Abstract

BACKGROUND: Early and mid-term safety and efficacy of aneurysm treatment with the Pipeline Embolization Device (PED) has been well demonstrated in prior studies.

OBJECTIVE: To present 5-yr follow-up for patients treated in the Pipeline for Uncoilable or Failed Aneurysms clinical trial.

METHODS: In our prospective, multicenter trial, 109 complex internal carotid artery (ICA) aneurysms in 107 subjects were treated with the PED. Patients were followed per a standardized protocol at 180 d and 1, 3, and 5 yr. Aneurysm occlusion, in-stent stenosis, modified Rankin Scale scores, and complications were recorded.

RESULTS: The primary endpoint of complete aneurysm occlusion at 180 d (73.6%) was previously reported. Aneurysm occlusion for those patients with angiographic follow-up progressively increased over time to 86.8% (79/91), 93.4% (71/76), and 95.2% (60/63) at 1, 3, and 5 yr, respectively. Six aneurysms (5.7%) were retreated. New serious device-related events at 1, 3, and 5 yr were noted in 1% (1/96), 3.5% (3/85), and 0% (0/81) of subjects. There were 4 (3.7%) reported deaths in our trial. Seventy-eight (96.3%) of 81 patients with 5-yr clinical follow-up had modified Rankin Scale scores ≤2. No delayed neurological deaths or hemorrhagic or ischemic cerebrovascular events were reported beyond 6 mo. No recanalization of a previously occluded aneurysm was observed.

CONCLUSION: Our 5-yr findings demonstrate that PED is a safe and effective treatment for large and giant wide-necked aneurysms of the intracranial ICA, with high rates of complete occlusion and low rates of delayed adverse events.

Large and giant aneurysms of the internal carotid artery (ICA) have a poor prognosis when left untreated.1,2 Parent artery occlusion and endosaccular coiling have been considered the standard treatment modalities for these lesions in the past. Parent artery occlusion with or without bypass3 has been shown to result in high long-term occlusion rates in the subset of ICA aneurysms proximal to the ophthalmic segment but is associated with moderately high complication rates.4 Conversely, endosaccular coiling of complex aneurysms, with or without stent assistance, is considered relatively safe but has been shown to result in high rates of aneurysm recanalization, often requiring retreatment.5-7 Endoluminal reconstruction with the Pipeline Embolization Device (PED) has emerged as a viable and often preferable alternative to traditional endosaccular coiling and parent artery occlusion techniques.8,9 By excluding the aneurysm from the parent artery, the PED promotes thrombosis of the aneurysm sac and neointimal overgrowth at the aneurysm neck, both of which serve to prevent aneurysm growth and rupture and ultimately lead to aneurysm occlusion.10

The Pipeline for Uncoilable or Failed Aneurysms (PUFS) study is a multicenter, prospective, single-arm interventional trial of PED for treatment of uncoilable or failed aneurysms of the ICA. We previously reported 180-d, 1-, and 3-yr angiographic and clinical outcomes, which demonstrated high angiographic success rates and low complication rates.8,11 Here, we report results from patients presenting for 5-yr clinical and angiographic follow-up as part of the PUFS trial.

METHODS

Study Enrollment and Patient Selection

One hundred eight patients from 10 centers were enrolled in PUFS between November 2008 and July 2009 under a US Food and Drug Administration Investigational Device Exemption. Of these, 104 subjects with 106 aneurysms constituted the effectiveness cohort of the trial and were followed for angiographic occlusion, while 107 patients constituted the safety cohort and were followed for functional outcomes and adverse events. Complete enrollment and exclusion criteria have been previously described.8 Briefly, patients were included if they had an ICA aneurysm (petrous through superior hypophyseal segments) measuring greater than 10 mm in diameter and with a neck measuring greater than 4 mm. Patients who had undergone prior treatments (ie, endosaccular coiling or neurosurgical wrapping) of the target aneurysm were not excluded unless another endoluminal device had been previously placed across the aneurysm neck.

Treatment Description

Placement of the PED was performed as previously described.8 In brief, all procedures were performed under general anesthesia using standard transfemoral approaches. PEDs were deployed through a 0.027-inch diameter catheter (Hi-Flow Renegade; Boston Scientific, Fremont, CA; or Marksman Catheter; ev3/Covidien, Dublin, Ireland). The size and number of PEDs placed were at the discretion of each treating operator, with a median of 3 devices used per aneurysm.

Patients were prescribed aspirin (325 mg per day orally for 2 d) and clopidogrel (75 mg per day for 7 d or a 600-mg loading dose 1 d prior to the procedure) prior to PED placement. After the procedure, patients were prescribed 325 mg of aspirin daily for at least 6 mo and 75 mg of clopidogrel daily for at least 3 mo. Most patients were maintained on combined antiplatelet therapy for 6 mo. Platelet inhibition studies were optional.

Baseline and Follow-up Assessments

Prior to placement of the PED, patients underwent a baseline neurological examination and a detailed neuroophthalmologic assessment to document visual deficits or cranial neuropathy. Patients underwent repeat neurological examinations at 30 and 180 d, and 1, 3, and 5 yr after PED placement as well as angiography at 180 d, and 1, 3, and 5 yr after placement. Mandatory neuro-ophthalmologic follow-up was limited to 180 d.12 Follow-up imaging was interpreted by an independent Core Radiology Laboratory (CRL) consisting of 3 neuroradiologists. Each CRL member independently adjudicated aneurysms for degree of occlusion (complete occlusion, residual neck, or residual aneurysm), the presence and degree of in-stent stenosis, and the occurrence of implant migration.

At 5 yr post-treatment, 81 of 103 surviving patients in the safety cohort (78.6%) completed all mandated clinical evaluations. In the effectiveness cohort, 75 (75%) of 100 surviving subjects with 77 aneurysms had 5-yr imaging follow-up; 61 of these with 63 aneurysms underwent catheter angiography.

At 180-d follow-up, there were 15 patients who had angiographically documented incomplete aneurysm occlusions. Fourteen of these 15 subjects underwent additional angiographic studies at 3 and/or 5 yr (Table 2).

Safety Reporting and Endpoints

An independent Clinical Events Committee reviewed and adjudicated each serious adverse event. An event was considered a serious adverse event if it met international clinical trial standard criteria (ISO 14155). Investigators were asked to judge the relationship of each adverse event to the PED, PED placement procedure, antiplatelet medication, and any pre-existing conditions.

The primary effectiveness endpoint of the study was complete occlusion of the aneurysm without major (>50%) stenosis of the parent artery or adjunctive use of complimentary embolic agent as seen on 180-d angiography and judged by the independent CRL.8 A case was considered successful if at least 2 of the 3 CRL members agreed that it met these criteria. The same definition was used to judge effectiveness at 1, 3, and 5 yr.

The primary safety endpoint of the PUFS study was the incidence of major ipsilateral stroke (defined as an increase of ≥4 points on the National Institute of Health Stroke Scale and present after 7 d as adjudicated by the Clinical Events Committee) or neurological death within 180 d of PED placement. In the current report, we present the following prespecified secondary endpoints observed at the 5-yr time point: (1) complete aneurysm occlusion, (2) incidence of significant in-stent stenosis (>50%), (3) occurrence of delayed device-related adverse events, and (4) functional outcomes based on the modified Rankin Scale.

Statistical Analysis

Summary statistics (counts and percentages) were tabulated. Exact binomial confidence intervals for major study endpoints were calculated with binom.test from the base stats R library, as described here: https://stat.ethz.ch/R-manual/R-devel/library/stats/html/binom.test.html. A posterior Bayesian distribution of the rate of complete target aneurysm occlusion was calculated by using a noninformative beta prior (1,1) distribution. Missing data resulting from lack of patient follow-up were generally ignored. All statistical calculations were performed using R (Vienna, Austria; www.R-project.org).13

RESULTS

Baseline Characteristics

Baseline characteristics were reported in the initial PUFS study.8 Of the 108 patients with 110 aneurysms enrolled, 109 aneurysms in 107 patients were treated with PED (including 1 aneurysm treated with adjunctive coils). The mean aneurysm dome size was 18.2 mm and the mean neck size was 8.8 mm. Figure details the subject follow-up at the 5-yr time point.

FIGURE.

Subject follow-up flowchart.

FIGURE.

Subject follow-up flowchart.

Angiographic Results and Retreatment

Table 1 details angiographic occlusion rates of the aneurysms at 180 d, and 1, 3, and 5 yr post-treatment. The 180-d, 1-, and 3-yr occlusion rates were previously reported.8,11 At 5-yr follow-up, 61 patients in the effectiveness cohort, harboring 63 aneurysms, had catheter angiographic imaging. Among these subjects, the angiographic occlusion rate was 95.2% (95% posterior credible interval [PCI] 86.7%-99%). There were 2 neck remnants and 1 residual aneurysm (3/63; 4.76% [PCI 0.993%-13.3%]).

TABLE 1.

Aneurysm Occlusion at Follow-up Time Pointsa

 Follow-up time point 
Occlusion type 180 d 1 yr 3 yr 5 yr 
Complete occlusion 73.6% (78/106) 86.8% (79/91) 93.4% (71/76) 95.2% (60/63) 
Residual neck 7.5% (8/106) 5.5% (5/91) 2.6% (2/76) 1.6% (1/63) 
Residual aneurysm 5.7% (6/106) 5.5% (5/91) 2.6% (2/76) 3.2% (2/63) 
Other 13.2% (14/106)b 2.2% (2/91)c 1.3% (1/76)c 0.0% (0/63) 
 Follow-up time point 
Occlusion type 180 d 1 yr 3 yr 5 yr 
Complete occlusion 73.6% (78/106) 86.8% (79/91) 93.4% (71/76) 95.2% (60/63) 
Residual neck 7.5% (8/106) 5.5% (5/91) 2.6% (2/76) 1.6% (1/63) 
Residual aneurysm 5.7% (6/106) 5.5% (5/91) 2.6% (2/76) 3.2% (2/63) 
Other 13.2% (14/106)b 2.2% (2/91)c 1.3% (1/76)c 0.0% (0/63) 

aBased on angiographic imaging at each follow-up.

bIncludes 3 carotid occlusions.

cNew carotid occlusions.

As reported earlier,8 15 patients had aneurysm or neck remnants at 180 d. The mean diameter for these 15 aneurysms was 17.1 mm and the mean neck size was 9.1 mm. Six of these subjects (5.66% [PCI 2.11%-11.9%]) received additional endovascular treatment: 5 (4.72% [PCI 1.55%-10.7%]) with additional PEDs and 1 who developed a post-PED carotid-cavernous fistula that was occluded with transvenous coils. This resulted in 4 complete aneurysm occlusions, 1 persistent neck remnant, and 1 aneurysm remnant within this subgroup at 5 yr (Table 2). Of the 9 remnants at 180-d follow-up that were not retreated, 3 were occluded at 3-yr angiographic follow-up (but had no 5-yr imaging), 4 were found to be occluded both at the 3-yr catheter angiogram and at the 5-yr follow-up (3 by catheter angiography and 1 with magnetic resonance angiography), 1 had a residual neck at 5 yr, and 1 did not have follow-up imaging. Overall, 11 of the 15 aneurysms (73.3% [PCI 44.9%-92.2%]) with remnants at 6 mo had at least 1 angiographic study, demonstrating subsequent complete closure over the course of our study. Of the 4 aneurysms that were never angiographically demonstrated to be occluded, the mean diameter was 19.1 mm and the mean neck size was 14.8 mm.

TABLE 2.

Aneurysm Remnant Cases

Patient Dome Neck   Additional  
number (mm) (mm) 180-d occlusion 3-yr occlusion treatment? 5-yr occlusion 
19.2 12.2 Residual neck Residual neck Residual neck 
23 7.6 Residual neck Complete No imaging 
11.5 6.6 Residual neck Complete No imaging 
25.7 6.25 Residual aneurysm Complete No imaging 
10.9 8.0 Residual aneurysm Complete Complete 
28.0 6.1 Residual aneurysm Complete Complete 
10.1 6.95 Residual aneurysm Complete Complete 
20.25 9.37 Residual neck Complete Complete (MRA) 
14.5 9.4 Residual neck Lost to f/u Lost to f/u 
10a 15.1 5.8 Residual aneurysm (CCF) Complete Y (CCF coil) Complete (MRA) 
11b 12.6 8.5 Residual neck Complete Y (PED) Complete 
12c 10.2 6.9 Residual aneurysm Residual aneurysm Y (PED) Complete 
13 12.2 4.9 Residual neck Complete Y (PED) Complete 
14 16.7 7.2 Residual neck Residual neck Y (PED) Residual neck 
15 26.1 30.4 Residual aneurysm Residual aneurysm Y (PED) Residual aneurysm 
Patient Dome Neck   Additional  
number (mm) (mm) 180-d occlusion 3-yr occlusion treatment? 5-yr occlusion 
19.2 12.2 Residual neck Residual neck Residual neck 
23 7.6 Residual neck Complete No imaging 
11.5 6.6 Residual neck Complete No imaging 
25.7 6.25 Residual aneurysm Complete No imaging 
10.9 8.0 Residual aneurysm Complete Complete 
28.0 6.1 Residual aneurysm Complete Complete 
10.1 6.95 Residual aneurysm Complete Complete 
20.25 9.37 Residual neck Complete Complete (MRA) 
14.5 9.4 Residual neck Lost to f/u Lost to f/u 
10a 15.1 5.8 Residual aneurysm (CCF) Complete Y (CCF coil) Complete (MRA) 
11b 12.6 8.5 Residual neck Complete Y (PED) Complete 
12c 10.2 6.9 Residual aneurysm Residual aneurysm Y (PED) Complete 
13 12.2 4.9 Residual neck Complete Y (PED) Complete 
14 16.7 7.2 Residual neck Residual neck Y (PED) Residual neck 
15 26.1 30.4 Residual aneurysm Residual aneurysm Y (PED) Residual aneurysm 

CCF: cavernous-carotid fistula.

MRA: magnetic resonance angiography.

PED: Pipeline embolization device.

aTarget aneurysm ruptured into the cavernous sinus creating a cavernous-carotid fistula that was occluded with transvenous coils 1 mo after the 180-d visit.

bTwo additional PEDs were placed on the same date as the 1-yr follow-up visit.

cTarget aneurysm was retreated with the PED 1 d after the 3-yr visit.

The prevalence of parent vessel occlusion or in-stent stenosis in the safety cohort at 5-yr is summarized in Table 3. Four of 66 evaluated vessels (6.06%; [PCI 1.68%-14.8%]) had ≥50% stenosis, including 3 with parent artery occlusion (all known at 3-yr follow-up) and 1 nonocclusive (50%-75%) stenosis that progressed from a nonsignificant narrowing noted at 1 and 3 yr. No delayed neurological events were recorded in these patients. Between 3 and 5 yr, no additional delayed parent vessel occlusions were reported. Overall, the number of parent vessel occlusions in our trial through 5 yr was 6 out of 109 treated vessels (5.5%; [PCI 2.05%-11.6%]); all occurring within 3 yr post-treatment.8,11 As seen in Table 4, 5 of these were neurologically silent and 1 resulted in a major ischemic stroke within the 180-d primary outcome assessment window in a patient that was noncompliant with antiplatelet medications.8

TABLE 3.

In-Stent Stenosis Rates in Treated Vessels

 Percentage with In-Stent Stenosisa 
Degree of stenosis 1-yr (n = 91) 3-yr (n = 76) 5-yr (n = 66b
<25% stenosis 93.4% (85/91) 86.8% (66/76) 87.9% (58/66) 
25% to 50% stenosis 1.1% (1/91) 10.5% (8/76) 6% (4/66) 
50% to 75% stenosis 1.1% (1/91) 0 (0/76) 1.5% (1/66) 
>75% stenosis 1.1% (1/91) 2.6% (2/76) 4.5% (3/66)c 
Indeterminate 3.3% (3/91) N/A N/A 
 Percentage with In-Stent Stenosisa 
Degree of stenosis 1-yr (n = 91) 3-yr (n = 76) 5-yr (n = 66b
<25% stenosis 93.4% (85/91) 86.8% (66/76) 87.9% (58/66) 
25% to 50% stenosis 1.1% (1/91) 10.5% (8/76) 6% (4/66) 
50% to 75% stenosis 1.1% (1/91) 0 (0/76) 1.5% (1/66) 
>75% stenosis 1.1% (1/91) 2.6% (2/76) 4.5% (3/66)c 
Indeterminate 3.3% (3/91) N/A N/A 

aBased on total number of intracranial aneurysms at each follow-up.

bIncludes 3 treated patients excluded from the PUFS effectiveness arm as previously reported.7

cAll are complete carotid occlusions.

TABLE 4.

Carotid Occlusions

Patient Time of occlusion Neurological symptoms Headaches Delayed sequelae 
0 to 6 mo None Persistent, resolved with medication None 
0 to 6 mo None Ipsilateral, transient None 
0 to 6 mo None None None 
6 to 12 mo None None None 
6 to 12 mo Stroke None Visuospatial issues 
1 to 3 yr None Ipsilateral retroorbital, persistent Chronic headaches 
Patient Time of occlusion Neurological symptoms Headaches Delayed sequelae 
0 to 6 mo None Persistent, resolved with medication None 
0 to 6 mo None Ipsilateral, transient None 
0 to 6 mo None None None 
6 to 12 mo None None None 
6 to 12 mo Stroke None Visuospatial issues 
1 to 3 yr None Ipsilateral retroorbital, persistent Chronic headaches 
TABLE 5.

Overall 5-Yr Angiographic Outcomes for the PUFS Effectiveness Arm

Outcome n = 106 aneurysms 
Complete occlusion 96 (90.6%) 
Residual aneurysm/neck 4 (3.8%) 
No angiographic follow-up 6 (5.7%) 
Withdrawn 3 (2.8%) 
Dead 3 (2.8%) 
Outcome n = 106 aneurysms 
Complete occlusion 96 (90.6%) 
Residual aneurysm/neck 4 (3.8%) 
No angiographic follow-up 6 (5.7%) 
Withdrawn 3 (2.8%) 
Dead 3 (2.8%) 
TABLE 6.

Clinical Outcomes Based on the Modified Rankin Scale (mRS)

mRS 3 yr 5 yr 
60 62 
19 14 
6a 
No mRS 19 22 
mRS 3 yr 5 yr 
60 62 
19 14 
6a 
No mRS 19 22 

amRS score of 6 assigned to subjects who had died (all deaths occurred within 2 yr).

At the completion of the predefined 5-yr follow-up period, 94.3% of aneurysms in the PUFS effectiveness cohort had at least 1 follow-up catheter cerebral angiogram with a median time of 3.6 yr to the last follow-up angiogram. Overall, 90.6% of aneurysms in the effectiveness cohort had at least 1 catheter cerebral angiogram showing complete occlusion. Table 5 shows the overall angiographic results for the 5-yr study period.

Delayed Neurological Complications and Functional Outcomes

Six occurrences of the primary PUFS safety endpoint of major ipsilateral stroke or neurological death were previously reported.8 Between 180-d and 5-yr follow-ups, no additional patients experienced major ipsilateral stroke, intracranial hemorrhage, or neurological death. There were 4 deaths in our study (3.7% [PCI 1.02%-9.21%]), 3 of which occurred within the first 180 d of treatment.8 One additional non-neurological death was reported at 2 yr post-treatment in a patient with recurrent malignancy.

There were 3 device-related serious adverse events between 180 d and 3 yr post-treatment.11 Between 3 and 5 yr, there were no additional reported serious device-related adverse events. Eighty-one patients completed the mandated 5-yr clinical follow-up (Table 5), 78 of whom (96.3% [PCI 89.6%-99.2%]) had modified Rankin Scale scores of 0 to 2 (Table 6). Between 3 and 5 yr, there were no reported strokes or neurological deaths.

DISCUSSION

Our examination of long-term clinical and angiographic outcomes from the PUFS trial suggests the treatment of large and wide-necked ICA aneurysms with the PED is both safe and effective. Among this cohort of patients harboring large and giant wide-neck aneurysms who underwent a 5-yr angiogram, angiographic occlusion was confirmed in over 95% of subjects with low rates of delayed neurological complications. Importantly, in those subjects with documented aneurysm remnants at 180-d angiography, complete angiographic occlusion rates with or without additional PED treatment increased with time (documented in 11 of 15 aneurysms over the course of follow-up), suggesting that the aneurysm healing process following PED placement occurs progressively. Despite the generally high degree of effectiveness, 5 of 106 aneurysms evaluated in the effectiveness cohort of PUFS required retreatment with PED, with 3 of these progressing to aneurysm occlusion after retreatment. Similarly, 7 of the 9 remnants identified at 6 mo that were not subsequently retreated were shown to be occluded on subsequent angiographic or magnetic resonance studies. The causes of failed treatment can be related on a case-by-case basis to a variety of factors, including (1) poor apposition of the construct to the parent vessel wall, (2) persistent runoff into branch vessels arising from the aneurysm fundus preventing involution of the aneurysm, and (3) continued transmural exchange between the aneurysm and its parent artery due to suboptimal construct scaffolding that is insufficient to support intimal overgrowth of the aneurysm neck.14 This latter problem often complicates aneurysms arising from the outer curvature of the parent artery (where the metallic coverage provided by braided devices is lower than that at nominal opening) or under circumstances in which an inadequate number of devices or oversizing of devices is employed to treat a complex neck aneurysm.15,16

During the post-treatment follow-up period between 180 d and 5 yr, there were no instances of delayed aneurysm recanalization or rupture following complete angiographic occlusion. These findings are especially important in the PUFS population of wide-necked, large, and giant ICA aneurysms, given the poor natural history of these aneurysms and the challenges associated with their treatment by conventional endovascular or open surgical techniques.1,2,17

Other single-center18-23 and small multicenter24 studies have demonstrated high angiographic cure rates with flow diverter (including PED) treatment of ICA aneurysms. Complete occlusion rates in studies reporting 6 to 12-mo follow-up typically range from 70% to 93%. A meta-analysis of 29 studies reporting 1451 patients with 1654 aneurysms treated with flow diverters found complete occlusion rates of 76%. The vast majority of studies in this meta-analysis, however, reported less than 12 mo of follow-up. Studies reporting longer term follow-up generally are limited to 24 mo. One single-center series of 38 aneurysms with medium-term follow-up reported complete occlusion of all 27 ICA aneurysms by 18 mo, with progressively increased occlusion between 3 and 18 mo of follow-up.25 Another multicenter trial of 31 aneurysms reported complete angiographic occlusion in 29 of 30 aneurysms at 2 yr, although half of the aneurysms in this series were adjunctively coiled.24 In their single-center study of 251 large and giant aneurysms undergoing PED treatment, Saatci et al26 reported a 91.2% complete occlusion rate at 6 mo and 94.6% complete occlusion rate at 1 to 2 yr post-treatment, supporting our observation that aneurysm occlusion following PED treatment may occasionally require longer intervals, even beyond 1 yr. This may be especially true in very wide-neck and fusiform aneurysms.

As reported by Becske et al,8 5.6% of subjects enrolled in the safety cohort of the PUFS trial were captured by the primary safety endpoints of major ipsilateral stroke or neurological death. This included 2 patients with parenchymal hemorrhage, 1 with both parenchymal hemorrhage and ischemic stroke, 2 with ischemic stroke only, and 1 presumed neurological death of undefined etiology. Additionally, there were 2 patients with parenchymal hemorrhage resulting in transient minor neurological deficit. Between 180 d and 3 yr, there were 3 additional serious device-related adverse events (2 amaurosis fugax and 1 asymptomatic parent vessel occlusion).11 As reported in this study, no additional major device-related complications, major neurological morbidity, or neurological mortality were observed during the 3- to 5-yr postoperative interval.

A number of retrospective studies have examined the safety of aneurysm treatment with endoluminal devices such as the PED. The International Retrospective Study of the PED examined outcomes of 793 patients with 906 treated aneurysms (including posterior circulation and ruptured aneurysms) and found a neurological morbidity and mortality rate of 8% and a spontaneous parenchymal hemorrhage rate of 2.4%.9 In their meta-analysis, Brinjikji et al27 found procedure-related morbidity and mortality rates of 5% and 4%, respectively, with higher morbidities and mortality seen in the treatment of posterior circulation aneurysms. Parenchymal hemorrhage rates in this meta-analysis were 3%. In a meta-analysis performed by Arrese et al,28 delayed morbidity rates were 2.6% and late mortality rates were 1.3%. However, most patients included in these analyses did not have follow-up beyond 18 mo. Our study demonstrates no major or persistent minor neurological complications of delayed onset between 180 d and 5 yr when treating large and giant ICA aneurysms with PED alone.

Side-branch occlusion of ICU branches covered by a PED construct was not specifically recorded in the prospective PUFS trial but has been investigated in several other series. Angiographic occlusion of a covered ophthalmic artery after Pipeline embolization has been reported as high as 21% with a median of 1 to 2 PEDs.29-31 This resulted in minimal complications, presumably due to the presence of collateral circulation. Angiographic occlusion of a covered anterior choroidal artery after Pipeline embolization was shown to be just 3.4%, with a median of 1 PED covering the anterior choroidal artery.18 Reported occlusion rates of the posterior communicating artery are more widely variable due to variability in the anatomy of the posterior communicating artery.32-37 Factors likely associated with side-branch occlusion include the number of PEDs covering the branch, the flow demand on that branch, and potential collateral circulation.

The frequency and etiologies for delayed aneurysm rupture and parenchymal hemorrhage after flow diverter treatment of intracranial aneurysms are poorly understood. The estimated rate of post-treatment parenchymal hemorrhage is 1% to 5%. Delayed aneurysm rupture is thought to occur in approximately 1% of cases, 9,27,28 typically occurring within the first 6 mo of treatment.38 In our study, there were no aneurysmal subarachnoid hemorrhages. One cavernous aneurysm ruptured after PED placement, creating a minimally symptomatic (pulsatile tinnitus) carotid-cavernous fistula that was discovered at the 6-mo follow-up angiogram.8 Retrospectively, the patient acknowledged the onset of ipsilateral pulsatile tinnitus the day after treatment, suggesting early rupture of the aneurysm.

Parenchymal hemorrhage, when it occurs, generally does so within the first month following treatment39 and is exceedingly rare beyond 6 mo. The underlying cause is generally not known, although foreign-body emboli have been implicated in some cases.40 In our study, 5 parenchymal hemorrhages were observed, all occurring within 30 d of treatment.8 Of the 5 hemorrhages, 3 patients (1 with initial major neurological deficit, 2 with minor neurological deficit) recovered without neurological sequelae, and 2 died (1 due to hemispheric ischemic infarction following cessation of antiplatelet medications to manage the parenchymal hematoma). No cases of delayed intracranial bleeding were reported between 30 d and 5 yr following treatment. Thus, our trial observed a 1.9% (2/107) rate of permanent morbidity/mortality related to intracranial hemorrhage. All patients were maintained on dual antiplatelet therapy postembolization, with aspirin 325 mg daily for at least 6 mo and clopidogrel 75 mg daily for at least 3 mo. Extension of this regimen was neither regulated nor recorded for this study; however, most patients were not continued on antiplatelet medication beyond what was prescribed by the study, except when the patient may have been on such medication for other reasons (ie, cardiac). All recorded instances of postembolization intracerebral hemorrhage occurred within 14 d of embolization, suggesting etiologies related to periprocedural phenomena such as emboli. The lack of subsequent hemorrhagic events indicates to us that longer term dual antiplatelet use, remote from the procedure, is safe. Experience with intracranial and cardiac stents also tells us that the ischemic risk from in-stent thrombosis without adequate antiplatelet therapy is very high. The level of platelet reactivity was unknown, as testing was not required.

Limitations

We acknowledge several limitations to this study. First, approximately 15% of patients did not complete formal long-term follow-up mandated by the study. Thus, it is possible that the long-term angiographic and clinical outcomes may not accurately reflect true underlying success rates in the treated population. However, of the 106 aneurysms enrolled into the effectiveness cohort, only 10 (9.4%) did not have at least 1 angiographic study demonstrating complete aneurysm occlusion over the 5-yr follow-up course.

Because our trial was a prospective single-arm study employing a historical control, direct comparison of PED treatment to parent vessel occlusion and endosaccular coiling is not possible. However, in prospectively limiting enrollment to subjects harboring large (>10 mm) and giant aneurysms with wide necks (>4 mm), we used objective morphological criteria to select aneurysms unlikely to be successfully treated by other reconstructive endovascular or surgical techniques with acceptably low morbidity.

Our study was limited to aneurysms of the ICA from the petrous to superior hypophyseal segments. Thus, the safety and efficacy results of this study may not be generalizable to more distal anterior circulation or posterior circulation aneurysms. Moreover, the median number of PEDs used to treat aneurysms in PUFS was 3. However, the range in the number of devices used to treat each aneurysm and the variability in aneurysm sizes and morphologies included in the PUFS group make it difficult to draw conclusions as to the minimum number of devices necessary for the effective treatment of any specific aneurysm.

CONCLUSION

Endovascular treatment of large and giant wide-necked aneurysms of the ICA with the PED results in high rates of long-term angiographic occlusion with low rates of delayed neurological complications. No delayed recanalization of previously occluded aneurysms was observed. The clinical and angiographic outcomes of treating these aneurysms with the PED are superior to those reported in the literature with other treatment modalities. PED is effective and safe treatment for large and giant proximal ICA aneurysms.

Disclosures

This study was funded by Chestnut Medical and ev3/Covidien (Irvine, California). The PUFS trial was supported and funded by Chestnut Medical and eV3/Covidien. Dr Becske is a consultant/proctor in the use of the PED for Covidien. Dr Kallmes’ institution received a grant from Chestnut Medical and ev3/Covidien for funding of a clinical trial; support for travel to meetings for the study or other purposes from MictroVention; consultancy fees not related to this study from Medtronic and ev3/ Covidien; financial support for grants or pending grants from ev3/Covidien, MicroVention, Sequent, Micrus, and Benvenue Medical for preclinical studies and clinical trials; development funds for educational presentations from ev3/Covidien and CareFusion; and travel/accommodations/meeting expenses unrelated to this study from MicroVention and ev3/Covidien; and he received royalties from the University of Virginia Patent Foundation. Dr Shapiro is a consultant/proctor in the use of the PED for Covidien. Dr Moran is a chief medical officer for Covidien, and he received fees from ev3/Covidien for lectures and for development of educational presentations. Serves as a consultant for and receives honoraria from Medtronic NV. Dr Levy is a board member of Codman & Shurtleff; received consultancy fees from Codman & Shurtleff, ev3/Covidien, and TheraSyn Sensors; is a consultant for Pulsar; serves on the Acute Ischemic Stroke Clinical Advisory Board for Stryker; provided expert testimony for Medical Legal Review; has grants or grants pending from SARIS (Stent-Assisted Recanalization in acute Ischemic Stroke), Codman & Shurtleff, ev3/Covidien, and Boston Scientific; received honoraria and lecture fees from Boston Scientific and Covidien; received funds from stocks or stock options from Intratech Medical, Mynx, Medina Medical, Inc., and Blockade Medical, LLC; is National PI for Covidien's SWIFT Prime Trials; and received fees for carotid stent training from Abbott Vascular. Dr McDougall received consultancy fees from the Covidien Medical Advisory Board. Dr Szikora and his institution received consulting fees or honoraria from Chestnut Medical, and he received consultancy fees from Stryker Neurovascular and ev3/Covidien and speaker's and service fees for being on speakers’ bureaus of Stryker Neurovascular and ev3/Covidien. Dr Lanzino's institution received consulting fees or honoraria from ev3/Covidien and support for travel to meetings for the study or other purposes from ev3/Covidien for presentation to a FDA panel; his institution received an unrestricted educational grant from ev3/Covidien. Dr Woo's institution received funds for grants or for pending grants from Siemens and from MicroVention; he received royalties paid by Codman & Shurtleff for an acute stroke device after acquisition of Revasc and travel/accommodations/meeting expenses unrelated to this study from Siemens; and he is the President/CFO of Vascular Simulations. Dr Lopes is a consultant for Covidien and Stryker; shareholder/ownership in Penumbra, Chestnut Medical, and Blockade Medical; advisory board for Siemens and Medtronic from whom he has also received research grants and for whom he has been a consultant; principal investigator for Liberty Trial (Penumbra). Dr Siddiqui received financial compensation from Codman & Shurtleff for serving on its advisory board and from Codman & Shurtleff, Concentric Medical, ev3/Covidien, Guide Point Global Consulting, and Penumbra for consultancy; he has grants or grants pending from the National Institutes of Health and the University at Buffalo (SUNY); he received fees for lectures and for serving on speakers bureaus from Codman & Shurtleff and Genetech; he received funds for stock/stock options from Hotspur, Intratech Medical, Stim-Sox, and Valor Medical; his institution received funds for travel/accommodations/meeting expenses unrelated to this study from the 44th Neuroendovascular Complication Meeting and Asia Pacific Stroke Conference; he received honoraria from Abbott Vascular, American Association of Neurological Surgeons’ courses, and Genetech; and he has a nonfinancial relationship with Penumbra and serves on the National Steering Committee for 3D Separator Trial. Dr Fiorella received grant support from SAMMPRIS-NIH and consultancy fees from Codman & Shurtleff, ev3/Covidien, NFocus, and Cordis; has grants or grants pending from Siemens, MicroVention; serves as a consultant for Medtronic; and holds stocks or stock options from Vascular Simulators. Dr Saatci's institution received fees for participation in review activities from Chestnut Medical, and she received consultancy fees from ev3/Covidien. Dr Cekirge's institution received financial compensation for review activities from Chestnut Medical, and he received consultancy fees from ev3/Covidien (Medtronic), MicroVention, and Sequent. Dr Berez was an employee of Chestnut Medical and ev3/Covidien and received patent royalties from Covidien; he has stock/stock options in Chestnut Medical and Covidien, and is a founder and employee of Chestnut; he received travel/accommodations/meeting expenses unrelated to this study from Chestnut Medical and ev3/Covidien; and he is the founder and CEO of Chestnut Medical, the company that developed the PED, and is the inventor named on the relevant patents. Dr Cher was an employee of Chestnut Medical and ev3/Covidien during the data collection phase of this study, is currently a consultant for Covidien/Medtronic, and he had stock/stock options in Chestnut Medical and Covidien. Dr Nelson is a consultant/proctor in the use of the PED for Covidien/Medtronic. The other authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.

REFERENCES

1.
Wiebers
DO
,
Whisnant
JP
,
Huston
J
et al
Unruptured intracranial aneurysms: natural history, clinical outcome, and risks of surgical and endovascular treatment
.
Lancet
 .
2003
;
362
(
9378
):
103
-
110
.
2.
UCAS Japan Investigators
Morita
A
,
Kirino
T
et al
The natural course of unruptured cerebral aneurysms in a Japanese cohort
.
N Engl J Med
 .
2012
;
366
(
26
):
2474
-
2482
.
doi: 10.1056/NEJMoa1113260
.
3.
van Doormaal
TPC
,
Van Der Zwan
A
,
Verweij
BH
,
Langer
DJ
,
Tulleken
CAF
.
Treatment of giant and large internal carotid artery aneurysms with a high-flow replacement bypass using the excimer laser-assisted nonocclusive anastomosis technique
.
Neurosurgery
 .
2006
;
59
(
4 suppl 2
):
ONS328
-
34
;
discussionONS334–5. doi: 10.1227/01.NEU.0000233971.08409.F0
.
4.
Whisenant
JT
,
Kadkhodayan
Y
,
Cross
DT
,
Moran
CJ
,
Derdeyn
CP
.
Incidence and mechanisms of stroke after permanent carotid artery occlusion following temporary occlusion testing
.
J Neurointerv Surg
 .
2015
;
7
(
6
):
395
-
401
.
doi: 10.1136/neurintsurg-2014-011207
.
5.
Shapiro
M
,
Babb
J
,
Becske
T
,
Nelson
PK
.
Safety and efficacy of adjunctive balloon remodeling during endovascular treatment of intracranial aneurysms: a literature review
.
Am J Neuroradiol
 .
2008
;
29
(
9
):
1777
-
1781
.
doi: 10.3174/ajnr.A1216
.
6.
Shapiro
M
,
Becske
T
,
Sahlein
D
,
Babb
J
,
Nelson
PK
.
Stent-supported aneurysm coiling: a literature survey of treatment and follow-up
.
Am J Neuroradiol
 .
2012
;
33
(
1
):
159
-
163
.
doi: 10.3174/ajnr.A2719
.
7.
Turfe
ZA
,
Brinjikji
W
,
Murad
MH
,
Lanzino
G
,
Cloft
HJ
,
Kallmes
DF
.
Endovascular coiling versus parent artery occlusion for treatment of cavernous carotid aneurysms: a meta-analysis
.
J Neurointerv Surg
 .
2015
;
7
(
4
):
250
-
255
.
doi: 10.1136/neurintsurg-2014-011102
.
8.
Becske
T
,
Kallmes
DF
,
Saatci
I
et al
Pipeline for uncoilable or failed aneurysms: results from a multicenter clinical trial
.
Radiology
 .
2013
;
267
(
3
):
858
-
868
.
doi: 10.1148/radiol.13120099
.
9.
Kallmes
DF
,
Hanel
R
,
Lopes
D
et al
International retrospective study of the pipeline embolization device: a multicenter aneurysm treatment study
.
Am J Neuroradiol
 .
2015
;
36
(
1
):
108
-
115
.
doi: 10.3174/ajnr.A4111
.
10.
Kallmes
DF
,
Ding
YH
,
Dai
D
,
Kadirvel
R
,
Lewis
DA
,
Cloft
HJ
.
A new endoluminal, flow-disrupting device for treatment of saccular aneurysms
.
Stroke
 .
2007
;
38
(
8
):
2346
-
2352
.
doi: 10.1161/STROKEAHA.106.479576
.
11.
Becske
T
,
Potts
MB
,
Shapiro
M
et al
Pipeline for uncoilable or failed aneurysms: three-year follow-up results
.
J Neurosurg
 .
2016
;
14
(
10
):
1
-
8
.
Epub ahead of print
.
12.
Sahlein
DH
,
Fouladvand
M
,
Becske
T
et al
Neuroophthalmological outcomes associated with use of the Pipeline embolization device: analysis of the PUFS trial results
.
J Neurosurg
 .
2015
;
123
(
4
):
897
-
905
.
doi: 10.3171/2014.12.JNS141777
.
13.
R Core Team
.
R: A Language and Environment for Statistical Computing
 .
1-1. Available at: http://www.r-project.org/. Accessed November 2015
.
14.
Shapiro
M
,
Becske
T
,
Nelson
PK
.
Learning from failure: persistence of aneurysms following pipeline embolization
.
J Neurosurg
 .
2016
:
1
-
8
.
doi: 10.3171/2015.12.JNS152065
.
15.
Shapiro
M
,
Raz
E
,
Becske
T
,
Nelson
PK
.
Variable porosity of the pipeline embolization device in straight and curved vessels: a guide for optimal deployment strategy
.
Am J Neuroradiol
 .
2014
;
35
(
4
):
727
-
733
.
doi: 10.3174/ajnr.A3742
.
16.
Shapiro
M
,
Raz
E
,
Becske
T
,
Nelson
PK
.
Building multidevice pipeline constructs of favorable metal coverage: a practical guide
.
Am J Neuroradiol
 .
2014
;
35
(
8
):
1556
-
1561
.
doi: 10.3174/ajnr.A3902
.
17.
Chalouhi
N
,
Tjoumakaris
S
,
Gonzalez
LF
et al
Coiling of large and giant aneurysms: complications and long-term results of 334 cases
.
Am J Neuroradiol
 .
2014
;
35
(
3
):
546
-
552
.
doi: 10.3174/ajnr.A3696
.
18.
Raz
E
,
Shapiro
M
,
Becske
T
et al
Anterior choroidal artery patency and clinical follow-up after coverage with the pipeline embolization device
.
Am J Neuroradiol
 .
2015
;
36
(
5
):
937
-
942
.
doi: 10.3174/ajnr.A4217
.
19.
Tanweer
O
,
Raz
E
,
Brunswick
A
et al
Cavernous carotid aneurysms in the era of flow diversion: a need to revisit treatment paradigms
.
Am J Neuroradiol
 .
2014
;
35
(
12
):
2334
-
2340
.
doi: 10.3174/ajnr.A4081
.
20.
Puffer
RC
,
Piano
M
,
Lanzino
G
et al
Treatment of cavernous sinus aneurysms with flow diversion: results in 44 patients
.
Am J Neuroradiol
 .
2014
;
35
(
5
):
948
-
951
.
doi: 10.3174/ajnr.A3826
.
21.
Moon
K
,
Albuquerque
FC
,
Ducruet
AF
,
Crowley
RW
,
McDougall
CG
.
Resolution of cranial neuropathies following treatment of intracranial aneurysms with the Pipeline embolization device
.
J Neurosurg
 .
2014
;
121
(
5
):
1085
-
1092
.
doi: 10.3171/2014.7.JNS132677
.
22.
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
.
Am J Neuroradiol
 .
2010
;
31
(
6
):
1139
-
1147
.
doi: 10.3174/ajnr.A2023
.
23.
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
;
discussion642–3–quizN6. doi: 10.1227/01.NEU.0000339109.98070.65
.
24.
Nelson
PK
,
Lylyk
P
,
Szikora
I
,
Wetzel
SG
,
Wanke
I
,
Fiorella
D
.
The pipeline embolization device for the intracranial treatment of aneurysms trial
.
Am J Neuroradiol
 .
2011
;
32
(
1
):
34
-
40
.
doi: 10.3174/ajnr.A2421
.
25.
Briganti
F
,
Napoli
M
,
Leone
G
et al
Treatment of intracranial aneurysms by flow diverter devices: long-term results from a single center
.
Eur J Radiol
 .
2014
;
83
(
9
):
1683
-
1690
.
doi: 10.1016/j.ejrad.2014.05.029
.
26.
Saatci
I
,
Yavuz
K
,
Ozer
C
,
Geyik
S
,
Cekirge
HS
.
Treatment of intracranial aneurysms using the pipeline flow-diverter embolization device: a single-center experience with long-term follow-up results
.
Am J Neuroradiol
 .
2012
;
33
(
8
):
1436
-
1446
.
doi: 10.3174/ajnr.A3246
.
27.
Brinjikji
W
,
Murad
MH
,
Lanzino
G
,
Cloft
HJ
,
Kallmes
DF
.
Endovascular treatment of intracranial aneurysms with flow diverters: a meta-analysis
.
Stroke
 .
2013
;
44
(
2
):
442
-
447
.
doi: 10.1161/STROKEAHA.112.678151
.
28.
Arrese
I
,
Sarabia
R
,
Pintado
R
,
Delgado-Rodriguez
M
.
Flow-diverter devices for intracranial aneurysms
.
Neurosurgery
 .
2013
;
73
(
2
):
193
-
200
.
doi: 10.1227/01.neu.0000430297.17961.f1
.
29.
Puffer
RC
,
Kallmes
DF
,
Cloft
HJ
,
Lanzino
G
.
Patency of the ophthalmic artery after flow diversion treatment of paraclinoid aneurysms
.
J Neurosurg
 .
2012
;
116
(
4
):
892
-
896
.
doi: 10.3171/2011.11.JNS111612
.
30.
Moon
K
,
Albuquerque
FC
,
Ducruet
AF
,
Webster Crowley
R
,
McDougall
CG
.
Treatment of ophthalmic segment carotid aneurysms using the pipeline embolization device: clinical and angiographic follow-up
.
Neurol Res
 .
2014
;
36
(
4
):
344
-
350
.
doi: 10.1179/1743132814Y.0000000322
.
31.
Zanaty
M
,
Chalouhi
N
,
Barros
G
et al
Flow-diversion for ophthalmic segment aneurysms
.
Neurosurgery
 .
2015
;
76
(
3
):
286
-
289
.
doi: 10.1227/NEU.0000000000000607
.
32.
Brinjikji
W
,
Lanzino
G
,
Cloft
HJ
,
Kallmes
DF
.
Patency of the posterior communicating artery after flow diversion treatment of internal carotid artery aneurysms
.
Clin Neurol Neurosurg
 .
2014
;
120
:
84
-
88
.
doi: 10.1016/j.clineuro.2014.02.018
.
33.
Albuquerque
FC
,
Park
MS
,
Abla
AA
,
Crowley
RW
,
Ducruet
AF
,
McDougall
CG
.
A reappraisal of the Pipeline embolization device for the treatment of posterior circulation aneurysms
.
J Neurointerv Surg
 .
2014
;
7
(
9
):
641
-
345
.
doi: 10.1136/neurintsurg-2014-011340
.
34.
Tsang
AC-O
,
Fung
AMY
,
Tsang
FCP
,
Leung
GK-K
,
Lee
R
,
Lui
W-M
.
Failure of flow diverter treatment of intracranial aneurysms related to the fetal-type posterior communicating artery
.
Neurointervention
 .
2015
;
10
(
2
):
60
-
67
.
doi: 10.5469/neuroint.2015.10.2.60
.
35.
Kan
P
,
Duckworth
E
,
Puri
A
,
Velat
G
,
Wakhloo
A
.
Treatment failure of fetal posterior communicating artery aneurysms with the pipeline embolization device
.
J Neurointerv Surg
 .
2015
.
doi: 10.1136/neurintsurg-2015-011959
.
36.
Zanaty
M
,
Chalouhi
N
,
Starke
RM
et al
Failure of the pipeline embolization device in posterior communicating artery aneurysms associated with a fetal posterior cerebral artery
.
Case Rep Vasc Med
 .
2016
;
2016
:
1
-
4
.
doi: 10.1155/2016/4691275
.
37.
Daou
B
,
Valle-Giler
EP
,
Chalouhi
N
et al
Patency of the posterior communicating artery following treatment with the Pipeline embolization device
.
J Neurosurg
 .
2016
:
1
-
6
.
doi: 10.3171/2016.2.JNS152544
.
38.
Kulcsar
Z
,
Houdart
E
,
Bonafe
A
et al
Intra-aneurysmal thrombosis as a possible cause of delayed aneurysm rupture after flow-diversion treatment
.
Am J Neuroradiol
 .
2011
;
32
(
1
):
20
-
25
.
doi: 10.3174/ajnr.A2370
.
39.
Cruz
JP
,
Chow
M
,
O’Kelly
C
et al
Delayed ipsilateral parenchymal hemorrhage following flow diversion for the treatment of anterior circulation aneurysms
.
Am J Neuroradiol
 .
2012
;
33
(
4
):
603
-
608
.
doi: 10.3174/ajnr.A3065
.
40.
Shapiro
M
,
Ollenschleger
MD
,
Baccin
C
et al
Foreign body emboli following cerebrovascular interventions: clinical, radiographic, and histopathologic features
.
Am J Neuroradiol
 .
2015
;
36
(
11
):
2121
-
2126
.
doi:10.3174/ajnr.A4415
.

Acknowledgment

Three- and 5-yr results from the Pipeline for Uncoilable of Failed Aneurysms Trial were previously presented at the 2015 International Stroke Conference in Nashville, TN on February 11, 2015.

COMMENTS

The authors present a focused, well-written 5-yr follow-up analysis of the patients treated in the PED for the initial PUFS study. In this study, the authors present a 95.2% occlusion with 6% retreatment rate and no delayed neurological deficits or hemorrhage. Of those evaluated, 96.3% had a mRS < 2. This manuscript is a logical progression and, indeed, a completion of the original study as designed. Importantly, in addition to demonstrating the high 5-yr occlusion rate of this device, no new recanalization is noted and no new neurological events are described. The data clearly support the use of flow diverters for the treatment of complex and potentially previously untreatable aneurysms. An important observation to make from this is the increased incidence of complete (though seemingly asymptomatic) carotid occlusion (3/66). Though clinically irrelevant, it is an important factor to take into consideration. It will be interesting to note if this incidence shall increase when the next follow-up for this cohort is published as it might provide insight into the long-term need for antiplatelet therapies.

Charles J. Prestigiacomo

Newark, New Jersey

Advancement in neurointerventional surgery has been marked by several breakthrough technologies: Guglielmi detachable coils, balloon microcatheters, self-expandable intracranial stents, and liquid embolic agents. Now added to this impressive armamentarium is the flow diverter. The investigators of the PUFS study reported to us the 5-yr long-term results of this impressive trial. In addition to the gratifying results of long-term occlusion of almost all treated aneurysms (95.2%, including some of the most challenging intracranial aneurysms faced by neurovascular specialists), there is the reassuring finding of no increase of complication beyond the initial 180-d and 3-yr reports. The PUFS trial investigators shall be congratulated for this important success.

Since the initial adoption of flow diverter technology, the neurointerventional surgical community has refined the delivery and deployment techniques, expanded treatment indications, and learned etiologies for complications and treatment failures. Still, many challenges remain and intense research efforts are ongoing. Flow diversional therapy will continue to play an important and increasing role in neurovascular disease management.

Y. Jonathan Zhang

Houston, Texas