Bioresorbable Flow Diverters for the Treatment of Intracranial Aneurysms: Review of Current Literature and Future Directions

Abstract

The use of flow diverters is a rapidly growing endovascular approach for the treatment of intracranial aneurysms. All FDA approved flow diverters are composed of nitinol or cobalt-chromium, which will remain in the patient for the duration of their life. Bioresorbable flow diverters have been proposed by several independent investigators as the next generation of flow diverting devices. These devices aim to serve their transient function of occluding and healing the aneurysm prior to being safely resorbed by the body, eliminating complications associated with the permanent presence of conventional flow diverters. Theoretical advantages of bioresorbable flow diverters include 1) reduction in device induced thrombosis, 2) reduction in chronic inflammation and device induced stenosis, 3) reduction in side branch occlusion, 4) restoration of physiological vasomotor function, 5) reduction in imaging artifacts, and 6) use in pediatric applications. Advances made in the similar bioresorbable coronary stenting field highlight some of these advantages and demonstrate the feasibility and safety of bioresorbable endovascular devices in the clinic. The current work aims to review the progress of the bioresorbable flow diverter field, identify opportunities for further investigation, and ultimately stimulate the advancement of this technology.

Introduction

Intracranial aneurysms (IAs) are estimated to be present in ~5% of the U.S. population. The use of flow diverters (FDs) is a rapidly growing endovascular approach to treat IAs, as they have demonstrated relatively good safety profiles and high rates of IA occlusion[1]. Consequently, ~30% of IAs in the U.S. are now treated with FDs. Several FDs have achieved FDA approval, with more expected to come on the market in the near future[2]. Recent advances have focused on bio- and hemocompatible coating technologies[3], with several such devices approved for clinical use in Europe[4]. Multiple independent investigators have proposed that the next major advancement in FD technology will be bioresorbable flow diverters (BRFDs), which may provide numerous performance enhancements compared to current devices.

Characteristics of Clinical Flow Diverters

FDs are metal mesh stents that are implanted in the parent artery over the neck of the IA. They function by diverting a majority of blood flow past the aneurysm, promoting thrombus formation within it, and ultimately resulting in its occlusion. In the meantime, the FD acts as a bridge for endothelialization and neointimal growth over the neck, walling the aneurysm off from blood flow[5]. All market-approved FDs are composed of Nitinol or cobalt-chromium alloys with platinum, tungsten, and/or tantalum incorporated into the device for radio-opacity[2]. These devices will remain in the patient for the duration of their lives. Since the radial force required for FDs is much less than coronary stenting, they can be constructed with thinner struts, with diameters ranging between 30-50 μm[2]. Critical FD design characteristics include porosity (percent area of the device wall not covered by struts) and pore density (number of open cells/pores within the FD wall per area), with optimal values reported at 60-75% and 18-32 pores/mm², respectively[6–8]. Clinical FD characteristics are included in Table 1.

Table 1.

Characteristics of FDA approved flow diverters.

Device	Company	Primary Composition	Strut Thickness (um)	Radio-Opaque Marker	Porosity (%)	Pore Density (pores/mm2)	Device Dimensions (Diameter x length (mm))	Balloon Required	Microcatheter compatible	Refs
Pipeline Embolization Device	Medtronic	36 cobalt chromium wires	28-33	12 Platinum-tungsten wires	65-70	~13-14	2.5-5 x 10-35	No	Yes	[2] [7] [9]
Surpass Evolve	Stryker	52 cobalt chromium wires	25-32	12 Platinum-tungsten wires	70	15-30	2-5 x 12-50	No	Yes	[2]
FRED	Microvention	48 wire inner layer, 16 wire outer layer, nitinol	23-36 (inner) 51-56 (outer)	Tantalum markers	50-60	20	3.5-5 x 13-45	No	Yes	[2] [10]

Complications of Clinical Flow Diverters

While FDs represent an advancement in the endovascular treatment of IAs, there are several drawbacks. First, the risk of thromboembolism remains high relative to intrasaccular devices, resulting in the requirement of dual antiplatelet therapy (DAPT). However, even after DAPT, acute thromboembolism associated complications can occur at rates upward of 5%[11]. Very late thrombosis (>1 year) and associated remote ischemic events have also been reported[12]. Second, neointimal hyperplasia and stenosis of the parent artery is frequently reported following FD deployment, which may result in delayed ischemic stroke[13]. Another drawback to FD treatment is the occlusion of perforating branch arteries near the IA. Clinical studies have reported perforator and branch occlusion rates of 3% due to FD deployment, which may result in downstream infarction and neurological deficit[1].

Bioresorbable Flow Diverters

BRFDs are emerging as the next generation of devices for the treatment of IAs. The ideal BRFD that occludes and heals the aneurysm is harmlessly resorbed by the body, eliminating the risk of complications associated with the permanent presence of contemporary devices. While the concept of BRFDs is relatively new, bioresorbable coronary stents have been extensively studied, with many different device iterations and materials making their way into the clinic. Insights from these studies can be leveraged in the design of BRFDs and showcase some of the advantages of bioresorbable endovascular devices. Potential advantages of BRFDs over conventional permanent FDs include:

1. Reduced risk of acute and chronic device induced thrombosis: The innate surface properties of many bioresorbable materials are less thrombogenic than conventional materials, reducing the risk of acute device thrombosis[14, 15].

2. Reduction of chronic inflammation, neointimal hyperplasia, and ultimately induced stenosis: As with bioresorbable coronary stents, the temporary nature of BRFDs may reduce these long term biological responses to the implant [16].

3. Reduction of side branch occlusion: The incidence of FD induced occlusion of side branches remains persistent over time[17]. The progressive and ultimate complete resorption of BRFDs may mitigate the occurrence of late (>1 year) side branch occlusion.

4. Restoration of physiological vasomotion: Clinical studies of bioresorbable coronary stents have demonstrated a return in vasomotor function following device resorption[18].

5. Reduction in imaging artifacts: Follow up CT and/or MR imaging is used to evaluate FD performance. The innate properties of many bioresorbable materials reduce imaging artifact relative to conventional FD materials[19]. Additionally, progressive reduction in device volume shows promise to further reduce image artifact over time.

6. Pediatric applications: The temporary nature of BRFDs will not interfere with the growth of the patient and their vasculature. Although the use of FDs to treat IAs are not as common in pediatrics, many case studies exist[20].

Motivated by these potential advantages (Figure 1), many groups have begun manufacturing and testing bioresorbable devices for the treatment of IAs, with most publications occurring over the past several years.

Figure 1:

Theoretical advantages of bioresorbable flow diverters. A) Reduction in device induced thrombosis and embolism B) Reduction in side branch occlusion C) Reduction in foreign body response and neointimal hyperplasia D) Reduction in metal induced image artifact. Created with BioRender.com

Review of Present Bioresorbable Flow Diverter Literature

Detailed descriptions of device characteristics are compared in Table 2 for all reviewed BRFDs. Experimental results are summarized in Table 3 for all reviewed BRFDs that were investigated in vivo.

Table 2.

Characteristics of the experimental flow diverters reviewed in section 3.

Author and Device	Year	Device Material	Strut Thickness (um)	Radio-Opaque Marker	Porosity (%)	Pore Density (pores/mm^2)	Braid Angle (degrees)	Dimensions (Diameter x length (mm))	Fully Bioresorbable	Microcatheter compatible	Balloon Required	Ref
Wang et al. BRFD	2013	22 Nitinol Wires, 24 Polyglycolic acid wires	~45*	Two parallel platinum containing struts	50-60	12-20	~40*	3.5 and 4.5 mm diameters, length not provided	No	Yes	No	[21]
Wang et al. control FD	2013	22 Nitinol Wires	~25-30*	Two parallel platinum containing struts	80-85	3-5	~45	3.5 and 4.5 mm diameters, length not provided	Fully permanent	Yes	No	[21]
Nishi et al. BRFD	2019	48 Poly-L-lactic acid wires	40-45	Three radio-opaque gold markers at both device ends	~40*	NA	~40*	4 x 10, 4 x 12, and 4 x 15	Yes	Yes	Yes	[22]
Jamshidi et al. BRFD	2020	44/4. 46/2, and 48/0 Poly-L-lactic acid/tantalum coated nitinol wires	50	Tantalum coated nitinol wires	60	17	~50*	4 x 10-20	No	NA	No	[15]
Tidwell et al. BRFD	2021	Polycaprolactone	350	BaSO4 Coating	65	0.87	NA	5.5 x 15	Yes	NA	NA	[23]
Gruter et al. BRFD	2019	Magnesium alloy coated in poly-L-lactic acid	120	None	Can’t be determined	NA	NA	2.5 x 6	Yes	NA	Yes	[24]
Gruter et al. BRFD + coil	2019	BRFD relative to a control FD (described below) performance was investigated for use in stent assisted coiling. The coil used for both conditions was a target helical ultra-coil (Stryker, Kalamazoo, Michigan, USA) of 2cm length and 2 mm diameter										[24]

^*Indicates that the value was estimated using ImageJ to take measurements on published figures. “NA” means the given metric cannot be measured for the device. “Can’t be determined” means that the metric exists for the device, but figures that could be used to calculate the metric were not provided in the original manuscript.

Table 3.

Summary of experimental results for reviewed BRFDs investigated in vivo. OCT: optical coherence tomography

Author and Device	Study Model	Study Duration (Sample Size)	Aneurysm Occlusion Rates			In Device Thrombosis			Device Resorption		Ref
Gruter et al. Control FD + coil	2019	Cobalt-Chromium alloy	Can’t be determined	None	Can’t be determined	NA	NA	2.5 x 9	Fully permanent	NA	Yes	[24]
Wang et al. BRFD	Rabbit elastase induced aneurysm	6 weeks (n = 6) and 3 months (n = 7)	67% (4/6) at 6 weeks 83% (5/6) at 3 months			One case of device induced thrombosis that resulted in the complete occlusion of the parent artery at 3 months			Residual PGA struts were observed at 6 weeks but appeared to be completely resorbed by 3 months via stereomicroscope		[21]
Wang et al. control FD	Rabbit elastase induced aneurysm	3 months (n = 7)	0% (0/6)			No observed device induced thrombosis by gross or histological evaluation			NA		[21]
Nishi et al. BRFD	Rabbit elastase induced aneurysm	1 (n = 5), 3 (n = 5), 6 (n = 5), and 12 (n = 3) months	0% (0/5) at 1 month, 20% (1/5) at 3 months, 50% (2/4) at 6 months, and 33% (1/3) at 1 year			No observed downstream arterial thrombus formation or occlusion via angiography. By OCT, thrombus was observed in 1.9%, 1.8%, 0.0%, and 0.0% of frames at 1, 3, 6, months and 1 year, respectively. All thrombi detected by OCT were small white thrombi.			In vitro testing demonstrated 15%, 45%, 83%, and 95% reductions in weight average molecular weight at 3 months, 9 months, 1 year, and 1.5 years, respectively.		[22]
Jamshidi et al. BRFD	Rabbit infrarenal abdominal aorta	1 month (n = 3)	NA			By SEM analysis, some thrombus was observed on the abluminal side of struts that were not opposed well to the vessel wall.			No reported analysis		[15]
Gruter et al. BRFD	Rat aortic sidewall aneurysm	4 weeks (n = 26), serial OCT out to 6 months (n = 8)	85% (22/26) at 4 weeks			No reported analysis			Qualitative microCT and OCT demonstrated progressive resorption		[24]
Gruter et al. BRFD + coil	Rat aortic sidewall aneurysm	4 weeks (n = 7)	100% (7/7)			No reported analysis			NA		[24]
Gruter et al. Control FD + coil	Rat aortic sidewall aneurysm	4 weeks (n = 6)	83% (5/6)			No reported analysis			NA		[24]

Polymer Based Bioresorbable Flow Diverters

Wang et al. (2013) were the first to pioneer the concept of BRFDs[21]. Their devices featured a backbone composed of 22 braided nitinol wires plus 24 bioresorbable polyglycolic acid (PGA) wires to improve porosity and pore density. Device performance was evaluated in a rabbit elastase induced aneurysm model. PGA containing BRFDs were implanted over the neck of the aneurysm for either 6 weeks or 3 months. Devices composed of only the nitinol backbone served as the control group and were implanted over aneurysm necks for 3 months. Angiography demonstrated successful flow diversion in all PGA containing devices, but only in 57% of the control devices immediately after deployment. No aneurysms were completely occluded by 6 weeks for the PGA containing devices. However, by three months, 83.3% of the PGA containing devices caused complete aneurysm occlusion compared to 66.7% for the controls. This suggested that BRFDs must maintain structural integrity for >6 weeks to allow the aneurysm to occlude and heal. Partial resorption of the PGA struts was observed at 6 weeks and was completed by 3 months. Neointimal thickness decreased from 6 weeks to 3 months for the PGA devices, suggesting that strut resorption may reduce neointimal size and lumen occlusion in the parent artery. However, at 3 months the neointimal thickness was slightly higher for the PGA containing group than the control group which may be attributed to the increased density of struts in the PGA device. Immunohistochemistry demonstrated a moderate to low amount of CD68+ macrophages around the metal struts and partially degraded PGA struts at 6 weeks, but none were observed where PGA struts had completely resorbed by 3 months. In conclusion, Wang et al. were the first to demonstrate that the incorporation of bioresorbable wires in their FD could successfully contribute to flow diversion and aneurysm occlusion prior to their complete and safe resorption.

Nishi et al. (2019) were the first to develop fully bioresorbable FDs, which were composed of 48 braided poly-L-Lactic acid wires[22]. In vitro experiments suggested that the BRFDs would completely resorb after ~1.5 years. Device performance was evaluated in the rabbit elastase induced aneurysm model. Due to the absence of a self-expanding metal back bone, the devices were fully expanded with balloon angioplasty to ensure proper wall apposition. All devices were deployed successfully and follow up time points ranged from 1 to 12 months. Aneurysm occlusion rates never exceeded 50% throughout the study. The authors attributed the poor aneurysm occlusion rates to excentric strut distribution at the aneurysm neck due to the curvature of the parent vessel. No thrombosis was observed via angiography throughout the study. Optical coherence tomography demonstrated very low degrees of device thrombosis throughout the study, and that neointimal thickness increased out to 3 months but then progressively decreased out to 12 months. Immunostaining demonstrated a moderate amount of CD68+ inflammatory and proliferative ki67+ cells around the struts that decreased over time. SEM analysis demonstrated that most struts were completely covered by an endothelium by 1 month. Although these devices resulted in poor aneurysm occlusion rates, they demonstrate the feasibility and biocompatibility of fully bioresorbable FDs. The authors mention that future studies will investigate different manufacturing techniques aimed at improving strut distribution over the aneurysm neck to increase aneurysm occlusion rates.

Jamshidi et al. (2020) were the next to report a partially bioresorbable FD[15]. Their devices contained 48 wires with different device iterations composed of 44/4, 46/2, or 48/0 bioresorbable poly-L-lactic-acid/tantalum coated nitinol wires for radio-opacity, structural support, and self-expandability during deployment. Hemocompatibility testing demonstrated no difference in hemolysis and a slightly improved thrombus surface coverage in a static thrombosis test for their BRFD relative to an industry standard control. Their nitinol containing BRFDs had crush resistances in the range of commercially available FDs. To evaluate the device performance in vivo, the 44/4 composition devices were implanted into the infrarenal abdominal aortas of 3 female New Zealand white rabbits for 1 month. All devices demonstrated good wall apposition after delivery. Only small neointimal growth occurred by 1 month. SEM analysis demonstrated a smooth neointimal covering over most of the struts. However, they reported that in some locations, the struts were not well opposed to the vessel wall. In these regions, thrombus was observed on the abluminal side of the devices via SEM. A separate study by Morrish et al. assessed the medical imaging characteristics of these nitinol containing BRFDs relative to industry standard FD controls[19]. There was no difference in anticipated visibility during deployment or image quality in fluoroscopy and CT images, respectively, when qualitatively assessed by 8 blinded neuroradiologists. The BRFDs significantly improved MR image quality by their assessment. Additionally, the BRFDs quantitatively reduced metal induced artifact in both CT and MR images. In conclusion, these primarily bioresorbable FDs have demonstrated great deployability and wall apposition without the use of a balloon, excellent imaging characteristics, and biocompatibility in the rabbit aorta. Future studies are required to determine the in vivo efficacy for treating aneurysms.

Tidwell et al. (2021) manufactured BRFDs out of bioresorbable polycaprolactone (PCL) using a micro-melting fabrication technique to create FDs with a strut width of 350 μm[23]. In vitro testing demonstrated that the devices had no toxic effects to human umbilical vein endothelial cells and that they were able to proliferate and form a monolayer on the surface of the PCL devices. Although the strut size of the device precludes its use for the treatment of IAs, this manuscript offered a novel approach and material for manufacturing FDs and demonstrated great biocompatibility of PCL with endothelial cells in vitro.

Metallic Bioresorbable Flow Diverters

While most of the current literature focuses on polymeric bioresorbable materials, a few publications have investigated bioresorbable magnesium in FD applications. Nevzati et al. (2017) described a method for creating a rat abdominal aorta sidewall aneurysm model[25]. Balloon expandable magnesium alloy stents were deployed over the aneurysm necks and evaluated after 7 or 21 days. The devices elicited high rates of thrombosis, which was attributed to the authors’ decision not to use dual antiplatelet therapy as well as the excessive strut size of the BRFDs. The BRFDs used in this study appeared to be laser cut with geometries and strut sizes that more closely resemble a coronary stent than industry standard FDs. Regardless of the large strut size, histology demonstrated a thin neointimal covering over the struts and across the neck of the aneurysm.

Gruter et al. (2019) reported a more comprehensive evaluation of magnesium BRFD performance in the rat aortic sidewall aneurysm model[24]. Their magnesium device appeared similar to the device used in the Nevzati study but featured a poly-L-lactic acid coating. After 4 weeks, the aneurysm occlusion rate was 85% for the magnesium BRFDs, compared to 20% for aneurysms that received no FD. Progressive bioresorption was observed and aneurysm occlusion rates were preserved out to 6 months. The authors anticipated ~95% device resorption would occur by 12 months. The performance of the BRFDs was also compared to cobalt-chromium control stents when used in stent assisted coiling after 4 weeks. Histology demonstrated no difference in inflammation between the devices, no signs of negative BRFD-coil interactions, and all aneurysms treated by the BRFD-coil combination were occluded. No intraluminal debris was observed from any of the BRFDs used throughout the study. Although the BRFDs coronary-stent-like geometry (excessive strut size and low pore density) will likely preclude its use for IAs, this study demonstrated that magnesium BRFDs can safely corrode in the aneurysm environment with comparable biocompatibility to permanent control devices and pioneered the use of BRFDs in stent assisted coiling.

Applying bioresorbable metals to BRFDs for the treatment of IAs remains vastly unexplored. To the authors’ knowledge, no studies have investigated metallic BRFDs with devices fabricated to match the porosity, pore density, and strut diameters of clinically approved permanent FDs.

Promise of Bioresorbable Metals

Bioresorbable metals such as magnesium, iron, zinc, molybdenum, and their alloys show great promise for application as BRFDs. The field of bioresorbable metallic cardiovascular stents is rapidly growing, with several devices already demonstrating their ability to safely resorb without adverse effects in the clinic[26]. Although one bioresorbable polymeric PLLA coronary stent achieved FDA approval in 2015, its inferior mechanical properties relative to metals resulted in devices that required large strut sizes with poor wall apposition, ultimately increasing rates of stent thrombosis, chronic inflammation, and restenosis relative to industry standard controls [27]. Consequently, these devices were removed from the market. However, the lower radial force requirement in the FD application will allow for smaller strut sizes, which is more conducive to bioresorbable polymeric devices. Regardless, cardiovascular stenting investigations have demonstrated that the intrinsically high mechanical strength of metals, as well as their high ductility and biocompatibility, favors metals for use as vascular scaffolds. These same principles may be applied to design metallic BRFDs with smaller struts and improved wall apposition. Magnesium and iron alloys are the most advanced bioresorbable metal classes. Magnesium has demonstrated great safety and biocompatibility in extensive clinical studies[26]. Magnesium’s innate surface properties are less thrombogenic than conventional stent materials, making it a great candidate for endovascular devices[14]. Indeed, the first three clinical trials investigating bare metal bioresorbable magnesium alloy coronary stents did not result in a single case of acute or late scaffold thrombosis[28–30]. It was not until these devices started featuring polymeric drug eluting coatings that stent thrombosis was observed[31]. Conventional FDs do not require drug eluting coatings, and therefore bare metal magnesium alloy BRFDs show great promise to reduce the rates of device thrombosis. The primary drawback is their weaker mechanical properties and fast resorption rate, which may require larger strut sizes relative to conventional and bioresorbable iron devices.

Conversely, the primary advantage of iron is its mechanical strength, which can be leveraged to design BRFDs with very small struts (~25 um). Bioresorbable iron alloys corrode much slower than magnesium, which may be suitable for smaller strut sizes. However, the biocompatibility of bioresorbable iron is less favorable, as it is associated with higher degrees and persistence of vascular inflammation relative to magnesium [26]. Inflammatory cells appear to have a more difficult time removing corrosion products, resulting in its persistence even after the stent has mostly degraded[32]. Regardless, this has not impeded clinical use. Initial clinical investigations of bioresorbable iron based stents in coronary arteries have demonstrated minimal restenosis at 6 month follow up that was preserved for 26 months[33].

Taken together, the mechanical properties, biocompatibility, and clinical success in stenting applications suggests that bioresorbable metals have great potential for use as BRFDs. Their use in this application is widely unexplored and a great opportunity for further investigation.

Conclusions, Challenges, and Future Directions

BRFDs have many theoretical advantages over conventional permanent FDs. Many advantages have been experimentally demonstrated in the similar bioresorbable coronary stenting application. Additionally, the growing BRFD field has demonstrated the feasibility and efficacy of these devices for occluding aneurysms, with devices made from a spread of materials and manufacturing techniques. However, many challenges and questions remain.

The Challenge of a Fully Bioresorbable Device

The incorporation of permanent, metallic backbone materials such as nitinol can be used to bolster the device’s radial strength, radio-opacity, and self-expanding properties. As seen with the BRFD designed by Jamshidi et al., the incorporation of 4 nitinol wires into the mostly bioresorbable polymeric braid allowed the devices to be deployed without ballooning[15]. The ideal BRFD avoids the requirement of ballooning, as it makes navigation and delivery more cumbersome. Therefore: What percentage of the device composed of permanent reinforcing materials is optimal for delivery without sacrificing the advantages associated with device bioresorption? Furthermore, after complete dissolution of the bioresorbable components is complete, what happens to the remaining few wires? Determining the ideal relationship between permanent and bioresorbable device components warrants further investigation.

Device Lifetime

How long must the BRFD maintain mechanical integrity, allowing the aneurysm to occlude and heal before it can substantially resorb? For bioresorbable stenting applications, extensive clinical investigations in atherosclerotic environments were required before it was determined that they must maintain structural integrity for 6-12 months before substantial bioresorption is beneficial[26]. Wang et al.’s study in the rabbit elastase induced aneurysm model suggested that device integrity is required for at least 3 months[21]. However, approximately half of their device was permanent/still present at the 3 month follow up. Clinical data suggests that for conventional FDs, blood vessel remodeling and mature neointimal covering over the aneurysm neck typically occur in the 6-12 month range[12]. However, at this point could substantial device bioresorption increase the risk for recanalization? Future studies are needed with longer follow ups aimed at determining the ideal BRFD lifetime.

Corrosion Behavior

Bioresorbable device corrosion behavior is yet to be extensively studied in the aneurysm environment. Two immediately pressing questions are: 1) Do corrosion characteristics for struts located over the neck of the aneurysm differ from struts opposed to the parent vessel wall? Accelerated device corrosion over the aneurysm neck may require devices with slower corrosion rates than anticipated. 2) Will corrosion products accumulate in the healing aneurysm and influence the biological response?