Periprocedural thromboembolic complications from endovascular

Transcription

1 Rescue Treatment of Thromboembolic Complications During Endovascular Treatment of Cerebral Aneurysms Waleed Brinjikji, MD; Jennifer S. McDonald, PhD; David F. Kallmes, MD; Harry J. Cloft, MD, PhD Background and Purpose Acute intraprocedural thrombus formation complicating endovascular cerebral aneurysm treatment is often treated with intra-arterial or intravenous administration of thrombolytic agents or glycoprotein IIb/IIIa (GpIIb/IIIa) inhibitors. We sought to evaluate the morbidity and mortality associated with such treatments using a large multihospital database. Methods Using the Premier Perspective Database, we examined outcomes for patients receiving endovascular coiling for ruptured and unruptured aneurysms requiring rescue therapy, defined as treatment with GpIIb/IIIa inhibitors and fibrinolytic therapy. We compared discharge status, length of stay, and complication rates across 3 groups: (1) patients receiving GpIIb/IIIa inhibitors only, (2) patients receiving fibrinolytic therapy only, and (3) patients receiving both GpIIb/ IIIa inhibitors and fibrinolytics. Student t test was used to compare continuous variables, and Fisher exact test was used to compare categorical variables. Results Seven-percent (254/3627) of patients treated for unruptured aneurysms received rescue therapy. When compared with patients receiving GpIIb/IIIa inhibitors alone, patients receiving only fibrinolytics had significantly higher rates of discharge to institutions other than home (37.5% [9/24] versus 7.4% [15/201]; P<0.0001). Eight-percent of patients (338/4204) treated for ruptured aneurysms received rescue therapy. When compared with patients receiving GpIIb/IIIa inhibitors alone, patients receiving only fibrinolytics had significantly higher rates of mortality (26.0% [18/69] versus 14.5% [35/241]; P=0.02) and discharge to institutions other than home (59.4% [41/69] versus 36.5% [88/241]; P<0.0001). Conclusions Pharmacological rescue therapy occurred in 7% to 8% of endovascular coiling patients with unruptured and ruptured intracranial aneurysms. Rescue therapy with thrombolytic agents resulted in significantly more morbidity and mortality than rescue therapy with GpIIb/IIIa inhibitors. (Stroke. 2013;44: ) Key Words: endovascular treatment interventional neuroradiology outcomes subarachnoid hemorrhage Periprocedural thromboembolic complications from endovascular treatment of intracranial aneurysms are estimated to occur in 2% to 15% of patients. 1 Aggressive treatment of acute intraprocedural thrombus formation with intra-arterial or intravenous administration of fibrinolytics or glycoprotein IIb/IIIa (GpIIb/IIIa) inhibitors has become a standard treatment for these complications. Recanalization of the treated artery is usually essential to avoid permanent neurological deficit. Many studies have demonstrated benefit from the intraprocedural administration of GpIIb/IIIa inhibitors and thrombolytic agents during endovascular treatment of intracranial aneurysms. However, the real-world outcomes of patients requiring such treatments are not well known. Using a large multihospital database, we sought to determine the periprocedural morbidity and mortality of patients requiring rescue therapy (ie, intraprocedural administration of GpIIb/ IIIa inhibitors or thrombolytic agents) during endovascular treatment of intracranial aneurysms. Methods Study Population and Data Retrieval The Premier Perspective database is a voluntary, fee-supported collection of data developed by Premier, Inc. to assess quality and resource use. As of 2011, the Perspective database consisted of 15% of hospitalizations nationwide and represented >600 US hospitals. Detailed information of a patient s hospitalization, including patient demographics, hospital information, diagnoses, procedures, discharge status, and all billed items are recorded. Patients who presented with a ruptured aneurysm (International Classification of Diseases-Ninth Revision-Clinical Modification [ICD-9-CM] diagnostic code 430) or unruptured aneurysm (ICD-9-CM code without concurrent diagnostic code 430) from November 2005 through December 2011 were identified. Patients were included if they underwent aneurysmal coiling (ICD-9 procedural codes [other repair of aneurysm], [endovascular repair of occlusion of head and neck vessels], [endovascular embolization or occlusion of vessel(s) of head or neck using bare coils], [endovascular embolization or occlusion of vessel(s) of head or neck using bioactive coils], or [other endovascular repair (of aneurysm) of other vessels]) during hospitalization. Patients who received a stent during the coiling procedure were similarly identified using billing information. Received December 27, 2012; accepted February 20, From the Department of Radiology (W.B., J.S.McD., D.F.K., H.J.C.) and Department of Neurosurgery (D.F.K., H.J.C.), Mayo Clinic, Rochester, MN. Correspondence to Waleed Brinjikji, MD, Mayo Clinic Department of Radiology, 200 SW First St, Rochester, MN brinjikji.waleed@mayo.edu 2013 American Heart Association, Inc. Stroke is available at DOI: /STROKEAHA

2 1344 Stroke May 2013 Table 1. Demographics of Patients With Unruptured Aneurysm GpIIb /IIIa Inhibitors Only Fibrinolytic Only P Value Both P Value N (%) 202 (5.6) 24 (0.6) 28 (0.7) Mean (SD) age 56.0 (12.9) 54.5 (14.8) (10.3) 0.11 Female, N (%) 162 (80.2) 18 (75.0) (85.7) 0.49 Urban hospital, N (%) 201 (99.5) 23 (95.8) (96.4) 0.11 Teaching hospital, N (%) 115 (56.9) 19 (79.2) (71.4) 0.14 Stenting, N (%) 94 (46.5) 6 (25.0) (53.6) 0.49 Patients with ruptured and unruptured aneurysm who received rescue therapy were identified using billing records. GpIIb/IIIa therapy or fibrinolytic recue therapy was included only if they were administered on the day of or the day after the coiling procedure. Antiplatelet rescue therapy was defined as administration of Abciximab (ReoPro), Eptifibatide (Integrillin), or Tirofiban (Aggrastat). Fibrinolytic rescue therapy was defined as administration of Alteplase (Activase), Reteplase (Retavase), Tenecteplase (Tnkase), or Urokinase (Abbokinase). We compared demographics and outcomes from patients with unruptured aneurysm and patients with subarachnoid hemorrhage separately. For each patient population, we compared outcomes across 3 groups: (1) patients receiving GpIIb/IIIa inhibitors only, (2) patients receiving fibrinolytic therapy only, and (3) patients receiving both GpIIb/IIIa inhibitors and fibrinolytics. The following demographic variables were compared: (1) age, (2) sex, (3) hospital location (urban versus rural), (4) hospital teaching status, and (5) stent-assisted coiling rates. The following outcomes were compared: (1) discharge status (mortality, discharge to institution other than home, discharge to home/home care); (2) length of stay; and rates of (3) aphasia, (4) hemiparesis, (5) postoperative neurological complications, and (6) ventriculostomy. Discharge to institutions other than home was used as an indicator of morbidity. Hemorrhage rates were compared for the unruptured aneurysm group only. Statistics Data were extracted from the Perspective database using SAS (SAS, version 9.3; SAS Institute, Cary, NC) and analyzed using JMP (version 9, SAS Institute, Cary, NC). Continuous results are presented as mean (SD). Categorical results are presented as percentage. Comparison of continuous results was performed using the Student t test. Comparison of categorical results was performed using Fisher exact test. The group of patients receiving GpIIb/IIIa inhibitors only was the reference for statistical analysis. Table 2. Outcomes of Patients With Unruptured Aneurysm Results Unruptured Aneurysm Group A total of 3627 patients in the Premier Perspective database received endovascular therapy for treatment of unruptured aneurysm. Of which, 254 patients (7.0%) received rescue therapy with GpIIb/IIIa inhibitor and thrombolytic agents; 202 patients (5.6%) received GpIIb/ IIIa inhibitor rescue therapy; 24 patients (0.7%) received fibrinolytic therapy; and 28 patients (0.8%) received both fibrinolytic and GpIIb/IIIa inhibitor rescue therapy. The demographic characteristics of these patients are summarized in Table 1. Patients receiving GpIIbIIIa inhibitors had significantly better outcomes compared with the fibrinolytic group. GpIIbIIIa inhibitor patients had significantly higher rates of discharge to home/home care compared with patients receiving fibrinolytic therapy (90.6%, 183/202 versus 58.3%, 14/24, respectively; P<0.0001). There was a trend for GpIIbIIIa inhibitor patients to have higher rates of discharge to home/home care facilities compared with the combination therapy group (90.6%, 183/202 versus 78.6%, 22/28, respectively; P=0.06). GpIIbIIIa inhibitor patients had decreased rates of aphasia compared with the fibrinolytic therapy group (1.5%, 3/202 versus 8.3%, 2/24, respectively; P=0.03). Compared with patients receiving combination therapy, patients receiving GpIIbIIIa inhibitors had decreased rates of hemorrhage (2.0%, 4/202 versus 10.7%, 3/28, respectively; P=0.01). These data are summarized in Table 2. Discharge status Died 3 (1.5) 1 (4.2) (0.0) 0.52 Discharge home/home care 183 (90.6) 14 (58.3) < (78.6) 0.06 Discharge to institution other than home 15 (7.4) 9 (37.5) < (21.4) 0.02 Hemorrhage 4 (2.0) 2 (8.3) (10.7) 0.01 Mean (SD) length of stay in days 3.6 (5.0) 5.2 (6.4) (10.9) 0.11 Aphasia 3 (1.5) 2 (8.3) (7.1) 0.05 Hemiparesis 11 (5.5) 3 (12.5) (3.6) 0.68 Postoperative neurological complications 13 (6.4) 3 (12.5) (17.9) 0.04 Ventriculostomy 1 (0.5) 1 (4.2) (3.6) 0.10

3 Brinjikji et al Rescue Therapy in Endovascular Aneurysm Treatment 1345 Table 3. Demographics of Patients With Subarachnoid Hemorrhage N (%) 241 (5.7) 69 (1.6) 28 (0.7) Mean (SD) age 54.2 (11.5) 56.7 (11.8) (13.8) 0.97 Female, N (%) 178 (73.9) 45 (65.2) (57.1) 0.06 Urban hospital, N (%) 240 (99.6) 68 (98.6) (96.4) 0.07 Teaching hospital, N (%) 121 (50.2) 26 (37.7) (42.9) 0.46 Stenting, N (%) 7 (2.9) 0 (0.0) (0.0) 0.36 Ruptured Aneurysm Group A total of 4204 patients received endovascular therapy for treatment of ruptured intracranial aneurysm. Of which, 338 patients (8.0%) received GpIIb/IIIa inhibitor and thrombolytic rescue therapy; 241 patients (5.7%) received GpIIb/IIIa inhibitor therapy only; 69 patients (1.6%) received fibrinolytic therapy only; and 28 patients (0.7%) received a combination of GpIIb/IIIa inhibitor and fibrinolytic therapy. The demographic characteristics of these patients are summarized in Table 3. Patients receiving GpIIb/IIIa inhibitors had significantly better outcomes compared with the fibrinolytic group. GpIIb/ IIIa inhibitor patients had significantly higher rates of discharge to home/home care (49.0%, 118/241 versus 14.5%, 10/69, respectively; P<0.0001) and significantly lower rates of discharge to institutions other than home (36.5%, 88/241 versus 59.4%, 41/69, respectively; P<0.001) and mortality (14.5%, 35/241 versus 26.0%, 18/69, respectively; P=0.03). GpIIb/IIIa inhibitor patients had significantly higher rates of discharge to home/home care compared with the combination therapy group (49.0%, 118/241 versus 25.0%, 7/28, respectively; P=0.02). They also had significantly lower rates of aphasia (8.3%, 20/241 versus 25.0%, 7/28, respectively; P=0.005). These data are summarized in Table 4. Discussion Using a large administrative database, we demonstrated that 7% to 8% of patients who underwent endovascular treatment of unruptured or ruptured intracranial aneurysms received rescue therapy with intraprocedural administration of either GpIIb/IIIa inhibitor agents or thrombolytic agents. GpIIb/IIIa inhibitors were used at much higher rates than thrombolytics. Table 4. Outcomes of Patients With Subarachnoid Hemorrhage Notably, outcomes for patients with both unruptured and ruptured aneurysm, on the basis of numerous outcomes, including discharge status, aphasia, and other complications, were significantly better for patients receiving GpIIb/IIIa inhibitors than those treated with fibrinolytic agents. These data should prompt practitioners to avoid fibrinolytics in place of GpIIb/ IIIa agents for rescue therapy during coil embolization of both ruptured and unruptured aneurysms. In most case series, the rate of rescue therapy for thromboembolic complications ranges between 5% and 10% The infarction rate among patients receiving rescue therapy ranges between 10% and 40% with higher rates generally seen in studies with postoperative MR imaging 1 14 (Table 5). In a series of 477 patients with 515 intracranial aneurysms, Ries et al 1 reported that 48 patients (10%) had thromboembolic events; 42 of these patients received rescue therapy with GpIIbIIIa inhibitors, and 1 patient received rescue therapy with recombinant tissue plasminogen activator. Of the patients who did not receive rescue therapy, 3/5 had infarcts, whereas 31% of patients receiving GpIIbIIIa inhibitor rescue therapy had infarction on computed tomography. Linfante et al 10 demonstrated that 10% of patients undergoing endovascular embolization of intracranial aneurysms required intra-arterial GpIIbIIIa inhibitor rescue therapy with an infarct and hemorrhage rate of 0%. Many previous studies that have demonstrated high-infarct rates on postoperative imaging also highlight the fact that many of these infarcts are clinically silent. 1,2,9 We cannot evaluate infarct rates or success of recanalization with thrombolytic agents versus GpIIb/IIIa inhibitors in this study, but better outcomes suggest that GpIIb/IIIa inhibitors are a better option for rescue therapy. 15,16 Discharge status Died 35 (14.5) 18 (26.0) (25.0) 0.15 Discharge home/home care 118 (49.0) 10 (14.5) < (25.0) 0.02 Discharge to institution other than home 88 (36.5) 41 (59.4) (50.0) 0.16 Mean (SD) length of stay in days 15.4 (13.9) 22.9 (15.9) (8.8) 0.58 Aphasia 20 (8.3) 10 (14.5) (25.0) 0.01 Hemiparesis 43 (17.8) 15 (21.7) (17.9) 0.99 Postoperative neurological complications 23 (9.5) 3 (4.4) (21.4) 0.05 Ventriculostomy 74 (30.7) 44 (63.8) < (35.7) 0.59

4 1346 Stroke May 2013 Table 5. Literature Review Authors Rupture Status Rescue Therapy, N (%) Type of Rescue Therapy Complete or Partial Recanalization Rate, N (%) Mortality Rate, N (%) Postoperative Hemorrhagic Complications, N (%) Infarction on Postoperative Imaging, N (%) Imaging Modality Gralla et al 7 R/U 41 (3.3) IV/IA Abciximab 36 (87.8) 1 (2.4) 1 (2.4) NR NR Jones et al 8 R/U 38 (6.2) IV/IA Abciximab 24 (63.2) 0 (0.0) 0 (0.0) 8 (21.1) CT Aviv et al 3 R 13 (NR) IV Abciximab 13 (100.0) 1 (7.7) 1 (7.7) 3 (23.1) CT Ries et al 1 R/U 42 (8.8) IV/IA Abciximab 31 (73.8) 3 (7.1) 0 (0.0) 13 (31.0) CT Aggour et al 2 R/U 23 (5.9) IV/IA Abciximab 17 (73.9) 3 (13.0) 0 (0.0) 11 (47.8) MRI Brooks et al 4 R/U 10 (6.5) IA Abciximab NR NR NR 4 (40.0) MRI Linfante et al 10 R/U 19 (10.3) IA Abciximab 19 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) CT Park et al 12 R/U 32 (5.3) IA Abciximab 32 (100.0) 2 (6.3) 3 (9.4) 4 (12.5) MRI Mounayer et al 11 R/U 13 (5.7) IA Abciximab 13 (100.0) 1 (7.7) 0 (0.0) 1 (7.7) CT Kang et al 9 R/U 25 (9.6) IA Tirofiban 24 (96.0) 0 (0.0) 0 (0.0) 8 (33.3) MRI Bruening et al 5 R 16 (NR) IA Tirofiban 15 (93.8) 2 (12.5) 3 (18.8) NR NR Song et al 13 R/U 7 (7.0) IV/IA Abciximab 6 (85.7) 1 (14.3) 2 (28.6) 3 (42.9) CT Velat et al 14 R/U 26 (NR) IV/IA Abciximab 21 (80.8) 2 (7.7) 3 (11.5) NR NR and r-tpa Fiorella et al 6 R/U 10 (NR) IV/IA Abciximab 10 (100.0) 0 (0.0) 0 (0.0) 3 (30.0) NR and r-tpa Cronqvist et al 17 R/U 19 (NR) IA Urokinase 19 (100.0) 2 (10.5) 3 (15.8) 3 (15.8) CT Hähnel et al 15 R/U 7 (NR) IA r-tpa 6 (85.7) 2 (28.6) 0 (0.0) 7 (100.0) NR Cognard et al 16 R/U 12 (4.8) IA Urokinase NR 1 (8.3) 1 (8.3) NR NR CT indicates computed tomography; IA, intra-arterial; IV, intravenous; NR, not reported; R, ruptured aneurysms; and U, unruptured aneurysms; and r-tpa, recombinant tissue plasminogen activator. The use of GpIIb/IIIa inhibitors in patients with ruptured aneurysms has been reported to be relatively safe with most series reporting intracranial hemorrhage rates of <10%. 1 3,7 10,12 Aviv et al 3 reported 13 patients with ruptured aneurysm treated with abciximab rescue therapy and only 1 hemorrhagic complication. Aggour et al 2 included 11 patients with ruptured aneurysm in their abciximab rescue series without hemorrhagic complication. Park et al 12 reported hemorrhagic complications in 3 of 16 patients with ruptured cerebral aneurysms treated with abciximab rescue therapy. On the other hand, the literature has shown that rescue therapy with thrombolytic agents in patients with ruptured aneurysms is associated with a very high mortality. In a series reported by Cronqvist et al 17 of 19 patients undergoing rescue therapy with urokinase, 3 had intracranial bleeding complications. In the International Subarachnoid Aneurysm Trial (ISAT), 5 patients had thromboembolic complications treated with thrombolytic agents, and all 5 died from rebleeding of their ruptured aneurysms. 18 Catastrophic hemorrhage has also been reported with rescue using thrombolytic agents in setting of an unruptured aneurysm. 19 There are few reports on rescue therapy with thrombolytic agents, perhaps because their use has been largely replaced by GpIIbIIIa inhibitors. Despite these past reports, our study shows that thrombolytic agents continue to be used. Our data confirm the higher risk of morbidity and mortality of thrombolytic agents relative to GpIIbIIIa inhibitors, and might help to discourage continued use of thrombolytic agents for rescue therapy, especially in the setting of subarachnoid hemorrhage. Limitations Our study has limitations. Using this database, we are unable to determine whether some patients had intraprocedural thrombus formation and did not receive rescue therapy. Thus, we are unable to compare outcomes between treatment and nontreatment groups. Furthermore, we are unable to determine whether patients with subarachnoid hemorrhage had further bleeding as a result of rescue therapy. Thus, the bleeding rate can only be obtained for patients presenting with unruptured aneurysm. This database does not offer any data on presenting condition and aneurysm geometric features (aspect and dometo-neck ratio), size, or location. Thus, we are unable to correlate our outcomes with the anatomic features of the treated aneurysm. We are unable to assess whether the rescue agent was given by intra-arterial or intravenous route. Some patients may have received GpIIbIIIa inhibitors for prophylaxis of thromboembolism, but this should be relatively uncommon, as it is not standard practice and would be contraindicated with subarachnoid hemorrhage. Coding inaccuracies can affect the accuracy of any administrative database, including the Premier Perspective database. Conclusions Using a large administrative database, we demonstrated an intraprocedural rescue therapy rate of 7% to 8% for endovascular coiling patients with unruptured and ruptured intracranial aneurysms. Rescue therapy with thrombolytic agents resulted in significantly more morbidity and mortality than rescue therapy with GpIIbIIIa inhibitors.

5 Brinjikji et al Rescue Therapy in Endovascular Aneurysm Treatment 1347 Disclosures Dr Cloft is the Site PI at enrolling site for Stenting and Angioplasty with Protection in Patients and HIgh Risk for Endarterectomy registry sponsored by Cordis Endovascular. Dr Kallmes has received grant and ev3-funding for clinical trials and preclinical research, and has received grants and has grants pending from Penumbra, MicroVention, Micrus, Cordis. The other authors have no conflicts to report. References 1. Ries T, Siemonsen S, Grzyska U, Zeumer H, Fiehler J. Abciximab is a safe rescue therapy in thromboembolic events complicating cerebral aneurysm coil embolization: single center experience in 42 cases and review of the literature. Stroke. 2009;40: Aggour M, Pierot L, Kadziolka K, Gomis P, Graftieaux JP. Abciximab treatment modalities for thromboembolic events related to aneurysm coiling. Neurosurgery. 2010;67(2 suppl operative): Aviv RI, O Neill R, Patel MC, Colquhoun IR. Abciximab in patients with ruptured intracranial aneurysms. AJNR Am J Neuroradiol. 2005;26: Brooks NP, Turk AS, Niemann DB, Aagaard-Kienitz B, Pulfer K, Cook T. Frequency of thromboembolic events associated with endovascular aneurysm treatment: retrospective case series. J Neurosurg. 2008;108: Bruening R, Mueller-Schunk S, Morhard D, Seelos KC, Brueckmann H, Schmid-Elsaesser R, et al. Intraprocedural thrombus formation during coil placement in ruptured intracranial aneurysms: treatment with systemic application of the glycoprotein IIb/IIIa antagonist tirofiban. AJNR Am J Neuroradiol. 2006;27: Fiorella D, Albuquerque FC, Han P, McDougall CG. Strategies for the management of intraprocedural thromboembolic complications with abciximab (reopro). Neurosurgery. 2004;54: ; discussion Gralla J, Rennie AT, Corkill RA, Lalloo ST, Molyneux A, Byrne JV, et al. Abciximab for thrombolysis during intracranial aneurysm coiling. Neuroradiology. 2008;50: Jones RG, Davagnanam I, Colley S, West RJ, Yates DA. Abciximab for treatment of thromboembolic complications during endovascular coiling of intracranial aneurysms. AJNR Am J Neuroradiol. 2008;29: Kang HS, Kwon BJ, Roh HG, Yoon SW, Chang HW, Kim JE, et al. Intraarterial tirofiban infusion for thromboembolism during endovascular treatment of intracranial aneurysms. Neurosurgery. 2008;63: ; discussion Linfante I, Etezadi V, Andreone V, DeLeo M, Alehashemi S, Shaw K, et al. Intra-arterial abciximab for the treatment of thrombus formation during coil embolization of intracranial aneurysms. J Neurointerv Surg. 2010;2: Mounayer C, Piotin M, Baldi S, Spelle L, Moret J. Intraarterial administration of Abciximab for thromboembolic events occurring during aneurysm coil placement. AJNR Am J Neuroradiol. 2003;24: Park JH, Kim JE, Sheen SH, Jung CK, Kwon BJ, Kwon OK, et al. Intraarterial abciximab for treatment of thromboembolism during coil embolization of intracranial aneurysms: outcome and fatal hemorrhagic complications. J Neurosurg. 2008;108: Song JK, Niimi Y, Fernandez PM, Brisman JL, Buciuc R, Kupersmith MJ, et al. Thrombus formation during intracranial aneurysm coil placement: treatment with intra-arterial abciximab. AJNR Am J Neuroradiol. 2004;25: Velat GJ, Burry MV, Eskioglu E, Dettorre RR, Firment CS, Mericle RA. The use of abciximab in the treatment of acute cerebral thromboembolic events during neuroendovascular procedures. Surg Neurol. 2006;65: , discussion Hähnel S, Schellinger PD, Gutschalk A, Geletneky K, Hartmann M, Knauth M, et al. Local intra-arterial fibrinolysis of thromboemboli occurring during neuroendovascular procedures with recombinant tissue plasminogen activator. Stroke. 2003;34: Cognard C, Weill A, Castaings L, Rey A, Moret J. Intracranial berry aneurysms: angiographic and clinical results after endovascular treatment. Radiology. 1998;206: Cronqvist M, Pierot L, Boulin A, Cognard C, Castaings L, Moret J. Local intraarterial fibrinolysis of thromboemboli occurring during endovascular treatment of intracerebral aneurysm: a comparison of anatomic results and clinical outcome. AJNR Am J Neuroradiol. 1998;19: Molyneux A, Kerr R, Stratton I, Sandercock P, Clarke M, Shrimpton J, 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 randomised trial. Lancet. 2002;360: Koebbe CJ, Horowitz MB, Levy EI, Dutton K, Jungries CC, Purdy PD. Intraarterial thrombolysis for thromboemboli associated with endovascular aneurysm coiling. Report of five cases. Interv Neuroradiol. 2002;8: