Background Occlusion of canine bifurcation and sidewall aneurysms was undertaken with a new liquid embolic agent (PHIL 35) assisted by a high-density partially retrievable stent (FRED) with preservation of the carotid artery.
Methods Three dogs were used as acute preparations for development of the technique and two were used for chronic studies lasting 90 days. In one animal we intentionally did not completely fill the aneurysm to determine the long-term results of incomplete treatment. The degree of occlusion, carotid artery compromise, and dislodgement and/or migration of embolic material in treated aneurysms were assessed.
Results All aneurysms planned for complete obliteration were totally occluded successfully. By design, we partially occluded one aneurysm. In this aneurysm, angiography performed at 30 days revealed less filling, but at 90 days it had persistent small residual filling. We did not detect any distal embolization during the injection and no angiographic occlusions, change in configuration, or delayed migration of the embolic material were found. In the inspected stent, no foreign material was noted. In four animals we successfully removed the stent with preservation of the integrity of the carotid artery. In the fifth we intentionally left both stents deployed.
Conclusions We have developed a new treatment for cerebral aneurysms using a combination of a retrievable stent and a new liquid embolic agent.
- Flow Diverter
- Liquid Embolic Material
Statistics from Altmetric.com
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.
Endovascular or microsurgical treatment of wide-necked cerebral aneurysms, particularly giant and partially thrombosed aneurysms, remains a therapeutic challenge. Direct clipping of such lesions entails considerable risk and a relatively high failure rate of complete occlusion. Endosaccular therapy has a high recurrence rate which requires additional treatment1–3 despite advances in coil technology,4 ,5 balloon remodeling technique,6 and intracranial stents.7 The use of flow diverters can successfully occlude such aneurysms, but requires leaving the device in the parent artery and the use of long-term antiplatelet agents. Flow diverters can take months to achieve total occlusion and can result in delayed hemorrhage.8 Because of these shortcomings, a better endovascular method is needed.
In limited animal studies of wide-necked aneurysms, cyanoacrylate injections have improved the long-term results of occlusion with a lower rate of aneurysm recurrence and thicker aneurysm orifice tissue healing.9 ,10 However, this approach, with or without an aneurysm neck bridge device,11 ,12 has been deemed unsafe in experimental dog models.9 ,10 The use of the liquid embolic agent Onyx (Covidien Neurovascular, Irvine, California, USA), with a variety of protection devices such as balloon catheters, stents, and coils, was also found to be unsafe in an experimental swine model, with a relatively high risk of migration of the liquid embolic agent into the parent artery.13 There are few reports of a combination of Onyx and stent placement, and those that exist were done mainly to stabilize the Onyx in the aneurysm.14 ,15 To our knowledge, the only successful and safe delivery of n-butyl 2-cyanoacrylate (NBCA) in an experimental aneurysm was done with an embolic-containing device in dogs by our team16 but, due to the adhesive properties of NBCA, the embolic-containing device has to be deployed and left in the neck of the aneurysm. This technique is designed for bifurcation aneurysms but is not ideal for sidewall lateral aneurysms.16
A recent report5 (after our experiments) on using a flow-diverting stent to assist in the delivery of microsphere embolization of fusiform and sidewall aneurysms used a high-density device to retain the particles, but the stent had to be left in place.
In view of the failure of experimental obliteration of aneurysms with liquid embolic material, as outlined above, we have created an aneurysm obliteration technique that combines the use of a retrievable high metallic density flow diverter (FRED; Microvention, Tustin, California, USA) and a fast precipitating, non-adhesive liquid embolic agent (PHIL 35). The stent is reliably retrievable when only 80% is deployed. The embolic agent is a precipitating copolymer dissolved in dimethyl sulfoxide (DMSO) with chemically bonded iodine for radio-opacity. We hypothesized that such an approach would improve the success in obliterating aneurysmal filling, and sought to test our hypothesis in a well-established canine aneurysmal model.
Materials and methods
All animal experiments were conducted in accordance with policies set out by the Institutional Animal Care and Use Committee of UCLA Hospital Center Animal Care Facility. Five mongrel dogs (males) were used in this study. The materials and methods of surgical experimental aneurysm creation, endovascular procedures, and pathology handling have been previously described.17 Venous pouch aneurysms were constructed after a Y-type bifurcation was created between the two common carotid arteries and one sidewall aneurysm was created in the internal carotid artery.18–20 Each animal thus had two aneurysms: bifurcation and sidewall (lateral).
We prospectively divided the five dogs into acute and chronic groups. The acute group (n=3, dogs 1, 2 and 3) was used for development of the technique and the chronic group (n=2, dogs 4 and 5) was used to assess the long-term effects of embolic closure. All the animals underwent angiography immediately after aneurysm creation, and after initial treatment the animals in the chronic group also underwent angiography at 30 and 90 days. All animals were euthanized following the guidelines of the Institutional Animal Care and Use Committee of UCLA Hospital Center Animal Care Facility.
Formulation of embolic agent
PHIL35 (35% iodine) with a viscosity of about 90 centipoise was used. Five to seven minutes of diffusion of DMSO results in full hardening. The advantage of PHIL 35 is that, unlike NBCA, it is non-adhesive. This allows for controlled progressive injections, periodic repeat angiography, and stent retrieval when the stent is partially deployed. Compared with Onyx 500, PHIL 35 is more cohesive and precipitation starts almost immediately. This permits a very controlled injection without fragmentation, in addition to the use of iodine for radio-opacity and better visualization of the branches and neck of the aneurysm. However, Onyx 500 has not been tested.
The FRED device is a second-generation high porosity (33%) flow diverter with reposition/removal capability.
As this was a proof of concept study, all aneurysms were surgically created within 24 h of embolization. We accomplished embolization as follows: under general intubation, anesthesia was administered using inhalation agents and intravenous access. The right or left femoral artery was catheterized with a 6 Fr Chaperon sheath (internal diameter 0.071 inch, outer diameter 0.081 inch; Microvention) and heparin (50 units/kg) was administered intravenously. The 6 Fr guide catheter was then advanced 3 cm proximal to the aneurysm neck and selective common carotid artery angiograms were obtained. This initial series of arteriograms21 confirmed the patency of the aneurysms in multiple projections using the best working position for the treatment (figure 1A).
Next, a DMSO-compatible Duo microcatheter (internal diameter 0.165 inch, outer diameter 0.0275 inch; Microvention) was advanced through the guide catheter and the tip of the microcatheter was placed into the dome of the aneurysm (figure 1B). A 0.027 inch delivery catheter (Headway 27, also DMSO-compatible, internal diameter 0.027 inch, outer diameter 0.039 inch) was then advanced beyond the aneurysm for partial deployment of the FRED retrievable stent. Under roadmapping and live fluoroscopy, the retrievable stent was partially deployed with the higher density metal part positioned to cover the neck of the aneurysm (figure 1B). We then performed controlled angiography22 through the Duo microcatheter in the dome of the aneurysm to assess flow dynamics inside the aneurysm (figure 1C). This was done in order to predict the initial pathway that the PHIL would follow (figure 1D–F). The microcatheter was then flushed with 3 ml saline solution and the PHIL 35 embolic agent was slowly injected through the Duo microcatheter and monitored with roadmapping and native fluoroscopy (figure 1D–I). Because of the non-adhesive properties of the PHIL 35, we were able to easily resheath the stent and perform control angiography (figure 1J). Thus, we were able to gradually inject additional liquid embolic agent in a controlled and titrated fashion (figure 1K, L) until we achieved complete aneurysm obliteration (figure 1M). Lastly, with the stent partially deployed for protection, we withdrew the microcatheter from the inside of the aneurysm and then resheathed and removed the stent (figure 1N).
Analysis of angiograms
On follow-up angiograms 20 min after treatment (figure 1O), 30 days after treatment (figure 1P), and 90 days after treatment (figure 1Q), we assessed the degree of aneurysmal filling, stability (ie, movement) of the embolic material, and the status of the carotid artery and compared these parameters with those observed on the angiograms immediately following embolization. As previously described,4 ,6 ,13 we defined angiographic recanalization as a new ‘dog ear’, or as a filling beyond the neck, or as an increase in aneurysmal neck of >10%.
For the bifurcation aneurysms (figure 2A) we first crossed the aneurysm with the 0.027 inch stent delivery catheter and then catheterized the aneurysm. The stent was then partially deployed and, in addition, we gently advanced the partially opened stent to increase the angulations at the level of the neck to better cover the aneurysm opening prior to the injection of the PHIL embolic agent (figure 2B). Once the aneurysm was closed, the microcatheter and stent were removed (figure 2C–E).
Pre-embolization morphological features of aneurysms
All aneurysms (bifurcation and sidewall) in all five animals were patent and without thrombosis on the initial angiogram. Average aneurysm dimensions were 6–8 mm length, 4–6 mm width, and 2–4 mm neck. On the venous pouch in dog 4 the suture at the distal end of the sidewall aneurysm produced a severe stenosis of the carotid artery (figure 3A). Because of this stenosis, we recanalized the artery. We fully deployed the stent and left it in place and then started antiplatelet therapy (figure 3).
We divided our evaluation of treatment effects in the following ways: (1) location (sidewall and bifurcation); (2) complete or partial treatment; and (3) length of follow-up as acute or chronic (ie, 30 and 90 days). All five sidewall aneurysms, both acutely (n=3) and chronically studied (n=2), were totally occluded (figures 1⇑–3). One sidewall (lateral) aneurysm in dog 4 had a severe pretreatment stenosis at the suture line (figure 3A). We were able to open the carotid artery when we totally deployed the stent, and then we proceeded to completely close the aneurysm with PHIL 35. Although the carotid artery thrombosed in the immediate postoperative angiogram, on the 30-day and 90-day follow-up angiograms the carotid artery was patent and the aneurysm remained closed (figure 3B, C).
By design, two bifurcation aneurysms were not treated (dogs 1 and 2) and three were treated (dogs 3, 4, and 5). Once we realized the potential of the technique, we opted to treat the bifurcation aneurysms in dogs 3 and 4, obtaining total occlusion in both (figures 2 and 3). In dog 4 we deployed the stent in the sidewall aneurysm because of the stenosis in the carotid artery. As we were committed to long-term antiplatelet therapy, we decided to deploy the second stent at the level of the bifurcation aneurysm in order to assess the long-term effect of the PHIL 35 and the FRED together. Thus, both the lateral and bifurcation aneurysms treated with total stent deployment and antiplatelet agents remained totally occluded in chronic studies (figure 3C).
We only partially treated one of the bifurcation aneurysms in dog 5 to ensure that there was no interference (chance of embolization towards the distal carotid artery) with the totally occluded sidewall aneurysm (figure 1O), which was the main focus of the experiment. In addition, it permitted us to assess the long-term result of incomplete treatment.
In dogs 4 and 5 from the chronic group, both sidewall aneurysms and one bifurcation aneurysm were totally occluded on the 30-day angiogram compared with the immediate postoperative evaluation (figures 1P and 3B) where a small residual filling was present. The other bifurcation aneurysm (dog 5) showed an improvement in the degree of filling compared with that observed on the initial study (figure 1P). In summary, all small remnants progressed to total or near total occlusion in acute studies (ie, 20 min, figure 2C) and chronic evaluation (ie, 30 days, figure 1O).
In dog 4 there was no recanalization. There was some stenosis of the right internal carotid artery but the artery was patent (figure 3C). Pathological study showed endothelialization at the neck of the aneurysm (figure 3D, E), In dog 5 from the chronic group, the 90-day control angiogram demonstrated some recanalization at the base of both the sidewall and the bifurcation aneurysms (figure 2Q), but pathological examination showed endothelialization in both (figure 2R–U).
Migration of embolic agent
We observed no evidence of migration or change in configuration of the embolic material. After removal of the stent, no foreign material was noted in the stent.
Compromise of the parent vessel by PHIL 35
Angiographic studies showed that all parent arteries were patent without any luminal compromise by the PHIL.
90-day histological analysis
In pathological sectioning, the parent arteries inspected distal or proximal to the aneurysms showed no evidence of migration of the PHIL.
Histological analysis showed that the aneurysm sac varied from mild to marked endothelialization in the stent+polymer group, and from minimal to moderate endothelialization in the polymer only group, without a clear difference between the two groups. Neointimal formation scores tended to be slightly higher in the stent+polymer group. Organized neointima composed of smooth muscle cells and neovascularization (newly formed blood vessels) within a collagen matrix was observed in the aneurysms (bifurcation and sidewall) treated with polymer only (figure 1T, U), while unorganized fibrin was seen within the neointima in two of the three aneurysms (sidewall and sidewall) treated with stent+polymer (figure 3D, E).
In this experimental study we describe a new technique for occluding aneurysms which involves partial deployment of a high-density removable stent and injection of a new non-adhesive liquid embolic agent. The flow diverter mesh density and its modification in the intra-aneurysmal flow, as well as the rapid precipitating qualities of the PHIL, permitted us to contain the liquid embolic agent within the aneurysm. In the past, this has been the most challenging obstacle to the use of liquid embolic agents for aneurysm occlusion.3 ,12 ,15 ,16 ,23–25
Using this technique in both acute and chronic animals, we were able to demonstrate that we can successfully treat both sidewall and bifurcation aneurysms without luminal compromise of the parent vessel or embolic migration.
The canine model we used has been shown to best simulate the hemodynamics of cerebral aneurysms and correlates with the results when using bare platinum (Guglielmi detachable coils) and other coils.15 ,20 The results of these animal studies appear also to correlate with clinical results regarding higher rates of coil compaction and aneurysm recanalization in human aneurysms, whether treated with coils4 ,8 ,10 or liquid embolic agents such as Onyx.26 In previous experimental aneurysm investigations, other devices and aneurysm bridging techniques failed to reliably contain the liquid embolic agent, whether NBCA or Onyx.11 ,13 ,23 The only successful results have been achieved with our embolic-containing device, which had to be detached due to the adhesive nature of embolic NBCA16 and is designed for bifurcation aneurysms.
An example of these limitations was found in a similar experimental study in dogs by Raymond et al.23 In seven dogs, cyanoacrylate embolization was performed: three with an aneurysm bridge device (Trispan), three with two aneurysm bridge devices, and one with an aneurysm bridge device and coils. In one case, left carotid occlusion occurred because of NBCA escape into the parent vessel. In their assessment of this method, they stated that the microcatheter was often glued and histopathological sections showed extruded lipid globules and acrylic material surrounded by a severe granulomatous reaction.23 Despite improved control of glue injection with two aneurysm bridge devices, they concluded that it remains doubtful that this strategy could reach safety levels routinely found in clinical use.23
In another study, Murayama et al13 investigated the use of protection devices and Onyx in the treatment of 40 experimental sidewall aneurysms created in 20 swine. They concluded that protection devices and maneuvers such as flow control, balloon catheter, stent, and coil-assist did not reliably prevent migration of Onyx into the parent vessel (migration rate 9–33%).13 Even when using our embolic-containing device, the use of NBCA makes it necessary to detach the device.16 It is conceivable that, with PHIL 35, we could use the embolic-containing device in terminal and bifurcation aneurysms without the need to detach the device.
As demonstrated in our study, the use of a partially deployed flow-diverting high-density retrievable stent to contain the non-adhesive liquid embolic agent eliminates the need for permanent device placement and the consequent mandatory treatment with antiplatelet agents. The use of the retrievable stent and PHIL 35 provides versatility because it permits controlled monitoring and allows flexibility in the degree and rate of aneurysmal obliteration. The operator can cease or vary the rate of injection of PHIL 35 and obtain an angiogram to determine the degree of aneurysmal filling, potentially partially redeploying the stent and then augmenting the embolic agent until complete obliteration of the aneurysm is achieved.
In the 90-day control, we found some recanalization of the base of both aneurysms, which may be due in part to incomplete filling in the original treatment as we were cautious not to overinject. As we gain experience with the technique, this can be improved.
This study has several limitations including the limited number of animals, the slight size of the necks of the aneurysms, the small size of the aneurysms, and ensuring that there was no distal microembolization. The study was designed as a proof of concept approach. It is shown that this approach is feasible, and the results suggest the possibility that the use of a retrievable stent and PHIL 35 has a high potential for evolving into an effective clinical treatment for cerebral aneurysms. Further studies in a rabbit model are planned using larger aneurysms with wider necks, which will allow us to determine the reliability and durability of the treatment as well as the tissue response of the wall and neck of the aneurysm. In this way we can use MRI with diffusion to study the presence of any cerebral embolization.
The combination of a non-adhesive embolic agent and a retrievable stent provides a safe, easy and reproducible method for the obliteration of aneurysmal sacs. The retrievable stent avoids the need for long-term antiplatelet therapy. Even if the stent is fully deployed and left in the parent artery, the aneurysm is closed at that time. This allows us to avoid the waiting period needed when using flow diverters. The techniques reported in this feasibility study and the proofs of concept represent a potentially major advance in the endovascular treatment of experimental aneurysms.
Dr Richard Winn, Dr Srinivasan Paramasivam, and Erica Berenstein for their collaboration and assistance in the preparation of this manuscript; the R&D team in Terumo, Microvention for their support; and Rob Green, Matt Fitz and Shawn O'Leary in particular.
Contributors AB was involved in formulation of the idea, performance of the experiment, collection of data and preparation of the manuscript.
Funding This work was supported by Terumo Microvention.
Competing interests AB has consulting and royalty agreements with Terumo Microvention.
Provenance and peer review Not commissioned; externally peer reviewed.