneurysm neck, followed by placement of the microcatheter into the aneurysm fundus. The first coil was partially deployed into the aneurysm without detachment to anchor the microcatheter in place. Once the microcatheter was jailed by deploying the stent, the first coil was fully advanced and detached. The most common device used was the Neuroform NRD (116 of 133, 87.2%), usually advanced over an exchange wire (Transcend 14/300 cm; Boston Scientific/Target, Fremont, California; Accelerator 14/300 cm, ev3) and deployed with either the provided pusher system or a coil pusher. The Enterprise Stent (16 of 133, 12%) and Wingspan Stent (1 of 133, 0.8%) were deployed with their respective systems. In 15 of 133 cases (11.3%) in which coils began to unexpectedly herniate out of an aneurysm after deployment, a stent was deployed to maintain the parent vessel. During 14 elective cases (10.5%), the procedure was staged by deploying a stent in the first stage before a second procedure to embolize the aneurysm at least 1 month later. These cases were usually selected because of tortuous anatomy that necessitated a stable stent before manipulation with microcatheters. In such cases, jailing was not used. Stent placement was determined by aneurysm anatomy and location. For most aneurysms, including those at bifurcations, 1 stent was deployed in what shall be referred to as the standard configuration (Figure 1). On occasion, the anatomy of an aneurysm at a bifurcation necessitated 2 stents in a Y configuration (Figure 2), with the proximal ends of both stents in the same parent vessel and the distal ends each in one of the bifurcated arteries. A third stenting technique, called the waffle cone (Figure 3), involved 1 end of a stent placed into the aneurysm neck. 5 A final technique, the coaxial stent configuration, involved the placement of a stent partially or completely within a second stent to increase the FIGURE 1. Standard configuration. An ophthalmic artery aneurysm with Neuroform neck-remodeling device across the neck (arrows demark stent extent) VOLUME 67 NUMBER 6 DECEMBER

3 IN-STENT THROMBOSIS AND STENOSIS FIGURE 2. Y configuration. The microcatheter and undeployed stents are positioned. After the coils are deposited into the aneurysm, the stents are deployed consecutively to maximize apposition to the vessel lumen and to one another. surface area of stent overlying an aneurysm neck. Both the Y and coaxial configurations are considered 2-stent constructs. Data Collection At the time of any angiographic follow-up, 1 of 2 angiographers at our institution commented on the degree of in-stent stenosis. With approval from the institutional review board, data were collected on the types of stents used; the stent construct configuration; whether the microcatheter was jailed before stent deployment; the incidence of in-stent stenosis, thrombus, or occlusion; and complications associated with stent deployment. Results for these 127 patients and 133 stents were subject to statistical analysis with the x 2 and Fisher exact tests. RESULTS Anatomic results of stent-assisted coiling are reported for 127 patients and 133 stent constructs; morbidity and mortality are reported for 161 patients with 168 constructs (Figure 4). One hundred women (78.7%) and 27 men (21.3%) were treated, with an average age of 51.8 years (range, years). Eighty-two patients (64.6%) presented for elective treatment of unruptured aneurysms, and 45 (35.4%) presented with SAH. When each stent is considered its own entity for data collection, average angiographic follow-up was 15.4 months (range, 1 day to 55 months; median, 12.7 months). For all 168 stent-coilings performed (including those that did not meet formal inclusion criteria), 164 initial stent deployments (97.6%) were successful. Of the 133 procedures with imaging follow-up, 126 patients received eptifibatide at the time of stent placement, and 11 of 82 constructs done electively received pretreatment with aspirin and Plavix (Figure 5). Five of 127 patients (3.9%) died; all presented with SAH. One patient had an intraoperative rehemorrhage of an anterior communicating artery (Acomm) aneurysm, and repeat angiography at FIGURE 3. Waffle cone technique as previously published. 12 A, microcatheter and guidewire are advanced into the aneurysm. B, the undeployed stent advanced over the wire. C, The stent is deployed at the neck of aneurysm. D, the microcatheter is advanced back over the guidewire into the aneurysm. E, coiling of the aneurysm. NEUROSURGERY VOLUME 67 NUMBER 6 DECEMBER

4 KANAAN ET AL FIGURE 4. Patient and complication flow diagram by construct type. CVA, cerebrovascular accident; EVD, external ventricular drain; MI, myocardial infarction; SAH, subarachnoid hemorrhage. 1 day showed that the 2 stents placed in a Y configuration through the A2 segments were occluded. One patient died as a result of rebleeding of a fusiform basilar artery aneurysm that could only be stented. One patient died of hemorrhage into a procedure-related thalamic infarct 2 months after stent-coiling. Two patients died after progressive neurological decline and withdrawal of care by family. When procedural morbidity and mortality are included for the 34 patients who did not meet inclusion criteria on the basis of imaging follow-up, 12 of 161 patients (7.5%) died. Two of 168 procedures resulted directly in the death of a patient for a procedural mortality of 1.2%, whereas the remaining 10 patients died of progression of poor neurological grade, rehemorrhage, or medical decline. Twenty-five cases (14.9%) were complicated by procedural morbidity. These included 10 cerebrovascular accidents, 6 perforations (2 requiring FIGURE 5. Patient and complication flow diagram by antiplatelet regimen. ACA, anterior cerebral artery; ASA, acetylsalicylic acid; ICA, internal carotid artery; MCA, middle cerebral artery. temporary cerebrospinal fluid diversion), 4 dissections, 3 femoral access complications, 2 myocardial infarctions, and 2 cases of bacterial meningitis. Procedural neurological morbidity was 6.0%. For the purposes of this study, acute or subacute thrombosis and chronic stenosis or occlusion are grouped together so that the primary end point of in-stent events leading to morbidity or mortality would not be underestimated. The classification of acute or subacute stent thrombosis in our patient cohort is based on the cardiac literature convention 6 of patients presenting within 30 days of stenting and an angiogram showing thrombus (Figure 6). In our series, as described below, 3 patients did not fit these criteria exactly but were still grouped with the subacute thrombosis group because of the nature of their presentation and the likely etiology of symptomatic thrombus. Chronic in-stent stenosis was defined as $ 50% luminal narrowing. 1 month after NRD-assisted coiling (Figure 7). One patient who was found to have 30% asymptomatic luminal narrowing at 6 months had complete resolution of in-stent stenosis by 37 months. Six patients (4.5% of 133 stents) had acute or subacute thrombus, and all were symptomatic (Table 1). A 68-year-old woman previously described had acute thrombus of both stents on day 1 after coiling of a ruptured Acomm aneurysm initially complicated by a perforation. A 47-year-old woman had a right middle cerebral artery territory stroke with thrombus visualized in the right middle cerebral artery 10 days after coiling of an unruptured ophthalmic artery aneurysm. This patient had a history of multiple deep vein thromboses with pulmonary embolism but no documented hypercoagulable state. A 66-year-old woman had an occlusion of her left posterior cerebral artery 30 days after coiling of a ruptured posterior cerebral artery aneurysm after stopping her antiplatelet therapy for a gastrointestinal bleed. The first of the 3 patients classified with subacute thrombus outside of the conventional definition is a 76-year-old woman who presented with a 3-mm ruptured Acomm aneurysm and required a Neuroform NRD extending from the left A1 to the right A2. On posthemorrhage day 10, she underwent placement of a ventriculoperitoneal shunt and received platelets preoperatively for aspirin and Plavix. Postoperatively, she was found to have bilateral anterior cerebral artery strokes and an angiogram showing no evidence of vasospasm or stent thrombus. Given that an acute thromboembolic event was the most likely explanation based on the pattern of stroke, she was included in the acute thrombus patient group despite the absence of angiography-proven thrombus. The second and third patients are a 62-year-old man and a 59-year-old woman who both presented at 35 days (5 days after Plavix was stopped) with an acute thrombus and stroke. One patient had a Y stent configuration with 2 stents placed proximally in the basilar artery and each diverging distally into each P2. The other patient had a stent placed within an existing stent for the retreatment of a fusiform basilar artery aneurysm. With these 3 exceptions included, 6 patients had acute stent thrombosis (4.5%), and 3 had chronic in-stent stenosis (2.3%) (Table 1); instent events occurred in 9 of 133 constructs (6.8%), and 7 were symptomatic (5.3%) VOLUME 67 NUMBER 6 DECEMBER

5 IN-STENT THROMBOSIS AND STENOSIS FIGURE 6. Acute thrombus (arrow) in the middle cerebral artery found 10 days after stent-assisted coiling of 16-mm right ophthalmic artery aneurysm. NEUROSURGERY VOLUME 67 NUMBER 6 DECEMBER

6 KANAAN ET AL FIGURE 7. Chronic stenosis/occlusion of the left posterior cerebral artery (arrow) found 2 months after stent-assisted coiling of 16-mm left superior cerebellar artery aneurysm. An analysis of the techniques used in each patient was undertaken to determine which factors may predispose to in-stent complications (Table 2). We found no increased risk of in-stent thrombosis or stenosis with microcatheter jailing, female sex, age, presentation with SAH, or stent type. There were higher rates of in-stent events with the use of the Enterprise NRD, but this was not statistically significant. There was no significant difference between standard and nonstandard constructs for all in-stent events, but subgroup analysis showed a trend toward acute thrombosis in 3 of 15 aneurysms (20%) with 2-stent constructs 1528 VOLUME 67 NUMBER 6 DECEMBER

7 IN-STENT THROMBOSIS AND STENOSIS TABLE 1. Nine Patients With In-Stent Thrombosis or Stenosis a Age, y Sex Aneurysm Presentation In-Stent Finding Jailed Stent/Configuration 68 F Ruptured 4-mm Acomm 1 d Intraoperative rupture, both stents occluded, death 47 F Unruptured 16-mm R Ophth 10 d Subacute thrombus, R MCA territory infarct 76 F Ruptured 3-mm Acomm 10 d Bilateral ACA strokes, no thrombus or spasm seen 66 F Ruptured 6-mm L PCA 1 mo (aspirin, Plavix held for GI bleed) 62 M Unruptured 9-mm basilar 35 d (5 d after Plavix stopped) 59 F Ruptured fusiform VBJ with 35 d (5 d after residual Plavix stopped) L PCA occlusion, stroke, at 7 mo recanalized L PCA stroke, in-stent thrombus No Yes Yes Yes Yes Neuroform/Y configuration L A1-A2 and R A1-A2 Enterprise/standard R ICA Neuroform/standard L A1 to R A2 Neuroform/standard basilar to L PCA Enterprise/Y configuration basilar to bilateral PCAs Bilateral PICA infarcts, Yes Neuroform/coaxial basilar basilar occlusion 53 F Unruptured 16-mm L PCA 2 mo L PCA occlusion, stroke No Neuroform/standard basilar to L PCA 52 F Unruptured 8-mm Acomm 6 mo 50% stenosis, asymptomatic No Enterprise/standard L A1- A2 67 F Unruptured 8-mm R Ophth 7 mo R ICA occlusion, asymptomatic No Neuroform/standard R ICA a ACA, anterior cerebral artery; Acomm, anterior communicating artery; GI, gastrointestinal; ICA, internal carotid artery; MCA, middle cerebral artery; Opth, ophthalmic; PCA, posterior cerebral artery; PICA, posterior inferior cerebellar artery; VBJ, vertebrobasilar junction. (Y configuration and coaxial) vs 6 of 118 (5.1%) with 1 stent (P =.07) (Table 3). Three of 12 Acomm aneurysms (25%) treated with NRDs had a higher rate (P =.005) of in-stent events compared with other anterior circulation aneurysms or posterior circulation aneurysms (Table 3). The role of pretreatment with antiplatelet therapy in the rates of in-stent events was also studied. Among 117 stent constructs performed in patients who received eptifibatide but not pretreatment, 5 (4.3%) experienced an instent event compared with 1 of 9 patients (11.1%) who received both eptifibatide and pretreatment. Only 2 patients received pretreatment without eptifibatide, and neither had an in-stent event. Finally, 3 of 5 patients (60%) who received neither pretreatment nor eptifibatide had an in-stent event (Figure 7). In summary, when the glycoprotein (GP) IIb/IIIa inhibitor eptifibatide was given at the time of stent deployment without pretreatment with aspirin or Plavix, there was no significantly higher rate of in-stent events. DISCUSSION The Stenting of Symptomatic Atherosclerotic Lesions in the Vertebral or Intracranial Arteries trial published in 2004 was the first prospective study to evaluate intracranial stenting for atherosclerotic disease 7 using the balloon-expanded Neurolink stent (Guidant Corp, Santa Clara, California). Restenosis, defined as. 50% stenosis after 6 months of follow-up, was found in 12 of 37 stents (32.4%). The 2007 Wingspan study used a self-expanding stent (Boston Scientific) after balloon angioplasty 8 in 45 patients with symptomatic intracranial stenosis. After 6 months, 3 patients (7.5%) had restenosis based on narrowing. 50%. A larger study of 127 patients found an in-stent stenosis rate of 32.3% with the Wingspan stent. 9 An interim analysis of the same patient cohort attempted to explain the lower rate in the initial Wingspan study on the basis of the finding that there was a higher rate of instent stenosis in vessels treated in the anterior circulation and a higher proportion of patients in their own study with anterior circulation treatment. 10 Although there was no difference in rates of stenosis between anterior and posterior circulation aneurysms in our study, there was a higher rate of acute or subacute thrombosis in Acomm aneurysms compared with other locations. Unlike the devices used for neck remodeling of aneurysms, Neurolink and Wingspan stent systems use balloon angioplasty. Evidence suggests that vascular damage incurred during angioplasty balloon expansion leads to a neointimal response that mediates restenosis, 11 which may be an important factor explaining our results and those seen with atherosclerotic lesions. It would stand to reason that healthy, nonatherosclerotic vessels that are not traumatized during stent deployment would have lower in-stent stenosis rates. The authors of the largest published study to date of stent-assisted coilings also reached this conclusion. 12 In the 259 patients with aneurysms treated with the Neuroform stent, 156 patients in this study had follow-up, 9 (5.8%) were found to have delayed in-stent stenosis, and 2 were symptomatic. Unlike the follow-up in our study, which extended beyond 15 months with cerebral catheter arteriography, follow-up by Fiorella et al 12 was reported only as. 2 months and entailed cerebral angiography or noninvasive studies. These authors also found a case of spontaneous resolution of asymptomatic in-stent stenosis. Biondi et al 13 published an earlier prospective study of 45 aneurysms treated with 47 Neuroform stents. Angiographic follow-up was available for 33 of 45 treated aneurysms, with a mean interval of 9 months. Only 1 patient developed significant in-stent stenosis (3% of patients NEUROSURGERY VOLUME 67 NUMBER 6 DECEMBER

8 KANAAN ET AL TABLE 2. Thrombosis/Stenosis and Demographic and Technical Variables No Significant Thrombosis/ Stenosis (n = 124) Significant Thrombosis/ Stenosis (n = 9) n % n % P Age, y, Sex F M Jailing No Yes Subarachnoid hemorrhage No Yes Stent configuration Standard Nonstandard By type Waffle cone Y configuration Coaxial Stent type Enterprise Neuroform Wingspan with imaging follow-up) that was asymptomatic. Our in-stent thrombosis/stenosis rate of 6.8% is comparable to the rate in these 2 studies. The higher rate of thrombosis with 2-stent constructs and in 2 patients who recently stopped Plavix may suggest the need for longer Plavix therapy beyond 30 days. The higher rates of thrombosis in patients with constructs used to treat Acomm aneurysms may be a result of smaller artery diameter or complexity of aneurysm and parent vessel anatomy. Here, we analyzed acute or subacute thrombotic events and chronic stenosis. By combining these events, we would not underestimate the complications associated with the use of NRDs compared with considering each entity separately. Fiorella et al 12 and Biondi et al 13 reported no acute thrombotic events in their respective cohorts. Only the latter group routinely used the GP IIb/IIIa inhibitor abciximab (Lilly, Indianapolis, Indiana), and only in elective coilings in which the patients were not pretreated with dual antiplatelet therapy. 12 As mentioned previously, Fiorella et al 12 report pretreatment in the 9 patients who presented with stenosis but do not describe the regimen used in all patients. Biondi et al 13 report pretreatment with Plavix or Ticlid (Roche, Basel, Switzerland) in all elective cases. With the results of these 2 groups and those of our own study, it would be TABLE 3. Thrombosis or Stenosis Based on Location and Number of stents Within a Construct a No In-Stent In-Stent Event (n = 124) Event (n = 9) P n % n % Location Anterior, Acomm Anterior, non-acomm Posterior No. of stents a Acomm, anterior communicating artery. reasonable to conclude that the incidence of acute in-stent thrombosis is low and that oral pretreatment with antiplatelets or loading with aspirin, Plavix, and eptifibatide at the time of intervention is sufficient for prophylaxis. In contradiction to these conclusions, in a recently published study of a mixed group of 67 patients treated with stents for both atherosclerosis and intracerebral aneurysms, 7 patients (10.1%) had subacute thrombosis of their stent. 14 Three of 34 patients (8.8%) with aneurysms had acute thrombotic events. Angiograms were performed only in symptomatic patients, and the follow-up was 1 to 13 days. Patients did not receive GP IIb/IIIa inhibitors, but all patients were pretreated with dual antiplatelet therapy. The clinical importance of the use of GP IIb/IIIa inhibitors remains undetermined and needs further examination. Although antiplatelet agents are generally used with stents in cardiac or neurological applications, the cardiac literature has some debate on the effectiveness of GP IIb/IIIA inhibitors; current consensus, however, is that there is a lower rate of acute or subacute stent occlusion associated with their use. 15,16 With the exception of patients who experienced a known perforation, the majority of our procedures were performed with eptifibatide (126 of 133) in addition to oral antiplatelet agents either as part of a pretreatment protocol or at the time of intervention. There was no higher rate of in-stent events in patients who received antiplatelet periprocedurally compared with pretreatment dosing in our patients, and this may be explained by our standardized use of a GP IIb/IIIa inhibitor. In regard to the safety of using stents and associated anticoagulation with SAH, the literature provides no study that clearly contraindicates their use. This question was addressed in a recent retrospective study of 61 patients who underwent stentassisted coiling specifically in the context of SAH. 17 The mortality for this cohort was 12 of 61 patients (19.7%). In comparison, we report 11 deaths in 64 patients (17.2%) with SAH. These mortality rates are consistent with rates reported in the literature for both coiling and clipping of ruptured aneurysms. 3 Intuitively, however, care can become complicated in anticoagulated patients who have an intraprocedural perforation 1530 VOLUME 67 NUMBER 6 DECEMBER

9 IN-STENT THROMBOSIS AND STENOSIS or who may require cerebrospinal fluid diversion procedures or any other intervention during a long intensive care unit stay. We recommend administering antiplatelet agents only after NRD and coils are in place and angiographic imaging shows no evidence of contrast extravasation. For any patient on aspirin and Plavix who requires an invasive procedure during hospitalization, we transfuse platelets at the onset of the procedure. Because Plavix and/or aspirin insensitivity may underlie thrombotic or stenotic complications, 14 we have begun within the last year to run aspirin and Plavix assays on all patients receiving stents to determine whether there is any utility in these tests. We perform Plavix assays on all patients the morning after their procedure to avoid false-negative results from the latent effects of eptifibatide that might affect the interpretation (VerifyNow, San Diego, California). Patients who have an elective coiling are discharged the next day. If the Plavix assay reports insensitivity, they are instructed to double their Plavix dose. 18 For SAH patients with lengthy hospital stays, we rerun the Plavix assay after the patient has been reloaded with Plavix and started on a higher dose. If the patient remains Plavix insensitive, then we change to ticlopidine, which is effective in 96% of patients who are Plavix insensitive. 19 So far, 4 patients with acute thrombosis had Plavix assays performed, and 2 were classified as Plavix responders and 2 were not. The doses in the patients who did not respond to Plavix were doubled on discharge, and as described above, the thrombotic events occurred only after Plavix was stopped at 30 days. The clinical significance of these assays remains to be determined. CONCLUSIONS One hundred twenty-seven patients with 133 NRDs were followed up for a mean of 15.4 months (median, 12.7 months); this represents the largest study with the longest standardized follow-up with patients imaged by catheter-based arteriography. Acute in-stent thrombosis or chronic stenosis was found in 9 of 133 stent-coil constructs (6.8%), which is comparable to rates in other studies available in the literature. Six constructs (4.5%) had an acute or subacute thrombotic event, and 3 constructs (2.3%) had chronic stenosis. Symptomatic in-stent events were seen in 7 of 133 constructs (5.3%). Including patients not eligible for follow-up, 2 of 168 procedures (1.2%) resulted in the death of a patient, and procedural neurological morbidity was 6.0%. Jailing coil-bearing microcatheters between the stent and vessel lumen is not associated with a higher rate of in-stent stenosis. Two-stent constructs are associated with a strong trend toward instent events (P =.07), particularly if Plavix is stopped after 30 days. Acomm aneurysms treated with NRDs have a statistically significantly higher rate (25%) of in-stent thrombosis or stenosis than other anterior circulation aneurysms (2.3%) or posterior circulation aneurysms (12%). Antiplatelet pretreatment was used in 11 of 82 elective cases, and there was no higher rate of in-stent events in patients who did not receive pretreatment but did receive the GP IIb/IIIa inhibitor eptifibatide. This suggests that the use of eptifibatide, aspirin, and Plavix load at the time of intervention is no less effective than pretreatment with antiplatelet agents. The importance of aspirin and Plavix sensitivity assays is undetermined, but further prospective study should include their use in determining the factors involved in the complications seen with the use of NRDs. Disclosure Dr Horowitz has been a consultant for EV3. Dr Jovin has been a consultant for EV3, Coaxia, and Concentric Medical. The authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article. REFERENCES 1. Hopkins LN, Ecker RD. Cerebral endovascular neurosurgery. Neurosurgery. 2008;62(Suppl 3): Fraser JF, Riina H, Mitra D, Gobin YP, Simon AS, Steig PE. Treatment of ruptured intracranial aneurysms: looking to the past to the register the future. Neurosurgery. 2006;59(6): Molyneux JA, Kerr RS, Yu L-M, et al; International Subarachnoid Aneurysm Trial (ISAT) Collaborative Group. International Subarachnoid Aneurysm Trial (ISAT) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: a randomized comparison of effects on survival, dependency, seizures, rebleeding, subgroups, and aneurysm occlusion. Lancet. 2005;366(9488): Kim SR, Vora N, Jovin TG, et al. Anatomic results and complications of stentassisted coil embolization of intracranial aneurysms. Intervent Neuroradiol. 2008;14: Horowitz M, Levy E, Sauvageau E, et al. Intra/extra-aneurysmal stent placement for management of complex and wide-necked-bifurcation aneurysms: eight cases using the waffle cone technique. Neurosurgery. 2006;58(4)(Suppl 2): Windecker S, Meier B. Late coronary stent thrombosis. Circulation. 2007;116(17): SSYLVIA Study Investigators. Stenting of Symptomatic Atherosclerotic Lesions in the Vertebral or Intracranial Arteries (SSYLVIA). Stroke. 2004;35(6): Bose A, Hartmann M, Henkes H, et al. A novel, self-expanding, nitinol stent in medically refractory intracranial atherosclerotic stenoses: the Wingspan Study. Stroke. 2007;38(5): Albuquerque FC, Levy EI, Turk AS, et al. Angiographic patterns of Wingspan instent restenosis. Neurosurgery. 2008;63(1): Levy EI, Turk AS, Albuquerque FC, et al. Wingspan in-stent restenosis and thrombosis: incidence, clinical presentation, and management. Neurosurgery. 2007;61(3): Benett MR, O Sullivan M. Mechanisms of angioplasty and stent restenosis: implications for design of rational therapy. Pharmacol Ther. 2001;91(2): Fiorella D, Albuquerque FC, Woo H, Rasmussen PA, Masaryk TJ, McDougall CG. Neuroform in-stent stenosis: incidence, natural history, and treatment strategies. Neurosurgery. 2006;59(1): Biondi A, Janardhan V, Katz JM, Salvaggio K, Riina HA, Gobin YP. Neuroform stent-assisted coil embolization of wide-neck intracranial aneurysms: strategies in stent deployment and midterm follow-up. Neurosurgery. 2007;61(3): Reidel CH, Tietke M, Alfke K, Stingele R, Jansen O. Subacute stent thrombosis in intracranial stenting. Stroke. 2009;40(4): Cohen M, Hoekstra J, Giugliano R, et al. Treatment strategies in non-stelevation acute coronary syndromes in patients undergoing percutaneous coronary intervention: an evidence-based review of clinical trial results and treatment guidelines: report on a roundtable discussion. JInterventCardiol. 2008;21(4): De Luca G, Suryapranata H, Stone GW, et al. Abciximab as adjunctive therapy to reperfusion in acute ST-segment elevation myocardial infarction: a meta-analysis of randomized trials. JAMA. 2005;293(14): Tahiniten OI, Vanninen RL, Manninen HI, et al. Wide-necked intracranial aneurysms: treatment with stent-assisted coil embolization during acute (, 72 hours) subarachnoid hemorrhage: experience in 61 consecutive patients. Neuroradiology. 2009;253(1): Beckrath N, Kastrati A, Wieczorek A, et al. A double blind, randomized study on platelet aggregation in patients treated with a daily dose of 150 or 75 mg of clopidogrel for 30 days. Eur Soc Cardiol. 2007;28(15): NEUROSURGERY VOLUME 67 NUMBER 6 DECEMBER

10 KANAAN ET AL 19. Campo G, Valgimigli M, Gemmati D, et al. Poor responsiveness to clopidogrel: drug-specific or class-effect mechanism? Evidence from a clopidogrel-to-ticlopidine crossover study. J Am Coll Cardiol. 2007;50(12): COMMENTS This is a retrospective review of 127 patients with intracranial aneurysms treated with stent-assisted coiling. The authors review their experience with in-stent thrombosis and stenosis for stent-assisted coiling of ruptured and unruptured aneurysms. They also discuss the use of antiplatelet agents for stents and possible pitfalls. As the authors note, this currently represents the largest study with the longest standardized follow-up with patients imaged by catheter-based angiography. Although the debate continues in the cardiac literature regarding the use of glycoprotein IIb/IIIa inhibitors, there does seem to be a consensus regarding their efficacy for lowering the rate of acute or subacute stent occlusion. However, there is currently no consensus in the neurointerventional literature regarding the use of glycoprotein IIb/IIIa inhibitors for stent placement during stent-assisted coil embolization of intracranial aneurysms. The authors note that they have recently adopted the practice of performing aspirin and Plavix assays on all patients receiving stents to determine the utility of these tests. The validity of this approach has not yet been proven in the neurointerventional literature, although as further data emerge, this may happen. We agree with the authors that further prospective studies should be performed to evaluate the utility of these assays in identifying risk factors for in-stent thrombosis and stenosis. The authors note that they double the dose of Plavix in patients who are classified as Plavix nonresponders. However, two of these patients had thrombotic events when Plavix was stopped at 30 days. It may be advisable in this subgroup of patients not only to double the dose of Plavix as the authors have done but also to maintain them on a lower dose of Plavix for an additional period of time. The authors also note that they observed a significantly higher rate of in-stent thrombosis when using 2-stent constructs. Given this finding, further studies should be conducted to determine whether increasing the antiplatelet therapy by increasing the dose of Plavix for patients whose aneurysms require the use of 2-stent constructs would decrease the rate of in-stent thrombosis for these patients. Advances in stent materials and possibly drug-eluting properties may tackle this issue in the future. The authors also note that the technique of jailing the microcatheter between the stent and the vessel lumen does not appear to increase the incidence of in-stent thrombosis. This is an important technical point because this technique has become increasingly popular for stent-assisted coiling of aneurysms. It is also interesting to note that no higher rates of in-stent thrombosis or stenosis were noted in patients who were loaded with antiplatelet agents at the time of the procedure compared with those who were pretreated with antiplatelet agents. The question of using stents in the setting of acute subarachnoid hemorrhage remains controversial. Our preference is to use balloon remodeling rather than stents in this setting and potentially complete treatment with a stent several weeks or months later. Further data ensuring the safety of using stents in the acute subarachnoid hemorrhage period is needed before wide adoption is likely to occur. Further refinements in stent technology and pharmacology are needed to further decrease delayed complications of intracranial stenting. Anitha Nimmagadda Bernard R. Bendok Chicago, Illinois The authors report the morbidity associated with neck-remodeling devices used as adjuncts for coil embolization of cerebral aneurysms in 127 patients. Acute or subacute thrombosis occurred in 4.5% of stentcoil constructs, and another 2.5% had chronic in-stent stenosis. Although these figures are comparable to rates reported in previous publications, one of the advantages of this series is the consistency of the follow-up protocol, with all patients receiving at least 1 catheter-based angiogram after the procedure. Thus, the sensitivity for detecting such events is likely to be high. There were no statistically different rates of complications associated with the use of NeuroForm (Boston Scientific, Natick, Massachusetts) vs Enterprise (Cordis, Warren, New Jersey) stents, anterior vs posterior circulation, or jailing vs crossing techniques of microcatheter placement. However, 2-stent constructs (eg, Y stent or overlapping stent configurations) were associated with higher complication rates, as was the treatment of anterior communicating artery aneurysms. Almost all the patients in this series received a dose of the glycoprotein IIb/IIIa inhibitor eptifibatide (Integrilin, Shering, Kenilworth, New Jersey) once the stent was deployed. No hemorrhages were detected as a result of this measure. Only a minority were pretreated with aspirin and Plavix, yet there were no higher rates of in-stent events among the patients who did not receive pretreatment. Collectively, these results reaffirm the role of stent-assisted coil embolization in carefully selected patients. Arun Paul Amar Los Angeles, California The authors present a retrospective review of a series of 161 patients who underwent coil embolization of 168 ruptured and unruptured aneurysms assisted by 1 or more neck-remodeling devices. The authors present several nuances, including nonstandard stent configurations and their technique for jailing the microcatheter. Catheter-based angiographic follow-up was available in 127 patients (133 stent-coil constructs) at a mean of 15.4 months. There were 9 in-stent events (6 acute or subacute thromboses and 3 delayed stenoses or occlusions). A significantly higher rate of in-stent events was seen with the use of constructs to treat anterior communicating artery aneurysms. Considering all patients (including those without angiographic follow-up), procedural mortality occurred in 2 patients (1.2%), and procedural morbidity was 14.9%. The authors have honestly presented their results using 1 or more neck-remodeling devices for the treatment of cerebral aneurysms. This is a large study from a high-volume center with significant experience and expertise. It is noteworthy that of the 127 patients with angiographic follow-up, 45 patients (35%) presented with subarachnoid hemorrhage. This demonstrates this group s aggressive endovascular approach toward patients with ruptured aneurysms not amenable to simple coil embolization. Other groups across the United States have cautioned against such an approach, 1,2 and the present study provides provocative data concerning the controversy surrounding the use of stents for acutely ruptured cerebral aneurysms. The complication rates reported in this study are not insignificant. Information such as this is useful in counseling patients and their families. It should be emphasized that the authors have treated what is likely an extremely challenging group of patients harboring difficult aneurysms. Detailed analysis of local results is an important step in 1532 VOLUME 67 NUMBER 6 DECEMBER

11 IN-STENT THROMBOSIS AND STENOSIS formulating an optimal plan for a given patient at a single institution. Microsurgery would be an option in many patients with challenging aneurysms for endovascular therapy, and whether microsurgery could be applied with a lower complication rate would depend on the expertise of the local treating team. The authors have provided a significant contribution to the existing literature. Reports such as this will help guide future therapy and research for patients harboring cerebral aneurysms. Aaron S. Dumont Philadelphia, Pennsylvania 1. Tumialan LN, Zhang YJ, Cawley CM, et al. Intracranial hemorrhage associated with stent-assisted coil embolization of cerebral aneurysms: a cautionary report. J Neurosurg. 2008;108(6): Dumont AS, Evans AJ, Jensen ME. Hemorrhagic complications associated with stent-assisted coil embolization. J Neurosurg. 2008;108(6): NEUROSURGERY VOLUME 67 NUMBER 6 DECEMBER