Article Text

Download PDFPDF

Estimating the proportion of intracranial aneurysms likely to be amenable to treatment with the pipeline embolization device
  1. Waleed Brinjikji,
  2. Harry J Cloft,
  3. David Fiorella,
  4. Giuseppe Lanzino,
  5. David F Kallmes
  1. Department of Radiology, Mayo Clinic, Rochester, Minnesota, USA
  1. Correspondence to Dr David F Kallmes, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905, USA; kallmes.david{at}


Objective A study was undertaken to estimate the fraction of intracranial aneurysms that might be amenable to treatment with the pipeline embolization device (PED), a current flow diverter device, and to determine the types of aneurysms that are probably not amenable to treatment with this device.

Methods A retrospective analysis of 200 consecutive intracranial aneurysms as seen on three-dimensional rotational angiography images was conducted. Based on aneurysm and parent artery morphology, four independent observers experienced in the use of the PED judged the likelihood that a given aneurysm could theoretically be treated with one or more PED using a 5-point scale: grades 1 and 2 indicated amenability to treatment with multiple devices; grade 3 indicated amenability to treatment with one device; and grades 4 and 5 indicated the aneurysm would not be amenable to treatment with the PED. Aneurysms were analyzed on the basis of anatomical factors only. Rupture status was not considered. Interobserver agreement was determined.

Results 94 of 200 aneurysms (47%) were judged to be likely to be amenable to treatment with the PED; 50 cases (25%) were judged to be amenable to the use of multiple overlapping devices while, in 44 cases (22%), coverage with a single device was preferred. There was a significant relationship between the location of the aneurysm and the grade of the aneurysm (p<0.0001).

Conclusion On the basis of anatomical configuration, nearly half of the aneurysms in this study were likely to be amenable to treatment with the PED.

  • Aneurysm
  • guidewire
  • history
  • face
  • flow diverter
  • artery
  • spinal cord
  • thrombectomy
  • navigation
  • complication

Statistics from

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.


Endoluminal flow diversion has been shown in preclinical and clinical studies to achieve high rates of aneurysm occlusion with acceptable safety profiles.1–3 One such flow diverter is the pipeline embolization device (PED), a tubular stent-like device that is placed within the parent, across the aneurysm neck, to achieve aneurysm occlusion.1 While the PED represents an intriguing new device for the treatment of aneurysms, important questions remain regarding the applicability of the device to ‘typical’ aneurysms. As currently constructed, the devices will probably have limitations based on aneurysm and parent artery morphology. Some aneurysms, including those of sidewall morphology, might be readily and safely treated with the PED. The efficacy of the device to treat other types of aneurysms, particularly those of bifurcation morphology or those in the vicinity of important perforating arteries, is currently unknown.

We sought to estimate the fraction of aneurysms that might be amenable to treatment with the PED. In the context of this ‘virtual’ experiment, we evaluated a consecutive series of aneurysms presenting to our institution and rated them using a 5-point scale developed to reflect the likely amenability of the lesions to flow diversion treatment. This newly developed rating scale incorporated both aneurysm morphology and the anatomy of the regional branch arteries and perforating vessels to gauge whether the PED would be appropriate treatment for a given aneurysm.

Materials and methods

Patient population

Following Institutional Review Board approval, a retrospective analysis of 179 patients with 200 intracranial aneurysms seen at our institution between 2005 and 2007 who underwent three-dimensional rotational angiography (3DRA) for evaluation of previously untreated aneurysms was conducted.

Angiographic technique

Typically, 5F or 6F catheters were placed into the internal carotid arteries or vertebral arteries. Biplane digital subtraction angiographic images of the entire circulation were usually performed, followed by ‘working projection’ biplane digital subtraction angiographic images based on views identified on 3DRA (see below). Working projection images were those images that offered ideal separation between the aneurysm neck and parent artery. Small fields of view, usually 5×7 inches, were used for these working projection views.

Three-dimensional rotational angiography (3DRA)

All of the 3DRA examinations were performed using a biplane C-arm angiographic unit (Integris Philips Medical Systems, Best, The Netherlands) with a field of view of 7 inches and a frame rate of 30 f/s. Images were acquired with a head-end propeller C-arm orientation at a rotational speed of 55 degrees/s covering +120 to −120 degrees. A volume of 16 ml non-ionic contrast medium was injected through a 5–6F catheter using an injector with a velocity of 4 ml/s. Image acquisition was started 1–3 s after injection of contrast material. The acquisition time of images was 4.4 s. Volume-rendered 3D images were reconstructed with a 100% magnification (a field of 37.56 cm2) and a matrix of 256 pixels3 using the 3DRA volumetric measurement of the system software. The threshold for volume-rendering image was fixed as the default value provided by the software.

Analysis of images

After 3DRA images were acquired, a database of the images from the aneurysms included in this study was created. Each aneurysm was presented with an anteroposterior, lateral and working projection for the readers to rate the amenability to treatment with the PED. In addition, the maximum dimension and location of the aneurysm were presented to each independent observer (all with at least 15 years of experience). Amenability to treatment was graded on a 5-point scale. For this study we made assumptions that multiple telescoped devices would offer added efficacy over a single device, but that a single device would be used if the anticipated span of the PED would have covered eloquent perforating arteries. The rating was performed independent of any knowledge of the rupture status of the aneurysms.

The grading scale was as follows:

  • Grade 1: sidewall aneurysms along vessels in which the intended treatment expanse includes no branch vessels, so multiple devices were considered appropriate.

  • Grade 2: sidewall aneurysms along vessels in which the anticipated treatment expanse includes branch vessels that could be occluded without substantial risk such as the ophthalmic artery, so multiple devices could probably be implemented.

  • Grade 3: sidewall aneurysms along vessels in which the anticipated treatment expanse includes branch vessels that would pose a significant risk of neurologic deficit such as the anterior choroidal artery, so a single device would be recommended.

  • Grade 4: sidewall aneurysms along vessels with maximum dimension <3 mm, so not amenable to the current PED.

  • Grade 5: non-sidewall morphology, so not amenable to the PED.

We defined aneurysms of grades 1–3 as ‘very likely amenable to treatment with the PED’ and grades 4 and 5 as ‘possibly not amenable to PED treatment’.

Statistical analysis

Using the statistical package JMP (, weighted κ values were calculated to rate interobserver variability. χ2 tests were implemented to evaluate whether factors such as parent artery location significantly influenced amenability to treatment with the PED. 95% Confidence Intervals (95% CI) were also calculated.


Patient population and aneurysm characteristics

The patient population consisted of 44 men and 135 women. The average age of the patients was 59±13 years and the average aneurysm size was 6.6±3.2 mm. Overall, 73 aneurysms (37%) were present on the internal carotid artery (ICA), 19 aneurysms (9.5%) were located on the middle cerebral artery (MCA), 46 (23%) were on the anterior cerebral artery or anterior communicating artery (ACA/ACOM), 18 (9%) were on the posterior communicating artery (PCOM), 26 (13%) were located on the basilar tip and 18 aneurysms (9%) were located in the posterior circulation but not on the basilar tip.

Aneurysm amenability to treatment

On average, the four reviewers determined that 47% (95% CI 43% to 50%) of aneurysms would have been likely to be amenable to treatment with flow diverter devices (ie, grades 1–3); 25% (95% CI 22% to 28%) would have been amenable to treatment with multiple devices (grades 1 or 2) and 22% (95% CI 19% to 25%) would have been amenable to treatment with a single device (grade 3). These data are summarized in table 1. Figure 1 shows examples of aneurysms graded 1–5 by the readers.

Table 1

Percentage of aneurysms by grade

Figure 1

Grading of aneurysms for amenability to treatment with flow diverter. (A–C) Internal carotid artery aneurysms grades 1, 2 and 3 that were unanimously determined to be amenable to treatment by the four readers; (D) grade 4 posterior inferior cerebellar artery aneurysm and (E) grade 5 basilar tip aneurysm.

There was a significant relationship between the location of the aneurysm and the amenability to treatment as determined by our grading system (p<0.0001). By location, 90% of ICA aneurysms, 79% of PCOM artery aneurysms, 7% of aneurysms on the basilar tip, 43% of aneurysms on the vertebrobasilar system excluding the basilar tip, 5% of aneurysms on the MCA and 6% of aneurysms on the ACOM and along the distal ACA were considered amenable to treatment with the PED. These data are summarized in table 2.

Table 2

Aneurysm location and amenability to treatment with pipeline embolization device

κ values were calculated rating the interobserver variability between pairs of observers. Overall, the κ values between observers ranged from 0.71 to 0.83, indicating good to very good agreement. The average κ value was 0.77, indicating that there was good agreement between observers.


In this study we have shown that, based on anatomical considerations alone, approximately one-half of aneurysms presenting to our institution would probably have been amenable to treatment with currently approved flow diversion devices. In almost half of these ‘amenable’ aneurysms we would have been very comfortable using multiple telescoped devices. Owing to the proximity of perforating arteries, for the remainder of the ‘amenable’ aneurysms we would have opted to treat with a single device. As expected, the location of the aneurysm is a significant predictor of amenability to PED treatment, with ICA aneurysms being highly amenable and other locations less so.

To our knowledge, no previous study has attempted to determine the fraction of aneurysms that would be likely to be amenable to treatment with flow diverting technology. Our results predict that, even in its current form, the PED might be widely applied to the treatment of typical unruptured intracranial aneurysms. This study also highlights a potential limitation of this device in that aneurysms arising from the basilar tip, ACA/ACOM and MCA in general may not be amenable to treatment with this device.

Several case series detailing use of the PED are available, yet none of these previous studies has evaluated the proportion of ‘typical’ aneurysms that might be amenable to treatment with the device. Fischer et al reported a series of 101 patients of which one-third harbored fusiform rather than saccular aneurysms, yet they offer no details on specific aneurysm location except that about 80% were in the anterior circulation.1 Lylyk et al reported a clinical series using the PED in which 85% were used in the ICA.4 The vast majority of these would probably have been in our ‘amenable’ group as they represented aneurysms mostly proximal to the PCOM. Similar proportions of ICA aneurysms were treated in other recent series.5 ,6 Our study provides an insight not available in the current literature since previous studies have only reported actual use rather than the likely proportion of typical aneurysms that would be amenable to treatment.

Devices that use braided technologies for construction of intrasaccular implants have recently been described.7 ,8 These braided ovoid devices probably work through similar mechanisms to endoluminal flow diverters but, given their morphology, are likely to be useful for the treatment of many aneurysms not amenable to treatment with endoluminal flow diversion devices.

Limitations of the study

In order to maximize our sample size, both ruptured and unruptured aneurysms were evaluated and graded and information regarding rupture status was not taken into account in our grading scale. Ruptured aneurysms would probably not be amenable to the PED because aspirin and clopidogrel would be needed with PED therapy and the use of anticoagulation medications is relatively contraindicated in cases of ruptured aneurysm. In the present study we are assuming that the anatomical characteristics of these ruptured aneurysms would not be expected to differ from the unruptured aneurysms in the group.

The main limitation of this study is its theoretical nature. This study was posed as a virtual experiment with four independent observers examining a database of 200 aneurysms and determining which aneurysms would very likely be amenable to treatment with the PED. While we have determined which aneurysms could potentially be treated with the PED, the question of whether the aneurysms in this study would be better treated with the PED than with clipping or endovascular coiling is one that cannot be answered until further studies are produced demonstrating which patients and aneurysms would benefit most from PED treatment.

Lastly, at the time of this experiment the smallest available PED had a diameter of 3 mm, but the current PED has a diameter of 2.5 mm. This may lead to a decrease in the number of grade 4 aneurysms in our study and an increase in the number of aneurysms that are amenable to treatment with this device.


This study serves as a means of providing a general consensus among four interventionalists regarding the potential role for the PED in endovascular aneurysm therapy. We have demonstrated that the PED could theoretically be used in the endovascular treatment of about half of typical saccular intracranial aneurysms. The likelihood of amenability to treatment has a very strong relationship with the location of the aneurysm, with >75% of aneurysms located on the ICA or PCOM probably being amenable to treatment with the PED compared with <10% of aneurysms located in the MCA, ACOM/ACA and basilar. Further studies are needed to determine which patients would benefit most from PED treatment.



  • Competing interests WB and HJC have no competing interests. DF consults (Pipeline proctor) for ev3 Codmanand Nfocus. He received institutional research grant funding and is an unpaid consultant for Microvention. GL receives unrestricted educational grants from ev3 and serves on the advisory board for Actelion Pharmaceuticals. DFK receives research support from ev3, Microventions Inc, Chestnut Inc, and Micrus Endovascular Inc.

  • Ethics approval Ethics approval was provided by IRB at Mayo Clinic, Rochester, MN.

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

Linked Articles

  • PostScript
    Joshua A Hirsch Ronil Vikesh Chandra Thabele M Leslie-Mazwi
  • PostScript
    David F Kallmes Harry J Cloft David Fiorella