CLINICAL ARTICLE
J Neurosurg 127:81–88, 2017
ABBREVIATIONS CEC = clinical events committee; CRL = core radiology laboratory; CTA = CT angiography; DSA = digital subtraction angiography; ICA = internal carotid
artery; MRA = MR angiography; mRS = modified Rankin Scale; PED = Pipeline Embolization Device; PITA = Pipeline Embolization Device for the Intracranial Treatment of
Aneurysms; PUFS = Pipeline for Uncoilable or Failed Aneurysms.
SUBMITTED February 9, 2015. ACCEPTED June 16, 2015.
INCLUDE WHEN CITING Published online October 14, 2016; DOI: 10.3171/2015.6.JNS15311.
Pipeline for uncoilable or failed aneurysms: 3-year
follow-up results
Tibor Becske, MD,
1,2
Matthew B. Potts, MD,
1,3
Maksim Shapiro, MD,
1,2
David F. Kallmes, MD,
4
Waleed Brinjikji, MD,
4
Isil Saatci, MD,
5
Cameron G. McDougall, MD,
6
István Szikora, MD, PhD,
7
Giuseppe Lanzino, MD,
4
Christopher J. Moran, MD,
8
Henry H. Woo, MD,
9
Demetrius K. Lopes, MD,
10
Aaron L. Berez, MD,
11
Daniel J. Cher, MD,
12
Adnan H. Siddiqui, MD, PhD,
13
Elad I. Levy, MD,
13
Felipe C. Albuquerque, MD,
6
David J. Fiorella, MD, PhD,
9
Zsolt Berentei, MD,
7
Miklós Marosföi, MD,
7
Saruhan H. Cekirge, MD,
5
and Peter K. Nelson, MD
1,3
Departments of
1
Radiology,
2
Neurology, and
3
Neurological Surgery, Neurointerventional Service, NYU School of Medicine, NYU
Langone Medical Center, New York, New York;
4
Department of Neurosurgery, Mayo Clinic, Rochester, Minnesota;
5
Department
of Radiology, Bayindir Hospital, Ankara, Turkey;
6
Department of Neurological Surgery, Barrow Neurological Institute, Phoenix,
Arizona;
7
National Institute of Neurosciences, Budapest, Hungary;
8
Division of Interventional Neuroradiology, Mallinckrodt
Institute of Radiology, Washington University School of Medicine, St. Louis, Missouri;
9
Department of Neurosurgery, Stony Brook
Hospital, Stony Brook, New York;
10
Department of Neurological Surgery, Rush University Medical Center, Chicago, Illinois;
11
Alembic, LLC, Mountain View;
12
Wild Iris Consulting, Palo Alto, California; and
13
Departments of Neurological Surgery and
Radiology, University of Buffalo, Buffalo, New York
OBJECTIVE The long-term effectiveness of endovascular treatment of large and giant wide-neck aneurysms using
traditional endovascular techniques has been disappointing, with high recanalization and re-treatment rates. Flow diver-
sion with the Pipeline Embolization Device (PED) has been recently used as a stand-alone therapy for complex aneu-
rysms, showing signicant improvement in effectiveness while demonstrating a similar safety prole to stent-supported
coil treatment. However, relatively little is known about its long-term safety and effectiveness. Here the authors report on
the 3-year safety and effectiveness of ow diversion with the PED in a prospective cohort of patients with large and giant
internal carotid artery aneurysms enrolled in the Pipeline for Uncoilable or Failed Aneurysms (PUFS) trial.
METHODS The PUFS trial is a prospective study of 107 patients with 109 aneurysms treated with the PED. Primary
effectiveness and safety end points were demonstrated based on independently monitored 180-day clinical and angio-
graphic data. Patients were enrolled in a long-term follow-up protocol including 1-, 3-, and 5-year clinical and imaging
follow-up. In this paper, the authors report the midstudy (3-year) effectiveness and safety data.
RESULTS At 3 years posttreatment, 74 subjects with 76 aneurysms underwent catheter angiography as required per
protocol. Overall, complete angiographic aneurysm occlusion was observed in 71 of these 76 aneurysms (93.4% cure
rate). Five aneurysms were re-treated, using either coils or additional PEDs, for failure to occlude, and 3 of these 5
were cured by the 3-year follow-up. Angiographic cure with one or two treatments of Pipeline embolization alone was
therefore achieved in 92.1%. No recanalization of a previously completely occluded aneurysm was noted on the 3-year
angiograms. There were 3 (2.6%) delayed device- or aneurysm-related serious adverse events, none of which led to
permanent neurological sequelae. No major or minor late-onset hemorrhagic or ischemic cerebrovascular events or
neurological deaths were observed in the 6-month through 3-year posttreatment period. Among 103 surviving patients,
85 underwent functional outcome assessment in which modied Rankin Scale scores of 0–1 were demonstrated in 80
subjects.
CONCLUSIONS Pipeline embolization is safe and effective in the treatment of complex large and giant aneurysms of
the intracranial internal carotid artery. Unlike more traditional endovascular treatments, ow diversion results in progres-
©AANS, 2017 J Neurosurg Volume 127 • July 2017 81
Unauthenticated | Downloaded 08/26/22 05:35 AM UTC
T. Becske et al.
J Neurosurg Volume 127 • July 201782
T
he treatment of complex wide-neck aneurysms of
the internal carotid artery (ICA) remains a chal-
lenge. If left untreated, large and giant aneurysms
of the ICA have a 5-year cumulative rupture risk of up to
40% depending on location of the lesion.
44,45
Surgical op-
tions, including clip occlusion or bypass and trapping, are
associated with a high morbidity rate,
25,38
while previous
endovascular options were associated with high recurrence
and re-treatment rates.
8,24,37
The Pipeline Embolization De-
vice (PED) was a novel endoluminal device for the treat-
ment of intracranial aneurysms introduced for human use
in 2006. Its safety and efcacy were demonstrated in two
prospective trials—the Pipeline Embolization Device for
the Intracranial Treatment of Aneurysms (PITA) trial
26
and
the Pipeline for Uncoilable or Failed Aneurysms (PUFS)
trial.
2
The latter was a multicenter, prospective study of
the PED for the treatment of large and giant aneurysms
of the petrous through superior hypophyseal segments of
the ICA.
2
In the PUFS trial, the primary effectiveness end
point of complete aneurysm occlusion at 6 months with-
out signicant (> 50%) stenosis or use of adjunctive em-
bolization devices was achieved in 73.6% of aneurysms,
while 5.6% of patients met the primary safety end point of
major ipsilateral stroke or neurological death at 6 month.
By 1 year, complete aneurysm occlusion was observed in
86.8% of aneurysms. These results led to Federal Drug
Administration approval for the PED in the treatment of
large and giant aneurysms of the petrous through superior
hypophyseal segments of the ICA. The long-term effec-
tiveness, durability, and safety of the PED, however, are
not known. Here we report the 3-year angiographic and
clinical follow-up results from the PUFS cohort.
Methods
Industry Support and Registration
The PUFS trial was supported by funding from ev3/Co-
vidien. It is registered at clinicaltrials.gov (NCT00777088).
Study Enrollment and Patient Selection
One-hundred eight patients from 10 centers were en-
rolled in this study between November 2008 and July 2009
under a US Food and Drug Administration investigational
device exemption. Complete enrollment criteria have been
previously described.
2
Briey, patients were included if
they had an aneurysm of the petrous through superior hy-
pophyseal segment of the ICA measuring larger than 10
mm in diameter with a neck of at least 4 mm. Exclusion
criteria included subarachnoid hemorrhage within the past
60 days, any intracranial hemorrhage or major surgery
within the last 42 days, a history of abnormal bleeding or
low platelet count, a previously placed stent at the target
aneurysm, a contraindication to CT or MRI, a known al-
lergy to platinum or cobalt and chromium alloys, evidence
of active infection, intolerance to antiplatelet medications,
or signicant stenosis of the ipsilateral carotid artery. Pa-
tients who had undergone prior treatments (i.e., coiling) of
the target aneurysm were not excluded unless a stent had
been previously placed across the neck of the aneurysm.
Baseline Assessments
Prior to placement of the PED, patients underwent a
baseline neurological examination and a detailed neur-
oophthalmological assessment to document the presence
of oculomotor cranial neuropathy or decits in visual acu-
ity and visual elds.
Dual Antiplatelet Therapy
Patients were asked to take aspirin (325 mg per day
orally for 2 days) and clopidogrel (75 mg per day for 7
days or loading dose of 600 mg) prior to PED placement.
After the procedure, patients were asked to take 325 mg
of aspirin daily for at least 6 months and 75 mg of clopi-
dogrel daily for at least 3 months (most were maintained
on combination antiplatelet therapy for 6 months). Platelet
inhibition studies were optional.
Description of the Study Device and Placement Procedure
Placement of the PED was performed as previously
described.
2
In brief, all procedures were performed after
induction of general anesthesia using standard transfemo-
ral approaches. PEDs were deployed through a standard
0.027-in-inner-diameter catheter (Hi-Flo Renegade, Bos-
ton Scientic; or Marksman Catheter, ev3/Covidien). The
size and number of stents placed were selected at the dis-
cretion of each treating operator.
Follow-Up Assessments
Patients underwent repeat neurological examinations
at 30 days, 180 days, 1 year, and 3 years, and angiogra-
phy at 180 days, 1 year, and 3 years after PED placement.
Follow-up imaging was interpreted by an independent
core radiology laboratory (CRL) consisting of 3 neurora-
diologists. Each CRL member independently adjudicated
aneurysms for degree of occlusion
33
(complete occlusion,
residual neck, or residual aneurysm), the presence and de-
gree of in-stent stenosis,
35
and the occurrence of implant
migration.
Safety Reporting
Any negative change in a patient’s health was consid-
ered an adverse event and recorded. Investigators were
asked to judge the relationship of each adverse event to
the PED, the PED placement procedure, and any preex-
sive vascular remodeling that leads to complete aneurysm obliteration over longer-term follow-up without delayed aneu-
rysm recanalization and/or growth.
Clinical trial registration no.: NCT00777088 (clinicaltrials.gov)
https://thejns.org/doi/abs/10.3171/2015.6.JNS15311
KEY WORDS cerebral aneurysm; giant; flow diversion; Pipeline Embolization Device; vascular disorders
Unauthenticated | Downloaded 08/26/22 05:35 AM UTC
Three-year PUFS results
J Neurosurg Volume 127 • July 2017 83
isting conditions. An independent clinical events commit-
tee (CEC) reviewed and adjudicated each serious adverse
event.
Study End Points
The primary effectiveness end point of the PUFS study
was complete occlusion of the target aneurysm without
major (> 50%) stenosis of the parent artery or adjunctive
use of a complementary embolic agent as seen on a 180-
day angiogram and judged by the independent CRL.
2
A
case was considered successful if at least 2 of the 3 core
laboratory members agreed that it met these criteria. In
addition, the primary safety end point of the PUFS study
was the incidence of major ipsilateral stroke (dened as an
increase of ≥ 4 points on the National Institutes of Health
Stroke Scale and present after 7 days as adjudicated by
the CEC) or neurological death within 180 days of PED
placement.
2
We now report the following prespecied secondary
end points observed at the 3-year time point: 1) complete
aneurysm occlusion, 2) incidence of signicant in-stent
stenosis, 3) occurrence of delayed device-related adverse
events, and 4) functional outcomes based on the modied
Rankin Scale (mRS) score.
42
Statistical Analysis
Summary statistics (counts and percentages) were
tabulated. Exact binomial condence intervals for major
study end points were calculated. All statistical calcula-
tions were performed using R.
Results
Baseline characteristics were reported in the initial
PUFS report.
2
Overall patient disposition at the 3-year
time point is shown in Fig. 1. Of the 107 patients with
109 treated aneurysms, there were 4 recorded deaths (3
occurring within the rst 180 days,
2
and 1 additional non-
neurological death occurring at 2 years posttreatment in
a patient with a recurrent malignancy). In addition, 9 pa-
tients were discontinued from the trial before their 3-year
examination (6 subjects were lost to follow-up and consent
was withdrawn in 3 cases). An additional 6 subjects did
not undergo 3-year imaging evaluation. Overall, 85 pa-
tients with 87 aneurysms had 3-year follow-up imaging,
and 85 patients underwent a 3-year neurological examina-
tion. Follow-up imaging modalities included conventional
angiography in 74 subjects (76 aneurysms), MRI or MR
angiography (MRA) in 7 subjects (7 aneurysms), and CT
angiography (CTA) in 4 subjects (4 aneurysms). Five pa-
tients with incomplete occlusions received additional en-
dovascular treatment prior to 3-year follow-up examina-
tion—1 with a postembolization carotid-cavernous stula
was treated underwent aneurysm coiling while the remain-
ing 4 received additional PEDs.
Secondary Effectiveness End Points
Table 1 details the angiographic occlusion rates for the
76 aneurysms in the 74 patients who had a 3-year conven-
tional angiogram. Overall, 71 aneurysms were completely
occluded (93.4%, 95% CI 85.3%–97.8%, p = 5.249e−16).
The CRL could not determine completeness of angio-
graphic occlusion in 1 case (1.3%). Two aneurysm rem-
nants (2.6%) and 2 neck remnants (2.6%) were observed.
In comparison, at 180 days after treatment, there were
6 aneurysm remnants, 8 neck remnants, and 1 carotid-
cavernous stula. Table 2 details the angiographic fate of
those 15 aneurysms at 180-day, 1-year, and 3-year follow-
ups. Of the 5 patients who received additional endovas-
cular treatment, 3 had complete occlusion by 3 years. Of
the 10 remnants that were not re-treated, 7 were occluded
at 3 years (Fig. 2), 2 remained open (1 was subsequently
re-treated), and 1 patient had no angiographic follow-up.
Excluding cases that were re-treated prior to the 3-year
angiogram, complete occlusion in a single session of Pipe-
line embolization was achieved in 68 patients (89.5%). Ex-
FIG. 1. Patient disposition at the 3-year PUFS time point. *Reasons for
lack of 3-year follow-up include patients who withdrew consent, those
who were lost to follow-up, and those who remained in contact with
study coordinators but were unable to undergo a 3-year examination.
F/up = follow-up.
Unauthenticated | Downloaded 08/26/22 05:35 AM UTC
T. Becske et al.
J Neurosurg Volume 127 • July 201784
cluding the case that was re-treated with coils, complete
occlusion was achieved with one or two sessions of PED
alone in 70 patients (92.1%). Of the 11 aneurysms with
imaging other than digital subtraction angiography (DSA)
at 3 years, 9 had had prior angiograms showing complete
aneurysm occlusion, and none had imaging ndings sug-
gesting recanalization.
Table 3 shows the degree of parent artery stenosis in
aneurysms of patients with 3-year angiographic follow-up.
Overall 74 (97.4%) of 76 treated vessels were free of sig-
nicant (> 50%) stenosis 3 years after PED placement.
Secondary Safety End Points
Six occurrences of the primary PUFS safety end point
of major ipsilateral stroke or neurological death were re-
ported in the PUFS report.
2
Between the 180-day and
3-year follow-ups, there were no additional occurrences
of either end point. There were 3 serious adverse events
deemed by the CEC as “probably” or “denitely” related
to the device after the 180-day time point. These included
2 cases of amaurosis fugax and 1 case of a neurologically
silent parent vessel occlusion. None of these resulted in
permanent neurological sequelae.
Functional Outcome
Table 4 summarizes functional outcomes in the safety
arm of the trial. Eighty-ve of these patients underwent
3-year clinical follow-up examinations, and an mRS score
of 0 or 1 was recorded in 80 (95.2%).
Discussion
Based on the results of the multicenter PITA
26
and
PUFS
2
trials, as well as several other multicenter
4,17, 2 0,
28, 47
and institutional
3,10,12,15,16,19,31,32,34,39
experiences, treat-
ment of large and giant ICA aneurysms with the PED is
becoming more widespread. Given that the PED only re-
ceived CE Mark in Europe in 2009 and FDA approval in
the United States in 2011, long-term experience with it is
limited. The PUFS trial is the second-largest prospective
trial of PED for large and giant ICA aneurysms published
to date
47
and the rst to report 3-year angiographic and
clinical outcomes. Complete occlusion rates were dem-
onstrated in 93.4% of the aneurysms with 3-year catheter
angiography studies (up from 73.6% observed at 180 days)
and without any additional cases of major stroke or neu-
rological death.
Other endovascular therapies for large and giant an-
eurysms, including coiling alone or balloon- and stent-
assisted coiling, are associated with high recurrence
rates.
6,24,37
In contrast, the majority of treated aneurysms
in the PUFS cohort that were incompletely occluded at
6 months continued to progressively occlude over time
(Fig. 3). Five aneurysms in the effectiveness arm of the
study required additional intervention with a PED (4 an-
eurysms) or coils (1 aneurysm) between 6 months and 3
years after initial treatment, leading to complete occlusion
in 3 by the 3-year follow-up. In total, 10 of 15 remnants
were completely occluded at 3 years. Importantly, no re-
canalization involving a previously completely occluded
aneurysm was observed at 3 years. In addition, there was
no evidence of recanalization in the 11 patients who un-
derwent non-catheter angiography (CTA or MRA) at 3
years postembolization.
In the initial PUFS report, 5 intracranial hemorrhagic
events were observed before 6 months—2 that met the pri-
mary safety end point and an additional 3 that were mini-
mally symptomatic.
2
All occurred within 1 month of treat-
ment and none were due to rupture of the target aneurysm.
Delayed hemorrhagic events after Pipeline embolization,
either rupture of the target aneurysm or hemorrhage distal
to the site of embolization, have been reported in the liter-
ature.
9,11,13,29,40,43
In the PUFS cohort, no additional cases of
intracranial hemorrhage were observed between 1 month
and 3 years, suggesting that the risk of delayed hemor-
rhage dramatically decreases after the early postembo-
lization period. Overall, there have been no instances of
target aneurysm rupture in the PUFS cohort at any time
postembolization.
To date, this is the largest cohort of patients treated with
the PED to have 3-year clinical and angiographic follow-
TABLE 2. Long-term angiographic outcomes of patients with
incompletely occluded aneurysms
Patient
No.
FU Time Point
Additional
Treatment180 Days 1 Yr 3 Yrs
1 Residual neck Residual neck Residual neck No
2 Residual neck Complete Complete No
3 Residual neck Residual neck Complete No
4 Residual An* Complete Complete Coil*
5 Residual neck Residual An Complete PED
6 Residual An Residual An Residual An No†
7 Residual neck Residual An Complete PED
8 Residual An Complete Complete No
9 Residual neck Residual neck Residual neck PED
10 Residual An Complete Complete No
11 Residual An Complete Complete No
12 Residual An Residual An Residual An PED
13 Residual An Residual An Complete No
14 Residual neck Complete Complete No
15 Residual neck Residual neck Lost to FU No
An = aneurysm; FU = follow-up.
* The target aneurysm ruptured into the cavernous sinus, creating a carotid-
cavernous stula that was coil-occluded 1 month after the 180-day examination.
† This patient was re-treated after the 3-year examination.
TABLE 1. Angiographic occlusion rates at 180 day and 3 years
after treatment
Occlusion Type 180 Days (n = 106*) 3 Yrs (n = 76†)
Complete occlusion 81 (76.4%) 71 (93.4%)
Residual neck 8 (7.5%) 2 (2.6%)
Residual aneurysm 6 (5.7%) 2 (2.6%)
Cannot determine 3 (2.8%) 1 (1.3%)
* 106 aneurysms in 104 patients.
† 76 aneurysms in 74 patients; 5 patients underwent additional endovascular
treatment prior to the 3-year angiogram.
Unauthenticated | Downloaded 08/26/22 05:35 AM UTC
Three-year PUFS results
J Neurosurg Volume 127 • July 2017 85
up. It is also one of the largest cohorts of large and giant
aneurysms managed with any treatment modality to have
3-year clinical and angiographic follow-up. Of the original
107 treated patients, an mRS score was known at 3 years
in 89. This included 4 deaths, 3 of which were considered
neurological in etiology. Of the 85 patients examined at 3
years, 94% had an mRS of 0 or 1. Such long-term results
further conrm the safety of Pipeline embolization for
large and giant ICA aneurysms.
It is important to emphasize that the PUFS results can-
not necessarily be generalized to small aneurysms
6,18
or
aneurysm in other locations for which the PED has been
used, including the ICA terminus,
27
middle cerebral ar-
te r y,
5,46,48
anterior cerebral artery, and posterior circula-
tion.
1,4,7,10,14,17,20–23,28,30,36,41
The incidence of large and giant
aneurysms in these locations is relatively rare compared
with that in those included in PUFS and the experience
using the PED to treat such aneurysms is therefore more
limited. Future prospective studies of Pipeline emboliza-
tion for these indications will be required.
Conclusions
In summary, ow diversion with the PED is a safe and
effective treatment of large and giant ICA aneurysms with-
TABLE 3. Stenosis of the parent artery on 3-year angiogram
Stenosis Grade No. of Lesions (n = 76*)
0%–25% 66 (86.8%)
>25%–50% 8 (10.5%)
>50%–75% 0
>75% 2 (2.6%)
* 76 aneurysms in 74 patients.
FIG. 2. Delayed closure of a large dural cave aneurysm. This 75-year-old woman presented with diplopia and was found to have
bilateral cavernous segment ICA aneurysms. A giant, right-sided aneurysm was treated with stent-assisted coiling prior to the
development of the PED. She subsequently elected to have the remaining left-sided aneurysm treated with Pipeline embolization.
Pretreatment oblique (A) and lateral (B) DSA studies demonstrate this large cavernous segment aneurysm. 3D rotational angiog-
raphy revealed an aneurysm measuring 1.15 cm in maximal dimension with a 6.65-mm neck (C). This aneurysm was treated with 3
overlapping PEDs without complication, and stagnation within the aneurysm dome was immediately evident (D). Six-month follow-
up oblique (E) and lateral (F) angiographic studies showed a small residual neck. A slight reduction in the size of this residual
neck was observed on 1-year follow-up angiography (oblique [G] and lateral [H] projections), which was notable for only a double
density in the lateral projection (H, arrow). Finally, no residual aneurysm or neck was detected on 3-year follow-up angiography
(oblique [I] and lateral [J] projections), with no subsequent treatments after the initial Pipeline embolization.
TABLE 4. Overall 3-year functional outcomes
mRS Score No. of Patients
0 60
1 20
2 2
3 1
4 2
5 0
6 4
Unknown* 18
* No 3-year mRS recorded.
Unauthenticated | Downloaded 08/26/22 05:35 AM UTC