Article Text

other Versions

Original research
Treatment of recurrent intracranial aneurysms with the Pipeline Embolization Device
  1. Nohra Chalouhi,
  2. Rohan Chitale,
  3. Robert M Starke,
  4. Pascal Jabbour,
  5. Stavropoula Tjoumakaris,
  6. Aaron S Dumont,
  7. Robert H Rosenwasser,
  8. L Fernando Gonzalez
  1. Department of Neurological Surgery, Thomas Jefferson University and Jefferson Hospital for Neuroscience, Philadelphia, Pennsylvania, USA
  1. Correspondence to Dr L Fernando Gonzalez, Department of Neurological Surgery, Division of Neurovascular Surgery and Endovascular Neurosurgery, Thomas Jefferson University Hospital, 901 Walnut Street, 3rd Floor, Philadelphia, PA 19107, USA; fernando.gonzalez{at}


Background The treatment of recurrent aneurysms after previous surgery or embolization is challenging. Little is known regarding the use of the Pipeline Embolization Device (PED) for recurrent aneurysms.

Objective To analyze the safety and results of PED therapy for recurrent aneurysms.

Methods Fifteen patients with recurrent intracranial aneurysms after previous embolization or surgical clipping were treated with the PED at our institution between 2011 and 2012. Procedural complications and clinical and angiographic outcomes were analyzed.

Results Median aneurysm size was 12 mm. Previous aneurysm treatment consisted of coiling in eight patients, stent coiling in four, a telescoping stent technique in two and surgical clipping in one. Major procedural complications (leaving significant morbidity) occurred in one patient (6.7%) and minor procedural or technical complications (no or minor morbidity) occurred in four patients (26.7%). Fourteen of the 15 patients (93.3%) had a favorable outcome (modified Rankin Scale score 0–2). Of 14 patients with angiographic follow-up, nine (64.3%) had complete aneurysm occlusion (100%), four (28.6%) had near-complete occlusion (≥90%) and only one (7.1%) had incomplete occlusion (<90%). Four of the five patients with less than 100% occlusion at follow-up had a previous stent in situ.

Conclusions Treatment of recurrent aneurysms with the PED appears to be effective, but patients with a previous stent in situ may achieve lower obliteration rates. The morbidity rate associated with PED therapy may be higher than with more standard endovascular techniques using historical data. Larger studies are needed to assess this question better.

  • Aneurysm

Statistics from


Endovascular therapy has emerged as a first-line treatment for an increasing population of intracranial aneurysms. Aneurysm recanalization is considered the major shortcoming of endosaccular coil embolization. The rates of aneurysm recanalization and retreatment after coiling have been reported to be as high as 34% and 9%, respectively.1 Likewise, rebleeding rates are reportedly higher with endovascular treatment than with surgical treatment of intracranial aneurysms.2 ,3 Recent advances in stent technology have improved the durability of endovascular treatment, with several studies reporting lower aneurysm recurrence rates with stent-assisted coiling compared with conventional coiling.4 ,5 Also, endoluminal vessel reconstruction with the Pipeline Embolization Device (PED) has emerged as a highly efficient treatment for intracranial aneurysms.6–9 However, experience with the PED is somewhat preliminary, and some concerns remain regarding the safety of this approach.

As intracranial aneurysms are increasingly managed by endovascular means, the number of patients presenting with aneurysm recurrence is expected to grow rapidly. Aneurysm recurrence can be managed by surgical clipping, coil/stent placement, Onyx HD 500 (eV3, Irvine, California, USA) embolization, vessel sacrifice or flow diversion. There are few data in the literature on the management of recurrent intracranial aneurysms. Moreover, very few studies have specifically reported on the use of PED for treating aneurysm recurrences. The purpose of this study was to analyze the safety and results of using the PED to treat aneurysms recurring after endovascular or surgical treatment.


Fifteen patients with recurrent intracranial aneurysms after a previous embolization or surgical clipping were treated with the PED at our institution between May 2011 and May 2012. Patients were offered PED therapy if they had an aneurysm with complex or wide-necked morphology for which additional coil placement was thought to be challenging or to carry a high risk of recurrence. Patients with fusiform aneurysms, partially thrombosed aneurysms or multiple recurrences were also more likely to undergo treatment with the PED. In patients with posterior circulation aneurysms, the potential benefits as well as the uncertainties accompanying the off-label use of the PED were discussed.


All patients received 75 mg/day clopidogrel and 81 mg/day aspirin for 10 days before the intervention. Platelet function tests were performed on all patients using the aspirin assay and P2Y12 assay (VerifyNow; Accumetrics, San Diego, California, USA). The procedure was performed only if the platelet inhibition level was >30%. Patients with inhibition <30% were reloaded and the assay rechecked. Patients found to be poor responders to clopidogrel were then switched to prasugrel (40 mg loading dose followed by a 5 mg daily maintenance dose). An initial 70 U/kg heparin bolus was administered and activated clotting time was maintained at twice the patient's baseline level intraoperatively. Heparin was discontinued, but not reversed, at the end of the procedure. Patients were maintained on dual antiplatelet therapy for at least 6 months after the procedure followed by aspirin indefinitely.

Procedures were performed under general endotracheal anesthesia and continuous neurophysiological monitoring, including electroencephalography, somatosensory evoked potentials and brainstem auditory-evoked responses. Three-dimensional rotational angiography was performed in all patients and working projections were determined. The PED was accurately sized according to the width of the inflow vessel to avoid any endoleak. The expansion of the PED was documented under fluoroscopy or with additional DynaCT angiography at the operator's discretion. Femoral access was obtained with an 8 F gauge femoral sheath, and a 6 F shuttle sheath (Cook Medical, Bloomington, Indiana, USA) was placed in the distal cervical segment of the vessel leading to the target intracranial vessel, followed by navigation of a 6 F 070 Neuron catheter (Penumbra, Alameda, California, USA). PEDs were then deployed through a Marksman microcatheter (ev3, Irvine) using this triaxial guide catheter system to maximize support during forward loading of the system and to optimize stent opening and apposition. The number of stents deployed was left to the operator's discretion. Overlapping several PEDs was typically considered in fusiform aneurysms or if there was a persistent jet inflow into the aneurysm. Multiple coverage of clinically significant side branches was avoided, especially in the posterior circulation. Inadequate vessel wall apposition was remedied with Hyperglide (ev3 Neurovascular) balloon angioplasty when needed.

Outcome analysis

All procedural technical and clinical complications were documented. Delayed complications were also recorded during follow-up. Clinical follow-up was performed at 1, 3 and 6 months. Clinical outcomes were assessed with the modified Rankin Scale (mRS). Angiographic follow-up was scheduled at 3–6 months, 1 year, 2 years and 5 years after treatment. Aneurysm occlusion at follow-up was categorized as complete (100%), near-complete (≥90%) or incomplete (<90%). In-stent stenosis was graded as none (0–25%), mild (25–50%) or severe (>50%).


Demographics and aneurysm characteristics

Of the 15 patients, 12 were women (80%) and three were men (20%). The mean age of the patients was 54 years (range 31–80 years). The median aneurysm size was 12 mm (range 4.5–25 mm). Thirteen aneurysms were saccular and two were fusiform. Aneurysm locations were as follows: paraclinoid (n=5), cavernous internal carotid artery (n=3), carotid ophthalmic artery (n=3), vertebrobasilar artery (n=3) and middle cerebral artery (n=1).

Previous treatment

Previous aneurysm treatment consisted of coiling in eight patients (53.3%), stent coiling in four (26.7%) patients (Enterprise (Codman Neurovascular, Miami Lakes, Florida, USA) in two and Neuroform (Stryker Neurovascular, Fremont, California, USA) in two), a telescoping stent technique in two (13.3%) patients (two Enterprise and two Neuroform in one patient; two Neuroform in one patient) and surgical clipping in one (6.7%) patient (Figures 1 and 2). Four patients (26.7%) had experienced two aneurysm recurrences prior to PED treatment. The time interval between original treatment and aneurysm recurrence leading to PED retreatment was 48.5 months on average (range 9–84 months). Immediately following original coiling (n=12), there was complete aneurysm occlusion (Raymond grade I) in seven patients, a residual neck (Raymond grade II) in two patients and dome filling (Raymond grade III) in three patients. Two of the 15 aneurysms (13.3%) had previously ruptured. In 14 patients, aneurysm recurrence was discovered on routine follow-up angiography. In one patient, repeat angiography was performed after exacerbation of pre-existing sixth cranial nerve palsy.

Figure 1

(A) Lateral view of digital subtraction angiogram showing a carotid ophthalmic aneurysm that was stent coiled. (B) Follow-up angiogram revealing significant recurrence in the anterior compartment of the aneurysm. (C) Placement of a Pipeline Embolization Device across the neck of the aneurysm. (D) Follow-up angiogram 3 months later showing no further occlusion of the aneurysm.

Figure 2

(A) Frontal view of digital subtraction angiogram showing significant recurrence of a fusiform aneurysm of the middle cerebral artery that was previously treated with a clip-wrap technique. (B) Follow-up angiogram 6 months after placement of a Pipeline Embolization Device showing significant decrease in aneurysm size.

Treatment and outcome

A single pipeline stent was used in 10 patients (66.6%), two stents in four patients (26.7%) and five stents in one (6.7%). No patient had coil placement in addition to the PED. Treatment was successful in all 15 patients. All but one aneurysm (93.3%) showed complete or significant stasis of contrast material at the conclusion of the procedure. Three patients (20%) required balloon angioplasty because of inadequate PED expansion; all three had previously deployed self-expanding stents.

Major procedural complications (leaving significant morbidity) occurred in one patient (6.7%). This patient, a woman who had previously undergone two coil embolizations for a giant right cavernous aneurysm, experienced a right-sided insular intraparenchymal hemorrhage extending into the basal ganglia and temporal lobe on postoperative day 1. A Precise (Cordis, Endovascular, Miami Lakes, Florida, USA) stent was placed for a carotid dissection found on the diagnostic angiogram at the beginning of the procedure. She was left with facial droop and left hemiparesis. There was no angiographic evidence of vessel injury during the procedure, nor did MRI show any source of the hemorrhage. However, the patient was maintained on a heparin drip for 24 h after the procedure before the hemorrhage and platelet inhibition level was noted to be 96%. At follow-up she was able to walk with some assistance and was cleared to return to work.

Minor procedural or technical complications (no or minor morbidity) occurred in four other patients (26.7%). The first patient, with a previously clip-wrapped fusiform middle cerebral artery aneurysm, experienced vasospasm in the M2 segment intraprocedurally that resolved with nicardipine infusion. He underwent placement of two PEDs and remained neurologically intact after the procedure. Three months later, while on dual antiplatelet therapy, he developed right-sided hemiparesis due to an infarct of the left basal ganglia. At his latest clinical follow-up he had almost completely recovered with only minimal residual hemiparesis. There was a significant decrease in aneurysm size at the 6-month angiographic follow-up. The second patient, a woman with a right carotid ophthalmic aneurysm previously treated with a stent-in-stent technique (two Neuroform and two Enterprise), developed headaches, left-sided hemiparesis and blurry vision in her right eye post-procedurally. This patient required balloon angioplasty during the procedure because the PED did not open fully within the previously placed stents. CT and MRI showed no evidence of infarct but fundoscopy revealed a retinal infarct and cotton wool spots. At follow-up she had only mild visual symptoms. The third patient, with a previously coiled paraclinoid aneurysm, developed a focal carotid dissection that was treated medically with anticoagulation. She remained neurologically intact at follow-up. In the fourth patient the PED did not fully expand (40–50%) within the previously placed stent despite balloon angioplasty, and there was no contrast stasis inside the aneurysm. The patient's aneurysm remained open at the 3-month follow-up.

Clinical follow-up was available for all patients at a median of 6 months. Fourteen of the 15 patients (93.3%) had an overall favorable outcome (mRS 0–2). The patient who presented with sixth cranial nerve palsy had recovered completely at follow-up. Two previously asymptomatic patients experienced intermitted headaches during the follow-up period.

Angiographic follow-up was available for 14 of the 15 patients at a median of 7.5 months. One patient with a previously coiled aneurysm has been recently treated and is awaiting follow-up (retreatment with PED was uneventful). Of 14 patients with angiographic follow-up, nine (64.3%) had complete aneurysm occlusion (100%), four (28.6%) had near-complete occlusion (≥90%) and one (7.1%) had incomplete occlusion (<90%). All but one patient had a marked decrease in aneurysm size. Four of the five patients with <100% occlusion at follow-up had previous stent in situ. Of these four patients, two had been treated with a telescoping stent technique and two with stent-assisted coiling (original Raymond grade II for both). As such, only two aneurysms (33.3%) with a previous stent in situ were completely obliterated (100%) at follow-up. One patient (7.2%) had an asymptomatic in-stent stenosis on follow-up angiography; the stenosis was mild and was managed conservatively.



The PED is a dedicated flow diverter that provides up to 35% metal surface area coverage, as opposed to other self-expanding stents such as the Neuroform and Enterprise that provide only around 7–10% coverage. By redirecting flow away from the aneurysm lumen, the PED induces stasis and thrombosis within the aneurysm. The parent vessel is also reconstructed through a process of neointimal tissue growth along the stented segment, which eventually excludes the aneurysm from the circulation and promotes its involution. The PED has emerged as an efficient treatment option for large, wide-necked, fusiform, blister-like and dissecting aneurysms.7 ,10 ,11 PED therapy has also been increasingly employed to treat aneurysms that failed previous endovascular or surgical treatment. Little is known, however, about the safety and efficacy of PED therapy in recurrent aneurysms as very few reports have provided a separate analysis of this aneurysm population.

Treatment efficacy

Nearly 93% of patients in our study had complete or near-complete occlusion at follow-up and most aneurysms significantly decreased in size after PED placement. Likewise, Yu et al12 reported a 0% recanalization rate and a 53.8% occlusion rate of previously treated aneurysms after PED placement. These data suggest that aneurysms recurring after previous treatment may be effectively managed with the PED. On the other hand, recoiling is associated with a significant risk of aneurysm recurrence. Renowden et al13 reported that 11% of recoiled aneurysms developed a second recurrence. Likewise, Sedat et al14 found that 16% of retreated aneurysms showed a second recurrence at follow-up. In a large and rigorous study on angiographic outcomes of coiled aneurysms, Raymond et al15 reported that nearly 50% of aneurysms retreated with coils will go on to recur at follow-up. Despite the encouraging results in the present study, long-term follow-up imaging and comparative studies are needed to better assess the efficacy of the PED in recurrent aneurysms.

Previously treated aneurysms may attain lower occlusion rates at follow-up after PED therapy than previously untreated aneurysms. O'Kelly et al16 identified previous aneurysm treatment as an independent predictor of persistent aneurysm filling after PED treatment. Similarly, in their series on 57 aneurysms treated with the PED, McAuliffe et al17 noted a 92.5% occlusion rate in previously untreated aneurysms compared with 68.7% in previously treated aneurysms, with only three of six cases with a previous stent in situ showing complete occlusion at follow-up. In our series, four of the five patients with <100% aneurysm occlusion following PED therapy had a previous stent in situ. It is therefore possible that a previously deployed stent may result in poor vessel wall apposition of the PED and reduce its hemodynamic effect. Treatment efficacy may thus be substantially reduced in the subgroup of patients with a previous stent in situ.

Of note, if PED therapy is not effective in achieving complete aneurysm obliteration, endovascular access to the aneurysm will have been permanently lost and the only options available for further treatment would be reduced to open surgery or additional placement of PEDs. When surgical clipping is considered, antiplatelet therapy should be discontinued, which may cause delayed thrombosis of the PED, a rare but potentially fatal complication.18 ,19 Also, clip application for proximal control is possible only proximal to the PED because the device is irreversibly deformed by clip application.20 Parent vessel deconstruction with or without extracranial-intracranial bypass is another alternative for recurrent aneurysms after PED therapy.

Special considerations in patients with a previous stent in situ

Six of our patients had a previous stent in situ. Indwelling stents may complicate PED delivery and deployment, and balloon angioplasty may be necessary to achieve adequate wall apposition (as in 50% of cases in the present study). The PED should be deployed distally to the stent in situ and sized accordingly—that is, the length of the device must be sufficient to allow anchoring on normal artery distally and proximally to the stent in situ. During deployment the distal end of the PED may ‘catch’ on the previously placed stent, which may cause anchoring and stretching of the device. Once the stent in situ has been crossed, further deployment is accomplished by unsheathing the device from the microcatheter and advancing the wire to minimize ‘dragging’ the entire construct. Dragging the system will stretch the PED since it is fixed distally on the in situ stent. Balloon angioplasty, if needed, may subsequently be performed to optimize device apposition to the vessel wall. Preserving perforator flow is also an important consideration in these cases because PEDs can have a considerable effect on perforating vessel flow, even exceeding that of three telescoping Neuroform stents.21 We do not have enough data to comment on the difference in deploying a PED within a previously placed open versus closed cell stent.

Safety considerations

Because the risk of rupture from aneurysm recurrence is very low, it is essential that retreatment—whether by flow diversion or recoiling—carries only a small additional burden of morbidity. Serious complications such as distal intracranial hemorrhage, late thrombosis of the construct and delayed aneurysm rupture have been reported with the PED.6 ,19 ,22 In this study, major complications occurred in 6.7% of patients and minor complications in 26.7%. Some complications, however, were not specific and were probably not related to the PED itself. One patient suffered a disabling distal intracranial hemorrhage postoperatively. Angiographic and MRI evaluation revealed no source of the hemorrhage and there were no reports of uncontrolled hypertension postoperatively to explain this complication. The underlying mechanism of such complications remains unknown, but a possible explanation is the hemorrhagic conversion of embolic infarcts in the setting of dual antiplatelet therapy. Another possible explanation is the alteration in pressure dynamics from flow redirection after PED placement.23 Thromboembolic events occurred in two other patients in this study but left no or minor morbidity. Such events are more likely to occur during PED procedures due to the size of delivery devices and the frequent use of multiple stents of high metal surface area coverage. One could argue that many patients treated with PED in the present study had a very low risk of aneurysm rupture and may question whether it is appropriate to expose such patients to the potential morbidity associated with PED therapy. Still, 93% of our patients achieved an excellent outcome following PED treatment of their aneurysms.

Based on the findings of this small series and by comparison with historical data, treatment of recurrent aneurysms with the PED may carry a higher morbidity rate than conventional endovascular techniques. In a multicenter study of 311 patients with recurrent intracranial aneurysms who underwent recoiling, the combined risk of death and permanent major disability was only 1.28%. Likewise, Kang et al24 and Slob et al25 observed no complications during repeat embolization of recurrent aneurysms and suggested that procedural morbidity for retreatment may be lower than for initial coiling. A multicenter study of 100 aneurysms requiring additional coiling because of an enlarging remnant and subtotal occlusion reported minor permanent neurological deficits in only 3% of cases.13 The small size of our study, however, precludes any confident conclusion regarding the safety of PED in this setting. It should also be noted that PED therapy may improve retreatment durability compared with recoiling, which may possibly reduce the number of repeat embolizations required and the associated morbidity.

This study is limited by its sample size and reflects the experience of a single neurovascular center. Comparison with a control group of patients retreated with coils is also lacking. Such a comparison, however, may be prone to bias due to inherent differences in aneurysm characteristics between the two groups.


In this study we analyzed the efficacy and safety profile of PED therapy for aneurysm recurrence and found that recurrent aneurysms may be effectively managed with the PED. Aneurysms with a previous stent in situ, however, may achieve lower obliteration rates at follow-up. Although the rate of favorable outcomes was high in this series, the morbidity rate associated with PED therapy may be higher than with more standard endovascular techniques. Further investigation of this question in larger studies is warranted.


View Abstract


  • Contributors All authors participated in the conception and design of the study, acquisition of data or analysis and interpretation of data; drafting the article or revising it critically for important intellectual content; and approval of the final published version. LFG is the guarantor of the study and takes responsibility for its content.

  • Competing interests None.

  • Ethics approval The paper adheres to all ethical principles and was approved by the Jefferson University Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.