Effects of Electret Coating Technology on Coronary Stent Thrombogenicity

Abstract

Stent thrombosis (ST) is a catastrophic event and efforts to reduce its incidence by altering blood-stent interactions are longstanding. A new electret coating technology that produces long-lasting negative charge on stent surface could make them intrinsically resistant to thrombosis. We assessed thrombogenicity of stents using an annular perfusion model with confocal microscopy, and determined the efficacy of electret coating technology to confer thrombo-resistant properties to standard stents. Using an annular perfusion chamber, Bare Metal Stent (BMS), standard uncoated DES (DES) and Electret-coated DES (e-DES) were exposed to human blood under arterial flow-conditions. Deposits of fibrinogen and platelets on the stent-surface were analyzed using immunofluorescence staining and confocal microscopy. Surface-coverage by fibrinogen and platelets and the deposit/aggregate-size were quantified using computerized morphometric analysis. The experimental methodology produced consistent, quantifiable results. Area of stent surface covered by fibrinogen and platelets and the average size of the deposits/aggregates were lowest for e-DES and highest on BMS, with DES in the middle. Size of fibrinogen-deposits showed no differences between the stents. The testing methodology used in our study successfully demonstrated that electret coating confers significant antithrombotic property to DES stents. These findings warrant confirmation in a larger study.

Introduction

Stent deployment is a critical implement in the management of obstructive coronary artery disease (CAD), but its benefits can turn in to catastrophic harm in case stent thrombosis (ST) develops.¹ The development of ST can acutely disrupt blood flow, resulting in myocardial infarction (MI), with relatively high mortality rates. Most cases of ST occur early (within the first 30 days) after stent deployment, with an incidence rate of approximately 0.5–1% according to analyses published in the past decade,^2–4 and the 1-year rate with the newer generation DES ranging from <1% to 1.7%, depending on the study.^5–8 The rates of ST have improved with the development of newer generation of drug eluting stents versus the older bare metal stents (BMS),^{6, 7} but the delayed re-endothelialization from cytostatic drugs used with the former to minimize smooth muscle cell growth, can lead to tissue factor expression and activation of extrinsic coagulation cascade.¹ Newer generation of drug eluting stents is characterized by faster stent re-endothelialization and lower rates of ST, and are the current gold standard for interventional treatment of obstructive CAD. However, rates of ST are still at levels that, given the volume of stent placements each year, remain a significant concern especially in high-risk patients.^{3, 4, 6, 7}

The role of platelets in vascular thrombus formation and the intervening mechanisms that modulate their interactions with damaged vessel walls are well characterized, but the processes implicated in their interactions with artificial biomaterials appear to have some differences.^9–12 Platelet attachment to vascular surfaces involves the interaction of platelet-membrane GPIb-IX-V receptor complex with the von Willebrand factor in the subendothelial matrix. In contrast, the interaction of platelets with artificial surfaces appears to be modulated by - maybe even dependent on - adhesion to pre-adsorbed plasma proteins, primarily fibrinogen.¹³ Immobilized fibrinogen is a substrate for platelet-arrest under flow conditions, the process selectively mediated by GPIIb-IIIa in the conformation on the surface of inactive platelets.¹⁴ Regardless of the mechanisms involved, the significance of platelets in ST is highlighted by the reductions attained in the rates of thrombosis with antiplatelet therapy. Long-term dual antiplatelet therapy (DAPT) is a requirement after stent placement but their benefits are accompanied by an increased risk of bleeding. Development of a stent intrinsically resistant to thrombosis could theoretically lower ST and bleeding risk at the same time, by minimizing the need for long-term DAPT.

The surface of blood vessel wall and blood cells is electro-negatively charged under physiological conditions, sustaining a repulsive electrical force that helps to maintain the integrity and patency of the blood vessels.¹⁵ Endothelial dysfunction, as in cardiovascular disease, or injury, as in during stenting, modifies their electrical charge, favoring the attachment of circulating platelets and proteins.¹⁶ This knowledge is applied in the development of new stents that could offer innate resistance to platelet adhesion and thrombus formation. A new negatively charged, quasi-permanent electret coating technology developed recently could enable existing stents to gain antithrombotic properties of their own. Electret coating may be able to retain its electric charge for several years making it an electrostatic equivalent of a permanent magnet.

Evaluating thrombogenicity of stents to gauge their post-deployment performance is challenging at multiples levels. Human studies are impractical because of ethical, cost and time-related issues, and the lack of testing methodology for the in vivo assessment of thrombogenicity. Animal models offer dependable representation of human atherosclerotic disease, but the mechanisms of platelet adherence and aggregation even in large animal models differ significantly from humans.¹⁷

In the present study, we have used the annular perfusion techniques in combination with confocal microscopy to investigate thrombogenicity of electret-coated stents. The utility of this model for the evaluation of stent thrombogenicity has never been tested before.

The aim of this study was to assess stent thrombogenicity using an established perfusion model and to determine if the new electret coating technology confers any intrinsic resistance to thrombus formation, to commercially available stents. To achieve our objective, we evaluated stents with and without electret coating for thrombogenicity after exposure to human blood.

Methods

In this proof-of-concept study, thrombogenic potential of 3 types of stents was investigated by measuring fibrinogen and platelet deposition on the study stents using the annular perfusion chamber model in conjunction with confocal microscopy. Each stent was exposed to circulating human blood from healthy volunteers in vitro, perfused through the annular chamber and the resulting platelet and fibrinogen deposition was quantified using confocal microscopy after antibody staining. The stents were coded and all analyses performed blinded. The endoluminal devices studied included: Bare metal stent (BMS; Multilink, Abbott), Standard drug eluting stent without electret coating (DES; Xience, Abbott), and Standard drug eluting stent with electret coating (e-DES; Amber Medical).

The study was approved by the Program for the Protection of Human Subjects (Institutional Review Board) of the Icahn School of Medicine at Mount Sinai. A written informed consent was obtained from each study participant before initiation of any study procedure.

Electret Coating

Electret coating is a decades old technology that generates a long-lasting state of polarization on the subject material, making it the electrostatic equivalent of a permanent magnet.^{18, 19} Based on extrapolation of short-term experiments, the electret charge is estimated to last for years, modulated by both the technological parameters and the material used.^{18, 20}

Standard DES (Xience, Abbott) were utilized for the electret coating. The coating material was tantalum pentoxide (Ta205), a dielectric material that is mechanically strong, chemically inert, biologically compatible with living tissues, and highly adhesive to the stent surface. This material can easily take the electret state with a negative charge. Vacuum plasma spraying was used for coating the stent in vacuum. The stent surface was thoroughly cleaned before coating. Tantalum pentoxide molecules and ions were sprayed on the stent surface at a high speed. Tantalum pentoxide coating is radiopaque and does not require additional contrast agents, making the manufacture of stents simpler and more cost-effective.

Blood Donors

Healthy male and female volunteers (n=9) served as blood donors. All participants abstained from taking aspirin or any other antiplatelet/anticoagulant medications for all least 2 weeks. Blood from each volunteer was collected in vacutainer tubes with Na-citrate as anticoagulant; and after gentle mixing split equally into 3 falcon tubes (approximately 25 ml each) to perfuse one stent of each type/donor. This design significantly reduces intra-subject variability.

Blood was not recalcified for use in experiments as our prior experience has shown that recalcified blood begins to clot in the circuit after approximately 5 minutes of perfusion. The relatively prolonged exposure of study stents to circulating blood warranted the use of citrate-anticoagulated blood without recalcification.

Perfusion Studies

Thrombogenic potential of the stents under investigation was assessed in vitro using annular perfusion chambers - an overview of the system is illustrated in Figure 1. The annular perfusion chamber is based on models tested over several.^21–23 Perfusion in the annular chamber is under user-defined shear rate conditions producing laminar flow that mimics in vivo flow conditions. The methodology was acquired for this study from the Escolar group in Madrid, where it has been used to investigate the alterations in hemostasis of uremic patients,²⁴ assess the hemostatic effects of platelet transfusion in patients with severe von Willebrand disease,^{25, 26} and more recently the reversal of direct oral anticoagulants.^{27, 28} However, the potential utility of this model for the assessment of stents has never been tested before. In preliminary experiments, we assessed and refined various aspects of the methodology – i.e. flow rate and duration, antibody titration for labelling immunofluorescence and confocal microscopy settings - until reliable and reproducible results were obtained. The study experiments were initiated only after all aspects of the methodology had been finalized.

Figure 1:

The Annular Perfusion Chamber System. Photograph of the annular chamber with the stent placed on the inner rod structure (A). Graphic illustration of the perfusion model (B).

In each study experiment, the stent was carefully placed in the annular chamber using forceps as depicted in Figure 1a. The chamber was then placed into the perfusion system (Figure 1b) where it was perfused with donor blood with the help of a peristaltic pump, exposing the stent to flowing blood at 800 s¹ for 20 minutes. The 20-minute duration was selected after preliminary testing with different experimental conditions, including shorter and longer perfusion times. Longer perfusion runs had the capacity to produce more thrombosis, but they also generate artifacts due to unavoidable excessive hemolysis by the rotor heads that recirculate the blood. At the 20-minute mark, the inflow channel was switched from blood to PBS without interrupting the flow and the system flushed for one minute to clear it of all blood and prevent excess thrombosis due to stasis. The stent was then carefully removed using surgical forceps and transferred to paraformaldehyde 4% for fixing. Detailed description of the perfusion system has been reported previously.²⁹

Each of the 3 stent types were evaluated using blood from each donor, yielding 27 perfusion studies in total (9 per stent type). To avoid confounding effect of testing order, the order was changed with each donor to yield equal runs at first, second and third for each stent type.

Immunostaining of Platelets and Fibrinogen

After completion of perfusion study, each stent was fixed with 4% paraformaldehyde (15 minutes) and immersed in 1% glycine (10 minutes) to reduce background/non-specific labeling. Thereafter, the stents were exposed to 1% bovine serum albumin (BSA) for 15 min prior to incubations with specific antibodies. To define the contribution of platelets and fibrinogen, combination of indirect and direct immunofluorescence labeling was used. Platelets were stained with a mouse anti-CD36 primary antibody for 1 hour at room temperature. Fibrinogen deposition was assessed by using a secondary antibody (anti-mouse Alexa fluor 488) incubated together with a conjugated antibody anti-fibrinogen (Alexa fluor 594) for 1 hour at room temperature.

Fluorescence Microscopic Examination

Each study stent was carefully examined under florescence microscopy (Leica DM5000B microscope and Leica DFC300 FX digital camera) prior to confocal analysis. This examination was performed at a low magnification (2.5X) with the objective to get a quick overview of the experiment and record the overall pattern of fibrinogen and platelet deposition on the stent surface. Any abnormalities in the deposition patterns (e.g. more aggregates in one area versus another) were to be further explored in order to identify the cause. This examination was conducted over the entire length of each stent.

Confocal Microscopic Analysis and Image Acquisition

The stents were mounted on a glass bottom chamber and imaged on a Leica TCS-SP5 Laser Scanning Confocal Microscope. Alexa Fluor 488 (green) and Alexa Fluor 594 (red) and reflection images were acquired using a 10X magnification objective. Excitation laser lines used were 488 and 561nm. Emission was acquired through internal photomultipliers at the emission ranges of 500–550 nm and 571–625 nm, 555–565 nm, for reflection, respectively. Confocal pinhole was set at 1 Airy unit. Bright field images were acquired simultaneously through a transmitted light detector.

Quantification of Platelets and Fibrinogen Deposition and Aggregate Size

Morphometric analysis was performed using Image-J software.³⁰ Areas of stent surface covered by platelets, by fibrinogen, and in total were quantified for each microscopic field. Platelet/fibrinogen deposition was determined by calculating the covered stent area as a percentage of total stent area for that field, with the results expressed as surface coverage (%).

The size of aggregates on the stent surface was measured using the ‘Analyze Particle’ function of imageJ. Scale was set using the specifications of the Xience stent as a reference. After careful review, particles less than 10 micrometer in size were assessed to be artefacts and the threshold for measurement was set at >10 micrometer.

For both deposition and aggregate size measurements, a total of 6–8 fields were analysed for each stent and the results averaged to generate final results for platelets and for fibrinogen, by stent type for each donor.

Statistical Analysis

The outcome measures analyzed were the percent of stent surface coverage and mean aggregate size. All results are expressed as median (95% CI) unless specified otherwise. Comparisons between the types of stents were performed using repeated measures ANOVA. Subsequent pairwise comparisons (uncorrected Fisher’s LSD test) were done when statistical significance was observed in ANOVA. The threshold for statistical significance was set at the nominal P=0.05 level. All analyses were performed using SPSS 20.

Results

The study was conducted with 9 healthy volunteers (7 males) serving as blood donors, with an average age of 40.6 ± 15.3 years old. The testing methodology, combining studies in established perfusion model with confocal analyses, produced consistent, measurable results that proved useful in determining the fibrinogen and platelet interactions with the metallic stent surface.

Low resolution examination of each stent under florescence microscope exhibited a uniform pattern of fibrinogen and platelet deposition on the stent surface. A representative collage of the confocal images acquired and analyzed in this study, is presented in Figure 2.

Figure 2.

Confocal microscopic representative images of study stent. Panels on the left show fibrinogen deposition and on the right show platelet deposition, on bare metal stent (A), standard/uncoated drug eluting stent (B) and Electret-coated drug eluting stent (C).

Stent surface coverage by fibrinogen and platelets

Fibrinogen deposition was highest on the surface of BMS and lowest on e-DES (Figure 3; left panel). Median coverage on e-DES was 33.8% lower than BMS and 18.7% lower than DES (p=0.005 and 0.043, respectively). Fibrinogen deposition on the standard, non-coated DES was also significantly lower than on BMS (p=0.043, Table 1 here).

Figure 3.

Box plot of fibrinogen and platelet deposition. Box plots summarizing surface coverage of each stent type by fibrinogen and platelets. Fibrinogen deposition (left panel) on the standard/uncoated drug eluting stent (DES) was significantly lower than on bare metal stent (BMS) and on the Electret-coated drug eluting stent (e-DES) was significantly lower than both DES and BMS. Similar findings were noted for platelet deposition (right panel) but the differences were statistically significant only between e-DES and BMS.

Table 1:

Platelet and Fibrinogen deposition

			Median	95 % CI	P value
					vs. BMS	vs. DES
Surface coverage (%)	Fibrinogen	BMS *	26.3 %	22.2 – 33.2 %	-	0.040
		DES †	21.4 %	15.3 – 29.3 %	0.040	-
		e-DES ‡	17.4 %	14.8 – 22.2 %	0.005	0.043
	Platelets	BMS *	24.0 %	21.6 % - 31.2 %	-	0.055
		DES †	20.0 %	15.7 % - 26.1 %	0.055	-
		e-DES ‡	18.1 %	12.6 % - 19.9 %	0.012	0.100
Aggregate size (μM)	Fibrinogen	BMS *	277 μm	223 – 453 μm	-	0.862
		DES †	269 μm	232 – 470 μm	0.862	-
		e-DES ‡	292 μm	231 – 352 μm	0.455	0.493
	Platelets	BMS *	562 μm	237 – 703 μm	-	0.053
		DES †	340 μm	177 – 596 μm	0.053	-
		e-DES ‡	282 μm	174 – 476 μm	0.021	0.141

^*Bare metal stent,

^†standard/uncoated drug eluting stent,

^‡Electret-coated drug eluting stent.

Findings for platelet deposition showed a similar trend, with surface coverage being lowest on e-DES and highest on BMS (Figure 3; right panel). Median surface coverage by platelets on e-DES was 24.6% lower than on BMS (p=0.012) and 9.5% lower than on DES (p=NS). The standard DES had 16.7% less platelet coverage than BMS, but the differences were not statistically significant (p=0.055, Table 1).

Size of deposits/aggregates formed on stent surface

Average size of the fibrinogen deposits on the surface of the 3 stent types did not show any statistically significant differences (Figure 4; left panel). The size of platelet aggregates formed on the stent surfaces however, did show substantial differences (Figure 4; right panel). The median aggregates on the surface of e-DES were 49.8% smaller than on BMS and 17.1% smaller than on DES, but only the former achieved statistical significance. The size of aggregates on the uncoated DES was also smaller than BMS, but the differences were not statistically significant (Table 1).

Figure 4.

Box plot of fibrinogen- and platelet-aggregate size. Box plots of the size of fibrinogen and platelet aggregates formed on the three types of study stents. No differences were observed in the size of fibrinogen aggregates were observed (left panel). The average size of platelet aggregates (right panel) on drug eluting stent (DES) was smaller than on bare metal stent (BMS) and those on the Electret-coated drug eluting stent (e-DES) smaller than both DES and BMS. However, these differences achieved statistical significance only between e-DES and BMS.

Discussion

The experimental methodology in our study shows that the electret coating technology we investigated has the ability to confer antithrombotic properties to standard drug eluting stents.

Evaluating the thrombogenicity of stents - whether in human or animal models, using in vivo or in vitro approach - is challenging for a number of reasons. Whereas animal model can provide valuable insights into long-term vascular responses to stent-deployment that can reliably be inferred to humans, in vitro flow-based models may be the principal option for assessing the acute reactivity of blood components to stents.³¹ The challenges go beyond study model selection and extend to the choice of testing modalities. Imaging technologies lack the spatial resolution and/or sensitivity to differentiate, quantify and detect small differences in components of a stent thrombus. Pathological evaluations offer better options, but require processing steps like cutting and mounting that are not possible with stents due to their firmness and durability.

A perfusion chamber model offers the most adaptable methodology for testing thrombogenicity of stents in an in vitro setting. Perfusion models with flat surface (Badimon chamber)^32–34 or those based on microfluidics (Maastricht flow chamber and its various derivative)^35–37 offer tremendous value, but do not have the capability to test complex structures like stents. The annular perfusion chamber, with its cylindrical design, is the only model with the capacity to test structures like stents under shear rate conditions producing laminar flow mimicking in vivo flow conditions. The use of the annular perfusion model, combined with confocal methodology for the quantification of fibrinogen and platelet deposits, exhibited sufficient sensitivity to detect measurable differences in the thrombogenicity of the study stents. Confocal microscopy has previously been used to study platelet interactions with coronary stents in an ex vivo arteriovenous shunt model using porcine blood but the interactions of platelets and fibrinogen were not quantified in that study.³⁸ Using our methodology, we were able to quantify the deposits of fibrinogen and platelets and also grossly examine the pattern of these deposits. In our in vitro experiments, the patterns of platelet deposits closely mirrored those of fibrinogen, apparently confirming earlier reporting that platelet reactivity to stent surface maybe modulated by pre-adsorbed fibrinogen.¹³ However, since our experiments were not designed to differentiate between plasma fibrinogen bound to integrin GPIIbIIIa or fibrinogen found in platelet granules, this cannot be stated with certainty.

The electret coating technology evaluated in the present study may confer significant resistance to thrombogenicity when applied to existing DES. The lower thrombogenicity in our study was recorded after a brief period of exposure to human blood and were more significant with fibrinogen deposition. Although the reduction in platelet deposition was on a comparable scale, the differences did not achieve statistical significance. It should be noted that this was a small, proof-of-concept study not powered to detect statistically significant differences. Furthermore, the lack of any prior data to indicate the degree of differences that could be expected, sample size/power calculation was not possible. Given the clear and consistent trends observed in our study, an absence of statistical significance is more likely the result of inadequate sample size.

Extrapolating the findings of our short in vitro study to gauge the possible performance of electret-coated stents in patients - where the contact with blood would be for exponentially longer periods – would be an educated guess. It is not unreasonable to posit though, that the thrombogenic differences apparent after a few minutes of testing may be magnified when the period of exposure to blood is extended to durations where acute (within 1 day) and subacute (within 1 month) stent thrombosis become a concern. Furthermore, a thrombo-resistant stent may also offer the possibility of reducing the need for some antiplatelet drug use. Long-term DAPT has become an integral part of any management plan involving stent placement and the absence / early discontinuation of DAPT has been shown to be the single most important predictor of early and late ST.^39–41 Use of DAPT is not without clinical consequences however, and is associated with increased bleeding rates. Among patients who currently undergo percutaneous coronary intervention both for acute coronary syndromes as well as stable angina indications, the proportion who have a high bleeding risk is significant, up to 2.8% at one year.⁴² Any possibility to lower the use of DAPT, either in terms of dosage or treatment duration, could prove to be beneficial. Therefore, developing a stent with antithrombotic properties that could reduce the use of DAPT is of very high clinical importance. These possibilities need to be explored in future studies.

Conclusions

The methodology reported in our study demonstrates the possibility of screening the thrombogenicity of stents by exposure to homologous blood components, before expanding the studies to animal models or clinical investigations. In our testing, electret coating proves to be a promising new technology to reduce the thrombogenic potential of stents.