Article Text

Original research
Flow diversion of bifurcation aneurysms is more effective when the jailed branch is occluded: an experimental study in a novel canine model
  1. R Fahed1,
  2. J C Gentric2,
  3. I Salazkin1,
  4. G Gevry1,
  5. J Raymond1,2,
  6. T E Darsaut3
  1. 1Laboratory of Interventional Neuroradiology, Centre Hospitalier de l'Université de Montréal, Notre-Dame Hospital Research Centre (CRCHUM), Montreal, Quebec, Canada
  2. 2Department of Radiology, Centre Hospitalier de l'Université de Montréal (CHUM), Notre-Dame Hospital, Montreal, Quebec, Canada
  3. 3Department of Surgery, Division of Neurosurgery, University of Alberta Hospital, Mackenzie Health Sciences Centre, Edmonton, Alberta, Canada
  1. Correspondence to Dr Tim E Darsaut, University of Alberta Hospital, 8440—112th Street, WMC 2C3.83, Edmonton, Alberta, Canada T6G 2B7; tdarsaut{at}


Background Flow diverters (FDs) are increasingly used for bifurcation aneurysms. Failure of aneurysm occlusion may be caused by residual flow maintaining patency of the jailed branch along with the aneurysm. Our aim was to test whether endovascular occlusion of the jailed branch could improve efficacy of flow diversion of bifurcation aneurysms.

Materials and methods Sixteen wide-necked lingual–carotid artery bifurcation aneurysms were created in eight canines. Patent aneurysms were randomly allocated 4 weeks later to flow diversion combined with jailed branch occlusion using coils and/or Onyx (n=6) or flow diversion alone (n=8). Angiographic results of aneurysm occlusion at 3 months were scored using an ordinal scale. Pathology specimens were photographed and neointimal coverage estimated using a semiquantitative scoring system.

Results Fourteen aneurysms were patent at 1 month. FD deployment was successful in all cases but, at 3-month follow-up, three devices had prolapsed into the aneurysm. None of the bifurcation aneurysms treated with FD alone were occluded at 3 months. Endovascular branch occlusion combined with flow diversion significantly improved aneurysm occlusion rates compared with flow diversion alone (median angiographic scores 2 vs 0: p=0.0137). Flow-limiting parent vessel stenosis was not observed in any arteries. Devices were covered with thick neointima in most cases, but patent aneurysms were associated with leaks or holes in the neointima covering the aneurysm neck.

Conclusions Treatment failures following flow diversion of bifurcation aneurysms can be caused by persistent flow to the jailed branch. Branch occlusion combined with flow diversion may improve angiographic occlusion scores of a canine bifurcation aneurysm model.

Statistics from


The role of flow diversion in the treatment of intracranial aneurysms remains to be defined.1 A recent survey has shown that clinicians favor flow diversion of aneurysms with lateral wall morphologies rather than other types of lesions, including bifurcation aneurysms.2 However, some bifurcation aneurysms have been treated successfully by flow diversion.2–8

One factor suggested to inhibit aneurysm occlusion, at least in preliminary animal and Computational Fluid Dynamics (CFD) studies, is the continued flow to the branch jailed by the flow diverter (FD).3 ,9–12 In the setting of poor collateral circulation, jailing of the branch may lead to ischemic complications.13 However, in circumstances where robust collaterals exist, branch occlusion may promote aneurysm occlusion. The effect of intentional deconstruction of one of the branches of the bifurcation on aneurysm occlusion rates following flow diversion has not been formally studied in animal models. We endeavored to test the hypothesis that branch occlusion could improve angiographic outcomes of flow diverted experimental bifurcation aneurysms.

Materials and methods

Surgical aneurysm creation

This report was prepared using the ARRIVE guidelines for animal research.14

Protocols for animal experimentation were approved by the Institutional Animal Care Committee in accordance with guidelines of the Canadian Council on Animal Care. All procedures were performed in 22–32 kg mongrel canines under general anesthesia. Wide-necked bifurcation aneurysms15 were created at the bifurcation of the carotid and lingual artery in eight animals, combining a previously published method to produce wide-necked carotid bifurcation aneurysms with a bilateral lingual aneurysm model to form a novel animal model.16 Briefly, through a paramedian vertical cervical incision, the left external jugular vein was harvested, inverted to remove valves, and placed in heparinized saline until vein pouch creation. The incision was closed in multiple layers over a separately tunneled drain, which was left for 48 h. Further rostrally, an additional paramedian vertical incision was carried down to expose the distal carotid artery at the level of the lingual artery. Temporary clips were applied, and an arteriotomy created on the medial aspect of the distal carotid artery, carried down to incorporate the lingual artery takeoff within the lateral aspect of the aneurysm fundus, just distal to the aneurysm neck. The harvested external jugular vein was divided in half, symmetric linear venotomies of appropriate length fashioned, and the vein pouch anastomosed to the carotid artery and lingual artery takeoff using continuous 7.0 Prolene suture (Ethicon, Cincinnati, Ohio, USA). The end of the vein pouch was secured and temporary clips were removed (figure 1). Any bleeding sites were suture repaired and the procedure was then repeated on the contralateral carotid artery to form bilateral wide-neck lingual artery aneurysms. Further details regarding animal anesthesia, housing, husbandry conditions, and welfare are available on request.

Figure 1

Schematic detailing the surgical aneurysm construction whereby the lingual artery branch origin is incorporated into the body of the aneurysm (A–C), with three-dimensional angiographic image of typical final aneurysm morphology obtained (D).

Endovascular treatment

Endovascular treatment was performed at least 4 weeks after surgical aneurysm construction through a coaxial microcatheter system introduced by a percutaneous transfemoral approach. Aneurysm and arterial dimensions were directly measured using a Siemens Leonardo workstation using Syngo software. Aneurysms were randomly allocated to treatment with FD with or without lingual branch occlusion. Randomization was performed by drawing lots. An Echelon-10 microcatheter (Covidien/Medtronic, Dublin, Ireland) was navigated into the lingual artery and platinum coils of diameter 2–4 mm and lengths 4–6 cm were used to occlude the lingual artery. For the last three such aneurysms, coils were combined with an injection of 0.1–0.2 mL liquid embolic (Onyx; Covidien/Medtronic). FDs were then deployed through a Marksman microcatheter (Covidien/Medtronic) bilaterally from the distal to the proximal carotid artery to span each aneurysm neck with adequate proximal and distal landing zones (5–10 mm), as in clinical practice. The FDs used were the Pipeline Embolization Device (PED; gifts from Covidien/Medtronic), 3.75–4.75 mm in diameter and 18–25 mm in length. The size of FD was selected according to the diameter of the proximal landing zone of the recipient vessel (3.5–4.5 mm in this model).

Antiplatelet regimen

Animals were treated with daily aspirin 325 mg and clopidogrel 75 mg for 4 days prior to flow diversion, according to previous experience.11 Clopidogrel therapy was discontinued 10 days post-PED implantation while aspirin was continued until sacrifice.


Transfemoral angiography was performed in all animals immediately before and after PED deployment, and immediately prior to euthanasia at 3 months. To prevent femoral hematomas on dual antiplatelet therapy, all punctured femoral arteries were surgically exposed through a small linear incision and ligated. Angiographic results were scored using an ordinal system modified from Kamran et al.17 A score of 0 indicated no change in aneurysm volume with treatment, a score of 1 indicated residual contrast filling >50% of the pre-treatment aneurysm volume, 2 indicated residual contrast filling <50% of the pre-treatment aneurysm volume, 3 indicated residual filling confined to the neck region, and a score of 4 indicated no residual filling (complete occlusion). The parent vessels were carefully checked at the proximal and distal landing zones for the presence of stenosis.

Sacrifice, photography, pathology

Euthanasia by barbiturate overdose was performed at 12 weeks. After fixation in 10% formalin for at least 7 days, the body of the aneurysm was sectioned longitudinally, making an attempt to capture the exiting lingual artery branch in the section. Specimens were photographed using a computerized imaging system (Vision PE, Clemex Technologies, Montreal, Canada). The carotid artery aneurysm construct was then opened along the longitudinal axis of the device with microsurgical dissecting tools under 4.5× loupe magnification. After the FD had been cut, the edges of the cylindrical device were opened and the edges were fixed and mounted to a stage to permit further magnified photography.

The degree of neointimal coverage of each device at the level of the aneurysm neck was scored independently by two observers (TED and JR) according to a semiquantitative scale using percentiles from 0 to 100% coverage (in 10% increments), with disagreements resolved in a consensus session as previously described.11 ,18


Aneurysm dimensions were compared across the two groups with t tests and aneurysm occlusion scores were compared with the Kruskal–Wallis test for non-parametric data. The significance level was set at 0.05.


Aneurysm creation

At initial angiography, 2/16 aneurysms were found to be occluded (prior to any treatment) and were excluded from subsequent analysis. Aneurysm characteristics for the treatment groups are presented in figure 2 and table 1. Aneurysm and neck dimensions were similar (p=0.6 and p=0.3, respectively).

Table 1

Aneurysm and flow diverter characteristics per treatment group

Figure 2

Schematic demonstrating the two groups of experimental wide-necked bifurcation aneurysms treated with (A) flow diverter (FD) alone and (B) FD with occlusion of the lingual branch.

Endovascular treatment

FDs were successfully deployed in all cases, with adequate landing zones and without technical difficulties. Coiling of the lingual artery did not lead to immediate complete occlusion in the first 2/3 animals. In these two animals, some flow was seen entering small arterial branches originating proximal to the coil pack, as well as continuing slowly through the coil mass. Coiling combined with Onyx led to immediate branch occlusion in all three cases.

Angiographic results

Angiographic results of flow diversion with or without branch occlusion are presented in table 2. After 3 months, 3/14 (18%) of FDs were found to have prolapsed into the aneurysms (1 in the FD with branch occlusion group, 2 in the FD alone group). These technical failures were excluded, leaving 11 aneurysms for final analysis. Aneurysms allocated to flow diversion plus branch occlusion had improved aneurysm occlusion scores at 3-month follow-up (median score 2) compared with aneurysms treated with FD alone (median score 0; p=0.0137). Typical angiographic results are presented in figure 3. The two aneurysms in the branch occlusion group with small branches originating proximal to the coil mass were occluded at 3 months (occlusion scores of 4 and 3).

Table 2

Angiographic and pathologic results at 3 months

Figure 3

Angiographic images showing similar aneurysms prior to treatment (A, C). (B) Typical 3-month follow-up results after treatment with a flow diverter (FD) alone, with a widely patent aneurysm. (D) Treatment with FD as well as branch occlusion with coils (arrow) led to complete aneurysm occlusion.

Stenosis within the body of the FD ranging from 10% to 30% was observed in 4/11 (36%) parent vessels (table 2); in no case was the stenosis flow-limiting.

Pathology results

All five aneurysms in the group treated with FD and branch occlusion had at least 80% of FD struts covered in dense neointima (figure 4). Four of the six aneurysms in the group treated with FD alone had at least 80% dense neointimal coverage of the FD struts while the other two aneurysms had only 10–20% coverage of the stent struts. There was no statistical difference in neointimal coverage between groups (p=0.28). In aneurysms where the lingual artery was occluded, the neointima at the level of the lingual artery origin was indistinguishable from that overlying the rest of the device at the level of the aneurysm ostium. However, when the lingual artery was kept patent, pores in the neointima were typically found immediately overlying the branch artery origin (figure 5).

Figure 4

Macroscopic photographs of experimental aneurysms treated with flow diverter (FD) and branch occlusion (A–C) or with FD alone (D–F). (A) Sagittal section showing the Onyx-filled branch (arrow) and aneurysmal cavity (*) with mature thrombus. (B, C) En-face view of device struts covering the aneurysm ostium with complete or near-complete neointimal coverage. (D) Sagittal section of aneurysm treated without branch occlusion showing a patent branch origin (*) exiting an aneurysm devoid of thrombus or organized tissue. (E, F) Neointimal coverage of aneurysms treated with FD alone (no branch occlusion), with large area of either bare or minimally covered stent struts and mostly immature (non-organized) tissue (arrow).

Figure 5

(A) Cross-section of an intentionally occluded lingual artery after treatment with coils and flow diversion showing complete obliteration of the vessel lumen. (B) Incomplete neointimal coverage of the flow diverter struts covering the origin of an unoccluded lingual artery.


This work demonstrates that flow diversion of bifurcation aneurysms may fail, at least in this model, but that intentional occlusion of one branch of the bifurcation can promote aneurysm occlusion.

This finding confirms our previous exploratory work in canines on mechanisms potentially associated with success or failure of flow diversion, where the anastomosis of an efferent branch from the apex of the fundus of two lateral wall aneurysms negated the effects of flow diversion,11 which was found to be a successful treatment for aneurysms without an efferent branch. It similarly confirms an experiment performed in a large fusiform aneurysm model, where persistent branch flows were associated with fully patent aneurysms whereas animals in which branches had been surgically clipped had improved angiographic occlusion scores.10 The small numbers of pilot animals included in those studies did not permit firm conclusions to be drawn. By using this new wide-necked lingual bifurcation model with bilateral aneurysms, we were able to formally demonstrate the influence of branch occlusion in a difficult aneurysm model using fewer animals.

The possible link between branch flow and aneurysm occlusion rates is difficult to extract from the clinical literature, as the series published to date have reported branch and aneurysm occlusion separately.19–22 It has been suggested that the presence of an adequate collateral supply to the territory of the branch covered by the FD may favor safe, usually asymptomatic, branch remodeling or occlusion.5 ,7 Slow branch remodeling after flow diversion of bifurcation aneurysm, where a major vessel is jailed with eventual development of robust collaterals, may occur without complication, but the risk of symptomatic ischemia of this approach remains undefined.5 ,21

In select anatomical circumstances and where robust collaterals exist, intentional occlusion of one limb of a bifurcation may promote durable occlusion of a bifurcation aneurysm.23 It should be emphasized that a strategy that intentionally occludes a normal arterial segment should be reserved for difficult aneurysms, as intentional arterial occlusion will pose a risk of ischemic neurological deficits, including that from perforating arteries originating from that segment. Occasionally, attempts to occlude an arterial segment are not completely successful and a small amount of flow persists. The two aneurysms with maintained flow through small arterial branches proximal to the coil pack were still completely or near completely occluded, suggesting that a small amount of residual flow can still be consistent with good results.

Animal studies can provide information relevant to clinical use. With animal models, a variable of interest (in this case branch occlusion) can be manipulated while keeping other variables constant. Varying one factor at a time permits study of the impact of that variable on the treatment outcome. A bilateral model allows comparison of two aneurysms while keeping all other factors (including individual biological variations such as coagulation and platelet function) constant.

The model has several limitations. In this experiment, flow-diverted lingual bifurcation aneurysms had high rates of device prolapse (3/14). In this model, aneurysms are constructed with very wide necks that may require longer landing zones. Furthermore, aneurysms located distally on the carotid artery under the angle of the canine mandible may be subjected to movements associated with deglutition and jaw opening, in addition to movements of the neck.

A small number of aneurysms were studied. It is possible to safely occlude the lingual arteries over a relatively long segment. Although recently introduced devices have been reported to facilitate safe vessel occlusion of human arteries over a short segment,24 ,25 the safety of deconstructive approaches in these circumstances has not been fully elucidated. A longer period of follow-up time after treatment may have resulted in improved occlusion scores, but this was not studied in this work. Cervical carotid aneurysm models differ from spontaneous intracranial aneurysms, and canine and human biology are different; any extrapolation of these results to the clinical realm must be cautious.


In this animal model, flow diversion of wide-necked bifurcation aneurysms did not lead to aneurysm occlusion by 3 months. When flow diversion was combined with intentional occlusion of the jailed branch, aneurysm occlusion scores were significantly improved.


View Abstract


  • Contributors JR and TED designed the experiments; JR, TED, IS, RF and JCG performed the experiments; JR, TED, IS, RF, JCG and GG analyzed and processed the data and drafted the manuscript. All authors revised the draft manuscript and gave final approval of the version to be published. JR and TED agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

  • Funding This work was supported by a research grant delivered by the Heart and Stroke Foundation of Alberta (grant number GIA-0200296) to TED. RF is the recipient of a research scholarship delivered by the Fondation pour la Recherche Médicale (FRM), Paris, France (grant number DEA20140630151).

  • Competing interests None declared.

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

  • Data sharing statement Unpublished/raw data may be available to academic researchers on a per request basis to the corresponding author.

  • Animal studies Principles of laboratory animal care (NIH publication No 86-23, revised 1985) were followed, as well as the guidelines of the Canadian Council on Animal Care. The protocol for animal experimentation was approved by the Institutional Animal Care Committee of our institution (CIPA).

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.