Intracranial Aneurysms: Midterm Outcome of Pipeline Embolization Device A Prospective Study in 143 Patients with 178 Aneurysms 1

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights. Simon Chun-Ho Yu, MD Ching-Kwong Kwok, FRCS Pui-Wai Cheng, FRCR Kwong-Yau Chan, FRCS Samuel Shun Lau, FRCR Wai-Man Lui, FRCS Ka-Ming Leung, FRCS Raymand Lee, FRCR Harold Kin-Ming Cheng, FRCS Yuk-Ling Cheung, FRCR Chi-Ming Chan, FRCR George Kwok-Chu Wong, FRCS Joyce Wai-Yi Hui, FRCR Yiu-Chung Wong, FRCR Chong-Boon Tan, FRCR Wai-Lun Poon, FRCR Kai-Yuen Pang, FRCS Alain Kai-Sing Wong, FRCS Kai-Hung Fung, FRCR 1 From the Department of Imaging and Interventional Radiology (S.C.H.Y., J.W.Y.H.) and Division of Neurosurgery, Department of Surgery (G.K.C.W.), the Chinese University of Hong Kong, Prince of Wales Hospital, 30-32 Ngan Shing Street, Room 2A061, 2/F, New Extension Block, Shatin, New Territories, Hong Kong; Departments of Neurosurgery (C.K.K., K.Y.C.) and Radiology (S.S.L.), Kwong Wah Hospital, Hong Kong; Scaing Department, St. Teresa Hospital, Hong Kong (P.W.C.); Division of Neurosurgery, Department of Surgery (W.M.L., K.M.L.) and Department of Radiology (R.L.), Queen Mary Hospital, Hong Kong; Department of Neurosurgery (H.K.M.C.) and Department of Radiology and Imaging (Y.L.C., C.M.C.), Queen Elizabeth Hospital, Hong Kong; Department of Radiology, Tuen Mun Hospital, Hong Kong (Y.C.W., C.B.T., W.L.P.); and Neurosurgical Unit, Department of Surgery (K.Y.P., A.K.S.W.) and Department of Radiology (K.H.F.), Pamela Youde Nethersole Eastern Hospital, Hong Kong. From the 2011 RSNA Aual Meeting. Received February 22, 2012; revision requested March 26; revision received April 20; accepted August 2; final version accepted August 2. Supported by the Vascular and Interventional Radiology Foundation. Address correspondence to S.C.H.Y. (e-mail: simonyu@cuhk.edu.hk). q RSNA, 2012 Intracranial Aneurysms: Midterm Outcome of Pipeline Embolization Device A Prospective Study in 143 Patients with 178 Aneurysms 1 Purpose: Materials and Methods: Results: Conclusion: To evaluate the midterm clinical and angiographic outcomes after pipeline embolization device (PED) placement for treatment of intracranial aneurysms. This prospective nonrandomized multicenter study was approved by the review boards of all involved centers; informed consent was obtained. Patients (143 patients, 178 aneurysms) with unruptured saccular or fusiform aneurysms or recurrent aneurysms after previous treatment were included and observed angiographically for up to 18 months and clinically for up to 3 years. Study endpoints included complete aneurysm occlusion; neurologic complications within 30 days and up to 3 years; clinical outcome of cranial nerve palsy after PED placement; angiographic evidence of occlusion or stenosis of parent artery and that of occlusion of covered side branches at 6, 12, and 18 months; and clinical and computed tomographic evidence of perforator infarction. There were five (3.5%) cases of periprocedural death or major stroke (modified Rankin Scale [mrs]. 3) (95% confidence interval [CI]: 1.3%, 8.4%), including two posttreatment delayed ruptures, two intracerebral hemorrhages, and one thromboembolism. Five (3.5%) patients had minor neurologic complications within 30 days (mrs = 1) (95% CI: 1.3%, 8.4%), including transient ischemic attack (n = 2), small cerebral infarction (n = 2), and cranial nerve palsy (n = 1). Beyond 30 days, there was one fatal intracerebral hemorrhage and one transient ischemic attack. Ten of 13 patients (95% CI: 46%, 93.8%) completely recovered from symptoms of cranial nerve palsy within a median of 3.5 months. Angiographic results at 18 months revealed a complete aneurysm occlusion rate of 84% (49 of 58; 95% CI: 72.1%, 92.2%), with no cases of parent artery occlusion, parent artery stenosis (,50%) in three patients, and occlusion of a covered side branch in two cases (posterior communicating arteries). Perforator infarction did not occur. PED placement is a reasonably safe and effective treatment for intracranial aneurysms. The treatment is promising for aneurysms of unfavorable morphologic features, such as wide neck, large size, fusiform morphology, incorporation of side branches, and posttreatment recanalization, and should be considered a first choice for treating unruptured aneurysms and recurrent aneurysms after previous treatments. q RSNA, 2012 Supplemental material: http://radiology.rsna.org/lookup /suppl/doi:10.1148/radiol.12120422/-/dc1 Original Research n Neuroradiology Radiology: Volume 265: Number 3 December 2012 n radiology.rsna.org 893

Although coil placement has been accepted as a treatment for intracranial aneurysms, unfavorable aneurysm features, such as wide neck, large size, fusiform morphology, and posttreatment recanalization, remain important challenges (1 6). Stent-assisted coil placement has been developed to address these challenges but has been associated with relatively high rates of aneurysm recurrence and procedure-induced mortality (4). Such limitations of coil placement have fueled the continual search for better endovascular options for treatment of intracranial aneurysms. Endovascular placement of flow diversion devices such as the pipeline embolization device (PED) (ev3 Neurovascular, Irvine, Calif) for endoluminal circumferential reconstruction of segmental vascular defects as a treatment for intracranial aneurysms is gaining widespread acceptance (7 10). The purpose of this study was to evaluate the midterm clinical and angiographic outcomes after PED placement for intracranial aneurysms. Materials and Methods Study Design This was a prospective multicenter study conducted in accordance to the Declaration of Helsinki and Declaration Advances in Knowledge Neurologic complications were rare beyond 30 days and did not occur beyond the 1st year. Approximately three-quarters of patients with cranial nerve palsy became asymptomatic after pipeline embolization device (PED) treatment. Delayed rupture after PED treatment did not occur in aneurysms smaller than 20 mm in size or in aneurysms treated with coil placement and 20 mm or larger in size. The occlusion rate of side branches covered by using PED was approximately 1%. of Good Clinical Practice and approved by the review boards of all involved centers. Signed informed consent was obtained. Inclusion criteria were as follows: (a) saccular or fusiform intracranial aneurysms; (b) untreated unruptured aneurysms or recurrent aneurysms after previous treatment; (c) aneurysms of a diameter of 10 mm or larger, dome-to-neck ratio of 1 or smaller, neck diameter of 4 mm or larger, or multiple aneurysms located within a 1-cm distance; and (d) parent vessel of a diameter of 2.5 5.0 mm distal or proximal to the target aneurysm. Exclusion criteria were as follows: (a) dissecting aneurysm (n = 14), (b) history of subarachnoid or intracerebral hemorrhage within the past 50 days (n = 13), (c) intracranial arteriovenous malformation (n = 1), or (d) parent artery stenosis of 50% or greater at the site of PED placement (n = 2). Between September 2008 and September 2011, 178 unruptured or previously treated aneurysms in 143 patients were treated in seven hospitals. There were 36 male patients (25.2%) and 107 female patients (74.8%), with a mean age of 54.9 years 6 11.4 (standard deviation) (median age, 55 years; age range, 27 82 years). At the time of data analysis Implications for Patient Care The use of concomitant endosaccular coil treatment at the time of flow diverter placement to build up a stable organized thrombus involving fibrin formation may be a reasonable treatment strategy to prevent posttreatment rupture. PED is a reasonably safe and effective treatment for intracranial aneurysms; the treatment is promising for aneurysms of unfavorable morphologic features, such as wide neck, large size, fusiform morphology, incorporation of side branches, and posttreatment recanalization. PED should be considered a first choice for treating unruptured aneurysms and recurrent aneurysms after previous treatments. in January 2012, the patients had been observed for an average of 17.6 months 6 10.8 (median, 18.2 months; range, 3 39.2 months). Study Endpoints The primary effectiveness endpoint was angiographic evidence of complete aneurysm occlusion at 6, 12, and 18 months and recanalization of the aneurysm. Complete occlusion was defined as complete aneurysm obliteration including the neck or class 1 by using the Montreal scale (6). Recanalization of aneurysm was defined as imaging evidence of incomplete aneurysm obliteration after an initial finding of complete obliteration. The primary safety endpoint was periprocedural death or major stroke within 30 days. A major stroke was defined as any stroke that resulted in severe disability, with modified Rankin Scale (mrs) score of 4 or 5. The secondary endpoints were as follows: (a) other complications within 30 days; (b) all neurologic complications as evaluated with mrs score during a follow-up period up to 3 years; (c) clinical outcome of cranial nerve palsy after PED placement up to 3 years; (d) angiographic evidence of occlusion or stenosis of a parent artery at 6, 12, and 18 Published online before print 10.1148/radiol.12120422 Content code: Radiology 2013; 265:893 901 Abbreviations: CI = confidence interval DSA = digital subtraction angiography mrs = modified Rankin Scale PED = pipeline embolization device Author contributions: Guarantors of integrity of entire study, S.C.H.Y., Y.L.C., J.W.Y.H., Y.C.W.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, S.C.H.Y., S.S.L., R.L., H.K.M.C., G.K.C.W., Y.C.W., K.Y.P.; clinical studies, S.C.H.Y., C.K.K., P.W.C., K.Y.C., S.S.L., W.M.L., K.M.L., R.L., H.K.M.C., Y.L.C., C.M.C., G.K.C.W., J.W.Y.H., Y.C.W., C.B.T., W.L.P., K.Y.P., K.H.F.; statistical analysis, S.C.H.Y., R.L., Y.C.W., C.B.T., K.H.F.; and manuscript editing, S.C.H.Y., C.K.K., P.W.C., K.Y.C., W.M.L., Y.L.C., G.K.C.W., Y.C.W., K.Y.P., A.K.S.W., K.H.F. Conflicts of interest are listed at the end of this article. 894 radiology.rsna.org n Radiology: Volume 265: Number 3 December 2012

months; (e) angiographic evidence of occlusion of covered arterial branches at 6, 12, and 18 months; and (f) clinical and computed tomographic (CT) evidence of infarction due to occlusion of perforating arteries covered by the PED up to 3 years. CT angiography or magnetic resonance (MR) angiography was used alternatively if digital subtraction angiography (DSA) was refused by the patient. Angiographic findings were reviewed independently and anonymously by two reviewers. Discrepancy was settled by using consensus. PED placement was the sole intervention without coils in most cases. A standard procedure for PED placement was adopted (9). Concomitant coil treatment of aneurysms at the time of PED placement was performed for previously untreated aneurysms 20 mm or larger in size, after the occurrence of delayed rupture of aneurysms after PED placement in two aneurysms 20 mm or larger in diameter early in the study. Coil treatment was performed to induce a stable organized thrombus involving fibrin formation, not for complete coil occlusion of the aneurysm; therefore, dense packing was not performed. Coil treatment for smaller aneurysms (,20 mm) was optional. All procedures were performed with the use of general anesthesia through the right common femoral artery by using a 6-F guiding catheter (Neuron Delivery Catheter, 6F 053; Penumbra, Alameda, Calif) and microcatheters (Renegade HI-Flo, Boston Scientific, Natick, Mass; Marksman, ev3 Neurovascular). The operating team in each center consisted of at least one operator with more than 10 years of experience in endovascular treatment of cerebral aneurysms (S.C.H.Y., C.K.K., P.W.C., W.M.L., H.K.M.C., Y.L.C., C.B.T., K.Y.P., and K.H.F.). Device diameter was at least equivalent to that of the proximal parent artery. The target was to place one device for each patient, unless an additional device was necessary for stent placement across the aneurysmal segment from normal artery to normal artery. PED apposition to vessel wall was assessed with DSA and nonsubtracted angiographic images. Good wall apposition was defined as perfect conformity of PED to vessel wall. The procedure was considered successful if the PED was placed at the vascular segment to completely cover the aneurysm. Patients were treated with daily oral clopidogrel (75 mg) and aspirin (80 mg) for 3 days before PED treatment and for 3 months afterward. At least 100 mg of daily oral aspirin was then given for another 3 months. A screening test for clopidogrel resistance was not performed. Heparin was administered intravenously during the procedure to maintain an activated clotting time between 250 and 300 seconds and was not reversed at the end of procedure. Statistical Analysis Subgroup analysis of correlation between the occurrence of complete occlusion of aneurysm within 6 months and six potential predictors was performed by using the Pearson x 2 test and the Fisher exact test. The potential predictors were as follows: (a) previously treated aneurysm versus untreated aneurysm; (b) aneurysm size smaller than 10 mm versus 10 mm or larger; (c) aneurysm neck less than 4 mm versus 4 mm or greater; (d) aneurysm dometo-neck ratio greater than 1 versus 1 or smaller; (e) location of aneurysm not in the ophthalmic segment or supraclinoid segment; and (f) use of more than one PED versus one PED. Subgroup univariate and multivariate analysis of the six potential predictors for occurrence of complete occlusion of aneurysm within 6 months were performed with a binary logistic regression model. Because some patients had multiple aneurysms, the observation of angiographic study endpoints on individual aneurysms was not independent in all cases; the difference in endpoints between single aneurysms and multiple aneurysms was analyzed by using the Pearson x 2 test, the Fisher exact test, and univariate and multivariate analysis by using a binary logistic regression model. Because this study involved seven different centers, the center effect was studied by analyzing correlation of periprocedural death and major stroke with particular centers by using the Fisher exact test. Software (SPSS, version 20; SPSS, Chicago, Ill) was used. A P value less than.05 was considered to indicate a significant difference. Results Patient Characteristics Among all 143 patients, 116 were asymptomatic (81.1%; 95% confidence interval [CI]: 73.5%, 87%), 14 had cranial nerve palsy (9.8%; 95% CI: 5.7%, 16.2%), and 13 presented with headache (9.1%; 95% CI: 5.1%, 15.4%). Thirty-four of 178 aneurysms (19.1%; 95% CI: 13.8%, 25.8%) had been treated previously, with clipping procedure (n = 1), wrapping procedure (n = 1), clipping procedure and coil treatment (n = 2), stent-assisted coil treatment (n = 7), and coil treatment (n = 23). The previous treatment was given for ruptured aneurysm in 26 of 34 cases (76%; 95% CI: 58.4%, 88.6%), at a median of 1.5 days (range, 0 13 days) after rupture and at a median of 16.1 months (range, 1.7 290 months) before PED placement. The mrs score was 0 in 127 patients before PED treatment. In the 14 patients with cranial nerve palsy, mrs score was 1 (n = 10) or 2 (n = 4). In the two patients with residual symptoms due to previous subarachnoid hemorrhage, mrs was 1 (n = 1) or 3 (n = 1). Aneurysm Characteristics There were 173 saccular aneurysms (97.2%; 95% CI: 93.2%, 99%) and five fusiform aneurysms (2.8%; 95% CI: 1%, 6.8%). Other aneurysm characteristics are shown in Table 1. Average diameters of the proximal and distal parent artery measured 3.7 mm 6 0.6 (range, 2.5 5.3 mm) and 3.4 mm 6 0.6 (range, 2.5 4.9 mm), respectively. Treatment Characteristics The procedure was successfully accomplished in all aneurysms. One PED was placed in 145 aneurysms (81.5%; 145 of 178; 95% CI: 74.8%, 86.7%), two in 32 aneurysms (18%; 32 of 178; 95% CI: 12.8%, 24.6%), and four in one aneurysm (0.5%; one of 178; 95% CI: 0%, Radiology: Volume 265: Number 3 December 2012 n radiology.rsna.org 895

3.6%). Suboptimal vessel wall apposition of PED occurred in two cases, one at the curvature of the carotid siphon, and the other at a preexisting stent (Neuroform; Boston Scientific) placed for stent-assisted coil treatment. Good wall apposition was achieved in both cases after balloon remodeling (Gateway; Boston Scientific). Acute thrombosis of the parent artery occurred in the case of poor wall apposition at the carotid siphon 15 minutes after PED placement. The vessel was completely recanalized after abciximab administration without ischemic consequence, before the suboptimal wall apposition was corrected with a balloon. Concomitant coil placement at the time of PED placement was performed in nine aneurysms that had not been treated previously, including four aneurysms 20 mm or larger in size and five aneurysms from 10 mm up to 20 mm (Table 2). Other treatment characteristics are shown in Table 2. Daily oral clopidogrel (75 mg) was given for 3 days before PED treatment and for 3 months afterward for each patient. The median dose of aspirin was 100 mg (range, 80 300 mg). The median dose of heparin was 3000 IU (range, 1000 8000 IU). The average duration of hospital stay for PED placement was 3.8 days 6 3.2 (median, 3 days; range, 1 21 days). Primary Endpoints Among 210 follow-up angiographic examinations, there were 103 DSA examinations (49%; 95% CI: 42.1%, 56%), 64 MR angiographic examinations (30.5%; 95% CI: 24.4%, 37.3%), and 43 CT angiographic examinations (20.5%; 95% CI: 15.4%, 26.7%). Complete occlusion was found in 78 of 140 aneurysms (55.7%; 95% CI: 47.1%, 64%) at 6 months, 61 of 75 aneurysms (81%; 95% CI: 70.3%, 89.1%) at 12 months, and 49 of 58 aneurysms (84%; 95% CI: 72.1%, 92.2%) at 18 months (Table 3). In 66 aneurysms with complete occlusion at 6 months, 33 were followed to 12 months and 28 to 18 months; late recanalization of an aneurysm did not occur in this group. For the five cases of fusiform aneurysm, complete occlusion occurred in three cases (60%; 95% CI: 17%, 92.7%) at 6 months. No bleeding Table 1 Aneurysm Characteristics Characteristic No. of Occurrences (n = 178) Percentage* Aneurysm diameter (mean, 7.0 mm 6 5.5; median, 5.0 mm; range, 1.9 30 mm),10 mm 145 81.5 (74.8, 86.7) 10 25 mm 29 16.3 (11.4, 22.7).25 mm 4 2.2 (0.7, 6) Neck size (mean, 5.3 mm 6 4.0; median, 4.0 mm; range, 1.1 28 mm),4 mm 82 46.1 (38.6, 53.7) 4 mm 96 53.9 (46.3, 61.4) Dome-to-neck ratio (mean, 1.4 6 0.9; median, 1.0; range, 0.6 6.7) 1 95 53.4 (45.8, 60.8).1 83 46.6 (39.2, 54.2) Aneurysm location Ophthalmic segment ICA 55 30.9 (24.3, 38.3) Supraclinoid ICA 50 28.1 (21.8, 35.4) Siphon ICA 32 18 (12.8, 24.6) PCA segment of ICA 28 15.7 (10.9, 22.1) Distal ICA 4 2.2 (0.7, 6) Vertebral artery 4 2.2 (0.7, 6) M1 segment of MCA 3 1.7 (0.4 5.3) A1 segment of ACA 2 1.1 (0.2, 4.4) Angiographic evidence of intrasaccular thrombus 7 3.9 (1.7, 8.3) before treatment Stenosis of parent artery at the site of aneurysm before treatment 25% 2 1.1 (0.2, 4.4).25% to 50 11 6.2 (3.3, 11.1) Previous treatment Yes 34 19.1 (13.8, 25.8) No 144 80.9 (74.2, 86.2) Note. ACA = anterior cerebral artery, ICA = internal carotid artery, MCA = middle cerebral artery, PCA = posterior communicating artery. * Data in parentheses are 95% CIs. Pretreatment intrasaccular thrombus in three cases. Pretreatment intrasaccular thrombus in one case. from an aneurysm occurred in cases of incomplete aneurysm occlusion, apart from two cases of delayed rupture after PED placement that occurred within 30 days in large-size aneurysms ( 20 mm). These two aneurysms were not occluded immediately after PED. Subgroup analysis did not identify any factors correlated with early occurrence of complete occlusion of aneurysm within 6 months. No difference in 6-month aneurysm occlusion rate between single aneurysms and multiple aneurysms was observed (Table E1 [online]). The periprocedural death or major stroke rate was 3.5% (five of 143; 95% CI: 1.3%, 8.4%) (Table 4). There was one case of ipsilateral ischemic stroke due to a small left capsular infarct immediately after PED placement (mrs = 4). There were two cases of major hemorrhagic stroke due to a large left parieto-occipital intracerebral hemorrhage (mrs = 5) and a large cerebellar hemorrhage (mrs = 6), which occurred in vascular territories independent from locations of instrumentation or PED placement for aneurysms 896 radiology.rsna.org n Radiology: Volume 265: Number 3 December 2012

in the right carotid siphon and left internal carotid artery, respectively. The former occurred 27 days after PED placement. The latter occurred on the day of PED placement after abciximab administration for acute parent artery thrombosis. There were two cases of major hemorrhagic stroke (mrs = 5, 6) due to posttreatment delayed rupture of large aneurysms that were untreated previously (22 mm, 25 mm). Delayed rupture occurred in two of five (40%; 95% CI: 7.3%, 83%) uncoiled aneurysms 20 mm or larger in size and in none of six (0%) coiled aneurysms 20 mm or larger. These six coiled aneurysms included four with concomitant coil placement performed at the time of PED placement (Fig 1) and two treated previously with coil placement. No other aneurysm ruptures occurred. Periprocedural death or major stroke occurred in three centers, at a rate of 9% (three of 33), 7% (one of 15), and 3% (one of 32), respectively. No correlation with a particular center was observed (P =.64). Secondary Endpoints Apart from major stroke or death, there were six other cases of periprocedural complications occurring in five patients (3.5%; five of 143; 95% CI: 1.3%, 8.4%). All were minor, including two transient ischemic attacks within 1 day after PED placement (mrs = 1), one small ipsilateral occipital infarction immediately after PED placement (mrs = 1), one small contralateral thalamic and corona radiata infarct 1 week after self-withdrawal from aspirin (mrs = 1), one cranial nerve palsy that occurred after the procedure and subsided after 1 week (mrs = 1), and one retroperitoneal hematoma related to femoral arterial puncture (mrs = 0). Beyond 30 days, there were two neurologic complications (1.4%; two of 134; 95% CI: 0.2%, 5.5%). One patient died of a recurrent intracerebral hemorrhage at the left parieto-occipital lobe 11 months after PED placement. The site of hemorrhage was unrelated to the treated aneurysm or PED in the right carotid siphon. Repeated transient ischemic attacks within 6 months occurred in a Table 2 Treatment Characteristics Treatment Characteristic Concomitant coil treatment for previously untreated aneurysms (median, five coils; range, four to eight coils) Aneurysm size 10 mm to, 20 mm Aneurysm size 20 mm to 25 mm Aneurysm size. 25 mm Aneurysms previously treated with coil placement (,10 mm, n = 21; 10 mm, n = 10) Aneurysm size 10 mm to, 20 mm Aneurysm size 20 mm to 25 mm Aneurysm size. 25 mm No. of arterial side branches covered by PED per case (total 174 branches covered) No branch One branch Two branches Three branches Four branches No. of PEDs covering each arterial side branch (total 174 branches) One PED Two PEDs Four PEDs Arterial side branch covered by PED (total 174 branches) Ophthalmic artery Posterior communicating artery Table 3 Anterior choroidal artery Middle cerebral artery (with perforating arteries) Posterior inferior cerebellar artery Anterior cerebral artery Angiographic Outcome patient who often stopped taking aspirin (mrs = 1). Neurologic complication did not occur beyond the 1st year. In total, Datum 5 of 14 aneurysms 2 of 7 aneurysms 2 of 2 aneurysms 8 aneurysms 0 aneurysms 2 aneurysms 22 cases 104 cases 36 cases 10 cases 1 case 148 branches (two became occluded) 25 branches 1 branch 107 cases 37 cases (two became occluded) 22 cases 3 cases 3 cases 2 cases Parameter 6 Months 12 Months 18 Months No. of aneurysm or parent 140 75 58 artery examined Incidence (rate) of complete occlusion 78 (55.7) [47.1, 64] 61 (81) [70.3, 89.1] 49 (84) [72.1, 92.2] of aneurysm* Recanalization of aneurysm 0 0 Incidence of parent artery occlusion 0 0 0 Incidence of parent artery stenosis 50% 2 3 3.50% 0 0 0 No. of arterial side branches examined 140 70 58 Incidence of side branch occlusion 2 2 2 * Data in parentheses are percentages, and data in brackets are 95% CIs. neurologic complications occurred in 12 patients after PED treatment (8.4%; 12 of 143; 95% CI: 4.6%, 14.5%). Radiology: Volume 265: Number 3 December 2012 n radiology.rsna.org 897

In the 12 patients with neurologic complications (mrs 1) occurring within 30 days or beyond 30 days after PED placement, 11 of them had an mrs score of 0 before PED placement and one of them had an mrs score of 1 before PED placement because of cranial nerve palsy. One hundred sixteen patients with an mrs score of 0 before PED placement had a postprocedural mrs score that remained 0 throughout follow-up period. The mrs score had gradually improved in 10 of 13 patients with cranial nerve palsy and in two patients with residual disability due to previous subarachnoid hemorrhage. Cranial nerve palsy involving the third, fourth, or sixth nerve occurred in 14 patients before the PED procedure. One of them died of a periprocedural complication, 13 patients were left for follow-up. Aneurysm size was smaller than 10 mm (n = 6), 10 25 mm (n = 4), or greater than 25 mm (n = 3). The mrs score before the PED procedure was 1 (n = 9) and 2 (n = 4). During the follow-up period, 10 of 13 patients (77%; 95% CI: 46%, 93.8%) completely recovered from symptoms of cranial nerve palsy within a median of 3.5 months (range, 1 8.9 months). The symptoms remained unchanged in the other three patients. Among the 10 patients with cranial nerve function who recovered after PED placement, five had their aneurysms completely occluded at the time of recovery. Among the three patients with symptoms of cranial nerve palsy that remained unchanged after PED placement, two of them had their aneurysms completely occluded at the time of assessment. All parent arteries remained patent in all 178 aneurysms during the follow-up period. Parent artery stenosis occurred in two of 140 cases at 6 months (1.4%; 95% CI: 0.3%, 5.6%), to a degree of 20% and 30%, without clinical consequence. Angiographic follow-up at 12 and 18 months revealed one further case of 40% stenosis. All the stenoses were located at the internal carotid artery, involving the ophthalmic segment (n = 2) and the supraclinoid segment (n = 1). Occlusion rate of covered arterial side Figure 1 Figure 1: Cranial nerve palsy in a 55-year-old woman. (a) Lateral DSA image shows a large intracranial aneurysm (24-mm diameter) at supraclinoid internal carotid artery. (b) Frontal radiograph shows PED (between arrows) after placement across the neck of aneurysm after concomitant coil placement at same session. Dense packing was not performed because the purpose was not to achieve complete coil occlusion of the aneurysm. (c) Lateral DSA image subsequently shows partial stasis of contrast media at the dependent part of the aneurysm. (d) Frontal DSA image shows complete obliteration of the aneurysm (arrow) 6 months after treatment. Table 4 Death or Major Stroke Up to 3 Years Parameter 30 Days.30 Days to 12 Months 13 24 Months 25 36 Months No. of patients observed 143 132 96 45 within period No. of deaths 2 (1.4) [0.2, 5.5]* 1 (0.7) [0, 4.8] 0 1 No. of ipsilateral ischemic stroke 1 (0.7) [0, 4.4] 0 0 0 No. of ipsilateral subarachnoid 2 (1.4) [0.2, 5.5] 0 0 0 hemorrhage No. of contralateral 2 (1.4) [0.2, 5.5] 1 (0.7) [0, 4.8] 0 0 intracerebral hemorrhage All stroke 5 (3.5) [1.3, 8.4] 1 (0.7) [0, 4.8] 0 0 Death or stroke 5 (3.5) [1.3, 8.4] 1 (0.7) [0, 4.8] 0 1 Note. Data in parentheses are percentages, and data in brackets are 95% CIs. * Due to ipsilateral subarachnoid hemorrhage and contralateral intracerebral hemorrhage. Due to pulmonary tuberculosis at 27th month. 898 radiology.rsna.org n Radiology: Volume 265: Number 3 December 2012

Figure 2 Figure 2: Fusiform aneurysm of M1 segment of right middle cerebral artery that was incidentally discovered at CT angiography in a 32-year-old woman. (a) Oblique DSA image obtained just before PED placement shows the fusiform aneurysm (arrow), a small saccular aneurysm next to it, and adjacent perforating arteries. (b, c) After placement of two PEDs (size, 2.5 3 20 mm and 2.75 3 20 mm), oblique angiograms (b) without subtraction and (c) with subtraction show coverage of the two aneurysms and perforating arteries by PED (between arrows). (d) Oblique DSA image shows complete obliteration of the two aneurysms and preservation of the perforating arteries 3 months after treatment. branches at 6 months was 1.4% (two of 140; 95% CI: 0.3%, 5.6%). Both occluded branches were posterior communicating arteries, covered by one PED. There was no clinical consequence (Table 3). Because of the low incidence of these angiographic endpoints, difference between single aneurysms and multiple aneurysms was unlikely and not analyzed. In three patients with perforating arteries covered by PED at the M1 segment, which involved two PEDs (n = 1) and one PED (n = 2), clinical or CT evidence of perforator infarction did not occur (Fig 2). Discussion In our series of PED treatment for unruptured intracranial aneurysms or recurrent aneurysms after a previous treatment, high rates of complete aneurysm occlusion and preservation of the parent artery were achieved. Although complete aneurysm occlusion rate was high, occlusion occurred at a slow pace, and the reason for this delayed occlusion rate remains unclear. Periprocedural death or major stroke rate was low and appeared to be further improved with concomitant coil treatment to prevent delayed rupture after PED placement. Neurologic complications were rare beyond 30 days and did not occur beyond the 1st year. The majority of patients with cranial nerve palsy became asymptomatic after PED treatment. Arterial complications such as parent artery occlusion, parent artery stenosis, or side branch occlusion were nonexistent or rare. Perforator infarction did not occur. These findings suggest that PED therapy represents a viable treatment option for these patients. For intracranial aneurysms with unfavorable morphology for coil treatment and aneurysms with recurrence after coil treatment, the availability of an Radiology: Volume 265: Number 3 December 2012 n radiology.rsna.org 899

alternative treatment option of reasonable safety and effectiveness is of substantial clinical importance. Knowledge of the complication risks of coil placement with or without balloon or stent assistance in treatment of unruptured intracranial aneurysms may provide a reference on the safety of PED treatment for unruptured aneurysms. In the Analysis of Treatment by Endovascular Approach of Nonruptured Aneurysms, or ATENA, study of 649 patients with unruptured intracranial aneurysms, thromboembolic complications and intraprocedural aneurysm rupture occurred in 9.7% of procedures, and neurologic complications including death occurred in 5.4% of patients (11). With the relatively high aneurysm recurrence rate of 14.9% with endovascular coil treatment (4), PED is a reasonably promising alternative treatment. A review of six prior studies on flow diverters for unruptured intracranial aneurysms, including the PED and SILK flow diverter (Balt Extrusion, Montmorency, France) (7 10,12,13) showed that procedure success rates in terms of device placement were high for the two devices and had always been 100% successful for the PED (7,8,10). Complete aneurysm occlusion rates at 6 months had been high. Suboptimal apposition of device to vessel wall and acute or delayed thrombosis of parent artery were extremely uncommon or rare. Stenosis of parent artery at 6 months and delayed posttreatment aneurysm were also extremely rare. The results of these previous studies have suggested that the safety and treatment effectiveness of flow diverters are promising. These findings have been confirmed in our study. There are certain study parameters that have been evaluated in our study but not covered in the previous studies, including the evaluation of longer-term clinical outcome up to 3 years and longer-term angiographic outcome up to 6 months, evaluation of the effect of PED treatment on aneurysm-related cranial nerve palsy, and evaluation of the risk of side branch occlusion due to PED placement. The results of our study on these study parameters were all favorable. Our study showed a finding different from what was reported previously. Perforator infarction due to compromise of perforator perfusion had been reported in two patients within 2 days after double PED placement at M1 and A1 segments, respectively (9,14), and in three patients at 4 weeks to 7 months after SILK flow diverter placement at the basilar artery (15). Perforator infarction was not observed in our study, and the reason is unclear; perhaps further studies in larger scales are required to clarify this point. A major concern with flow diverters is their inability to immediately occlude the aneurysm with the risk of aneurysm rupture during the latency period (16). Delayed rupture of large-size aneurysms (17 35 mm) after treatment with flow diverters did occur with unknown incidence (12,13,17,18). Rupture may occur as early as 2 48 days or as late as 5 months (13,17). Reported cases so far involved the use of a SILK flow diverter (12,13,17,18), but not a PED (7 10); the current study showed that aneurysms treated with the PED are not exempted. Hypotheses for the mechanisms of rupture included proximal migration of the flow diverter with redirection of the jet inside the aneurysm sac (12) and aneurysm wall weakening due to a high content of proteolytic enzymes formed within nonorganized red thrombus as a result of flow stagnation developed within a short period (18). For large-size aneurysms harboring partial thrombosis, the approach of endoluminal coil placement had been challenged, and surgical excision or parent artery occlusion had been advocated, because these aneurysms are likely extraluminal disorders involving subacute or chronic dissections, repeated intramural hematomas, proliferating vasa vasorum, and triggering of inflammatory mechanisms, such that subarachnoid hemorrhage from the aneurysms originates from the periphery of their walls instead of the lumen of their sacs or necks (19). Whether treatment with flow diverters is reliable for this vasculopathy remains unknown at the moment. However, the use of concomitant endosaccular coil placement at the time of flow diverter placement to build up a stable organized thrombus involving fibrin formation may be a reasonable treatment strategy to prevent posttreatment rupture. Adoption of such preventive measures in the current study had resulted in no further incidence of aneurysm rupture; further observation in a larger study is warranted to verify such a proposal. The study was limited by imperfect follow-up. Furthermore, the majority of the aneurysms were located at the internal carotid artery and were small. Evolution of aneurysm volume was not studied. We caot confidently identify specific risk factors for given complications, given the small sample size and rare nature of these events. Further studies on the effect of PED treatment for fusiform aneurysms and the risk of delayed rupture of large-size aneurysms after PED treatment are necessary. In conclusion, our findings suggested that PED placement is a reasonably safe and effective treatment for intracranial aneurysms; the treatment is promising for aneurysms of unfavorable morphologic features, such as wide neck, large size, fusiform morphology, incorporation of side branches, and posttreatment recanalization, and should be considered a first choice for treating unruptured aneurysms and recurrent aneurysms after previous treatments. Disclosures of Conflicts of Interest: S.C.H.Y. No relevant conflicts of interest to disclose. C.K.K. No relevant conflicts of interest to disclose. P.W.C. No relevant conflicts of interest to disclose. K.Y.C. No relevant conflicts of interest to disclose. S.S.L. No relevant conflicts of interest to disclose. W.M.L. No relevant conflicts of interest to disclose. K.M.L. No relevant conflicts of interest to disclose. R.L. No relevant conflicts of interest to disclose. H.K.M.C. No relevant conflicts of interest to disclose. Y.L.C. No relevant conflicts of interest to disclose. C.M.C. No relevant conflicts of interest to disclose. G.K.C.W. No relevant conflicts of interest to disclose. J.W.Y.H. No relevant conflicts of interest to disclose. Y.C.W. No relevant conflicts of interest to disclose. C.B.T. No relevant conflicts of interest to disclose. W.L.P. No relevant conflicts of interest to disclose. K.Y.P. No relevant conflicts of interest to disclose. A.K.S.W. No relevant conflicts of interest to disclose. K.H.F. No relevant conflicts of interest to disclose. 900 radiology.rsna.org n Radiology: Volume 265: Number 3 December 2012

References 1. Molyneux A, Kerr R, Stratton I, et al. International Subarachnoid Aneurysm Trial (ISAT) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: a randomised trial. Lancet 2002;360(9342): 1267 1274. 2. Friedman JA, Nichols DA, Meyer FB, et al. Guglielmi detachable coil treatment of ruptured saccular cerebral aneurysms: retrospective review of a 10-year single-center experience. AJNR Am J Neuroradiol 2003; 24(3):526 533. 3. Vallee JN, Aymard A, Vicaut E, Reis M, Merland JJ. Endovascular treatment of basilar tip aneurysms with Guglielmi detachable coils: predictors of immediate and long-term results with multivariate analysis 6-year experience. Radiology 2003;226(3): 867 879. 4. Piotin M, Blanc R, Spelle L, et al. Stentassisted coiling of intracranial aneurysms: clinical and angiographic results in 216 consecutive aneurysms. Stroke 2010;41(1): 110 115. 5. Johnston SC, Dowd CF, Higashida RT, et al. Predictors of rehemorrhage after treatment of ruptured intracranial aneurysms: the Cerebral Aneurysm Rerupture After Treatment (CARAT) study. Stroke 2008;39(1):120 125. 6. Raymond J, Guilbert F, Weill A, et al. Longterm angiographic recurrences after selective endovascular treatment of aneurysms with detachable coils. Stroke 2003;34(6): 1398 1403. 7. Lylyk P, Miranda C, Ceratto R, et al. Curative endovascular reconstruction of cerebral aneurysms with the pipeline embolization device: the Buenos Aires experience. Neurosurgery 2009;64(4):632 642. 8. Szikora I, Berentei Z, Kulcsar Z, et al. Treatment of intracranial aneurysms by functional reconstruction of the parent artery: the Budapest experience with the pipeline embolization device. AJNR Am J Neuroradiol 2010;31(6):1139 1147. 9. Nelson PK, Lylyk P, Szikora I, Wetzel SG, Wanke I, Fiorella D. The pipeline embolization device for the intracranial treatment of aneurysms trial. AJNR Am J Neuroradiol 2011;32(1):34 40. 10. McAuliffe W, Wycoco V, Rice H, Phatouros C, Singh TJ, Wenderoth J. Immediate and midterm results following treatment of unruptured intracranial aneurysms with the pipeline embolization device. AJNR Am J Neuroradiol 2012;33(1):164 170. 11. Pierot L, Spelle L, Vitry F; ATENA Investigators. Immediate clinical outcome of patients harboring unruptured intracranial aneurysms treated by endovascular approach: results of the ATENA study. Stroke 2008; 39(9):2497 2504. 12. Lubicz B, Collignon L, Raphaeli G, et al. Flow-diverter stent for the endovascular treatment of intracranial aneurysms: a prospective study in 29 patients with 34 aneurysms. Stroke 2010;41(10):2247 2253. 13. Byrne JV, Beltechi R, Yarnold JA, Birks J, Kamran M. Early experience in the treatment of intra-cranial aneurysms by endovascular flow diversion: a multicentre prospective study. PLoS ONE 2010;5(9):e12492. 14. van Rooij WJ, Sluzewski M. Perforator infarction after placement of a pipeline flow-diverting stent for an unruptured A1 aneurysm. AJNR Am J Neuroradiol 2010;31(4):E43 E44. 15. Kulcsár Z, Ernema U, Wetzel SG, et al. High-profile flow diverter (silk) implantation in the basilar artery: efficacy in the treatment of aneurysms and the role of the perforators. Stroke 2010;41(8):1690 1696. 16. D Urso PI, Lanzino G, Cloft HJ, Kallmes DF. Flow diversion for intracranial aneurysms: a review. Stroke 2011;42(8):2363 2368. 17. Kulcsár Z, Houdart E, Bonafé A, et al. Intraaneurysmal thrombosis as a possible cause of delayed aneurysm rupture after flow-diversion treatment. AJNR Am J Neuroradiol 2011;32(1):20 25. 18. Turowski B, Macht S, Kulcsár Z, Hänggi D, Stummer W. Early fatal hemorrhage after endovascular cerebral aneurysm treatment with a flow diverter (SILK-Stent): do we need to rethink our concepts? Neuroradiology 2011;53(1):37 41. 19. Alvarez H. Etiology of giant aneurysms and their treatment. AJNR Am J Neuroradiol 2009;30(1):E8. Radiology: Volume 265: Number 3 December 2012 n radiology.rsna.org 901