Chitosan-Nanoclay Embolic Material for Catheter-Directed Arterial Embolization

Abstract

Minimally invasive transcatheter embolization is a common nonsurgical procedure in interventional radiology. It is used for the deliberate occlusion of blood vessels for the treatment of diseased or injured vasculature, including vascular malformation and malignant/benign tumors. Here, we introduce a gel embolic agent comprising of chitosan nanofibers and nanoclay with excellent catheter injectability and tunable mechanical properties for embolization. The properties of the gel were optimized by varying the ratio between each individual component and also adjusting the total solid content. The rheological studies confirm the shear thinning property and gel nature of the developed gel as well as their recoverability. Injection force was measured to record the force required to pass the embolic gel through a clinically relevant catheter, evaluating for practicality of hand-injection. Theoretical predicted injection force was calculated to reduce the development time and to enhance the physician’s experience. The stability of occlusion was also tested in-vitro by monitoring the pressure required to displace the gel. The engineered gels exhibited sterility, hemocompatibility and cell biocompatibility, highlighting their potential for transcatheter embolization.

Graphical Abstract

A shear thinning hydrogel embolic material comprising of laponite nanoclay and chitosan nanofibers was developed, characterized and studied for transcatheter based minimally invasive surgery. The developed gels exhibited tunable rheological properties which is beneficial for a wide range of embolization conditions and the injection force of the gel through a catheter was measured as well as predicted using rheological parameters for improving the physician’s experience. The engineered embolic gels demonstrated exceptional in-vitro stability to withstand physiological blood pressure, along with excellent sterility, hemo- and bio- compatibility, highlighting the potential application for catheter-directed arterial embolization.

1. INTRODUCTION

Catheter directed arterial embolization is a minimally invasive surgical procedure to deliberately occlude blood vessels using some embolic agents delivered through a catheter for the treatment for diseases such as hemorrhage, aneurysm, tumors, vascular malformation etc.^(1)–(3) Blood vessels, running through the whole body, are effective routes for embolic agents to enter a body and access any organs or parts of body using the aid of a catheter. Compared to the conventional open surgery, catheter directed arterial embolization is safer, with less recovery time, minimal complications, controlled bleeding and low cost.⁽¹⁾ Conventionally used embolic agents are solid (i.e., metallic coils and microspheres) and liquids (i.e., Onyx and nBCA systems). The metallic coils, usually made of platinum⁽⁴⁾, stainless steel⁽⁵⁾ or nitinol⁽⁶⁾, have a diameter between 0.2 to 1.3 mm ⁽⁷⁾, which can effectively occlude arteries (up to 1 -2 cm diameter) but cannot occlude small capillaries with diameters of 5-10 μm.^(1),(8) Microspheres with a size range of 20-1200 μm, on the contrary, are excellent candidates for occluding small blood vessels in tumors, such as liver carcinoma and uterine fibroids, but are not used in the intervention of aneurysms.⁽¹⁾ Recently, a new class of semi-solid embolic material, Obsidio™ conformable embolic agent (Boston Scientific, USA), has been FDA 510k cleared and developed for peripheral embolization.^(9),(10)

The size constraint of solid embolic agents makes them less versatile along with other drawbacks such as recanalization, risk of non-target embolization and failure in coagulopathic patients.⁽³⁾ On the other hand, liquid embolic agents can flow to vasculature of any diameter and successfully occlude blood vessels. However, most of the liquid embolic agents contain organic solvents (e.g. DMSO) which can cause both local and systematic cardiovascular toxicity and also the liquid glue adhere to the catheter resulting in improper embolization.^(1),(3) Therefore, it is of great importance to develop an embolic agent with the properties of both solid and liquid such as a shear thinning hydrogel, thereby the material can have an optimum mechanical strength similar to that of solids and can flow when pressure is applied. The flow property of the shear thinning gel makes the material more versatile and the absence of organic solvents can reduce the risk of toxicity.

Laponite nanoclay (NC, Na⁺ _0.7[(Mg_5.5Li₀.₃)Si₈O₂₀(OH)₄]^-_0.7) consists of synthetic silicate nanoparticles with a disc shaped nano-structure with 20-30 nm in diameter and approximately 1 nm in thickness.^(11)–(13) Laponite nanoclay possesses anisotropic charge distribution with positive charges along the rim and negative charges at the top and bottom surfaces, thereby enabling them to form self-assembled structures which can dynamically form and break with respect to the application of a shear force, making them an excellent shear thinning gels.^{(3),(13),(14)} The synthetic nanosilicates form gels at high concentrations due to the electrostatic and van der Waals interactions resulting in the formation of “house of cards” structure.⁽¹⁵⁾ Recently, NC is used in variety of biomedical applications including drug delivery agents,⁽¹⁶⁾ scaffolds for tissue,^(17),(18) rheological modifiers,⁽¹⁹⁾ solid hemostat products,⁽²⁰⁾ and osteogenic agents.⁽¹⁵⁾ Gaharwar et al. developed NC-gelatin nanocomposite hydrogel, an injectable hemostatic agent for the treatment of hemorrhage and found a decrease in in-vitro clotting time by 77%.⁽¹³⁾ Altun et al. developed a shear thinning hydrogel using NC and blood derived platelet rich fibrin with good injectable and regenerative properties that can achieve instant and durable intra-arterial hemostasis.⁽²¹⁾ Albadawi et al. developed a tantalum loaded nanocomposite hydrogel, which is visible by most of the clinically available imaging modalities including ultrasound, computed tomography, magnetic resonance imaging, and fluoroscopy without significant artifacts.⁽¹⁰⁾ Avery et al. engineered a shear thinning nanocomposite hydrogel containing gelatin and NC with comparable hemostatic property with that of conventional metallic coils.⁽²²⁾

Chitosan (CH), the second most abundant natural polymer, is obtained by alkaline deacetylation of chitin which is majorly derived from the shells of shrimps or crabs.⁽²³⁾ The main backbone of CH consists of N-acetyl-d-glucosamine and D-glucosamine units with one amino group (-NH₂) and two hydroxyl group (-OH) groups in each repeating glycosidic unit.⁽²⁴⁾ The free amino and hydroxyl group in the CH backbone acts as an active site for the drugs or image contrast agents to attach and also imparts positive charge making them soluble in dilute acids and neutral solvents.^(24),(25) They are used in wide range of application such as food packaging, artificial skin, drug delivery, cosmetics and dye adsorption.⁽²⁶⁾ Hydrogels containing CH are extensively studied in the biomedical applications such as drug delivery,⁽²⁷⁾ tissue scaffolds,^(28),(29) and wound healing,⁽³⁰⁾ owing to its biocompatibility, biodegradability, mucoadhesiveness and antimicrobial properties.^(27),(31) Zhang et al. developed a CH/polyaniline/NC hydrogel having good biocompatibility and photothermal conversion capacity for tumor therapy and tissue regeneration.⁽³²⁾ Fan et al. engineered a CH/gelatin/PVA hydrogel with good mechanical and hemostatic properties which can be considered as a potential candidate for wound healing.⁽³³⁾ Ranjbardamghani et al. developed a CH/NC hydrogel using a hybrid cross-linking method by genipin and β-glycerophosphate (BGP) for cartilage regeneration.⁽³⁴⁾

One key property of any embolic agent to be considered is its transcatheter injectability. The force required to pass the embolics through a catheter is directly related to physician’s experience in clinical setting.⁽³⁵⁾ The peak force required to initiate the movement of syringe plunger is known as the break loose force, which physicians need to overcome. It is important to recognize that the break loose force of the embolics should be in the comfortable range for physicians during hand injection. Once the break loose force is achieved, a constant force (less than break loose force) will be required to sustain the movement of the plunger to deliver the embolics, namely the injection force.

So far only a few studies have been reported on the use of hydrogels for endovascular embolization. Moreover, embolic efficiency of CH-based hydrogels has not been studied despite the material possesses good biocompatibility, bio-degradability, anti-microbial activity and potential active sites for drug and contrast agent conjugations. In this study, we developed shear thinning gel embolic agent comprised of NC and CH nanofiber, aiming to occlude varying vasculature sizes and geometries. The morphology of the CH nanofiber and the gels were examined using a scanning electron microscope (SEM). The rheological properties of the gels were investigated with tunable mechanical properties, suggesting their potential for a wide range of occlusion capability. In addition, transcatheter injectability was assessed both experimentally and theoretically to correlate between gel rheology and injection forces. The gels demonstrated excellent sterility, hemocompatibility, thrombogenicity, and biocompatibility, suggesting their potential to serve as embolic agents for transcatheter intervention. Finally, translational considerations, such as shelf-life, sterilization, and imageability are discussed.

2. MATERIALS AND METHOD

2.1. Materials

Laponite-XLG nanoclay (NC) was purchased from BYK USA Inc., Texas, USA (Lot No: 0002303785), Chitosan nanofibers (CH) was purchased from Sugino Machine Ltd, Japan. Molecular grade water from Intermountain life sciences (Utah, USA) was used for engineering the gels. Phosphate buffered saline (PBS) tablet was purchased from Sigma Aldrich (St. Louis, MO, USA).

2.2. Dynamic Light Scattering (DLS)

The hydrodynamic diameter of NC nanoparticle and zeta potential of both NC, CH nanofiber and selected CH-NC gel were determined using dynamic light scattering (Litesizer 500, Anton Paar, Austria). An aqueous solution of 0.0625 wt% of NC, CH and embolic gel were prepared for the measurement (n=3).

2.3. Scanning Electron Microscopy (SEM)

The morphology of CH nanofiber and lyophilized gels were imaged using a scanning electron microscope (Hitachi SU3900, Japan) at an accelerating voltage of 20 kV. The gels were pre-frozen at −80°C for 24 hours and then lyophilized for another 24 hours to ensure complete removal of water. Chitosan nanofibers were suspended in DI water (0.125 wt% concentration) and then drop-casted on to a glass slide for imaging.

2.4. Preparation of embolic gel

To engineer CH-NC gels for transcatheter delivery, NC, CH and molecular grade water were mixed at pre-determined weight ratios using a specialized high-speed mixer (DAC 330-100 SE, FlackTek Speed Mixer, USA) at 3000 rpm for 5 minutes and repeated for 3 times. The gels were named as xNCy, where x is the total solid content in the gel and y is the percentage of nanoclay in the total solid content (E.g.: 8NC50 represents gel with total solid content of 8 wt%. 50% of the total solid content is nanoclay (4 wt%) and remaining 50% of the total solid content is CH nanofibers (4 wt%)). The CH-NC systems with 4, 6 and 8% total solid content were prepared with varying NC content of 100, 75, 50, 25 and 0% in each system, as shown in Table S1.

2.5. Rheology studies

All rheological studies were conducted at 25°C using a rheometer (MCR 302e, Anton Paar, Austria). A sandblasted 25 mm upper plate and a sandblasted lower plate were used, with a 1 mm gap in between for all tests. A solvent trap was used to maintain the humidity and prevent the gels from drying. The flow curves were performed to evaluate the shear thinning behavior of the gels by varying shear rate between 0.01 s⁻¹ and 1000 s⁻¹. Amplitude sweeps were carried out with shear strain from 0.01% to 100% to assess the linear viscoelastic (LVE) region of gel, as well as its storage modulus (G’) and loss modulus (G”). The angular frequency was kept at 10 rad/s. Flow curves and amplitude sweeps were conducted three times for each test. Frequency sweeps were conducted with increasing angular frequency from 0.1 to 100 rad/s at 0.1% shear strain (LVE region). Lastly, thixotropic test was performed to evaluate gel’s recoverability by oscillating between 0.1% (low shear rate) and 100% (high shear rate) for 2 minutes each for a total of 20 minutes. All rheological tests were conducted at 25°C because it is the ambient temperature at which an injection would be performed in the clinic.⁽³⁶⁾

2.6. Yield stress and strain

Yield strain is determined from the intersection point of tangents drawn from the linear viscoelastic (LVE) region of G’ and from the section with low G’ which is at high shear strain. A schematic of yield strain estimation is shown in Fig S1(a). Yield stress can be obtained from the stress corresponding to the yield strain.

2.7. Injectability

The injectability of the embolic gel was measured using a mechanical tester (Univert, CellScale, Canada). The catheter was connected to a 3 mL syringe (3 mL BD Luer-Lok™ Syringe sterile, BD, USA) filled with the gel using the Luer lock. The force required to inject the gel through a 100 cm 5F catheter (Cordis, Florida, USA) at a flow rate of 2 mL/min was recorded. The force required for injecting the gels, loaded in 1 mL syringe (BD Luer-Lok 1 mL syringe, BD, USA), through a 1.8F 150 cm microcatheter (SuperCross microcatheter, Teleflex, USA) at a flow rate of 1 mL/min was also tested. Each test was run in triplicates.

2.8. In-vitro pressure test

A differential pressure transducer (Omega Engineering, Norwalk, CT, USA) was used to determine the ability of the CH-NC gel to resist the physiological blood pressure. The CH-NC gel was placed in a tube (diameter of 3/32” and 3/16”, Tygon PVC soft plastic tubing, USA). The pressure transducer was connected between the upstream and downstream of the gel as shown in Fig S1(b). Fresh whole porcine blood (LAMPIRE Biological Laboratories, Pipersville, PA, USA), was injected into the tube at a flow rate of 50 mL/min using a syringe pump (Legato 100, KD Scientific, USA). The in-vitro pressure required to displace the gel from the tube was recorded using the pressure transducer. Each sample was tested three times.

2.9. Swelling and degradation test

The swelling of the gels was tested. Approximately 0.5 mL of each gel was placed in an Eppendorf tube and lyophilized for 24 hours to remove all the water. The lyophilized gels were then weighed for their initial weight, W₀. The gels were then incubated with PBS of at pH of 4, 7 and 9, respectively, at 37° C (n=4). After incubating for 10 min, 20 min, 30 min, 40 min, 50 min, 1 hr, 2 hrs, 18hrs and 24 hrs, the PBS was carefully removed and the gels were weighed (W_r - final weight of the gel after incubation). The percentage degree of swelling was calculated using the following equation:

$$DegreeofSwelling\left(\mathrm{\%}\right)=\frac{W_R}{W_0}\times W_0\times 100$$ Equation 1

The degradation of the embolic gels was also tested. Approximately 200 μL of gel was taken in an Eppendorf tube and weighed (W₀ - initial weight of the gel). The gel was then incubated with 200 μL of PBS at pH of 4, 7 and 9 respectively at 37° C (n=4). After incubating for 1, 3, 6, 8 and 24 hours, the PBS was removed from the tube and then the gels were weighed (W_r - final weight of the gel after incubation). The degradation percentage of the gels was calculated using the following equation:

$$Degradation\left(\mathrm{\%}\right)=\frac{W_R}{W_0}\times 100$$ Equation 2

2.10. In-vitro cell viability

The in-vitro cytotoxicity of 8NC100 and 8NC50 were studied using L929 fibroblast cells (ATCC, CCL-1, Lot #70008726) and human umbilical vein endothelial cells (HUVEC, Corning, 354151, Lot #0715305) according to ISO 10993-5.⁽³⁷⁾ The L929 cells were grown in medium comprising of Dulbecco’s modified Eagle Medium (DMEM, Gibco BRL, Grand Island, NY), 10% heat-inactivated fetal bovine serum (Cytiva, Marlborough, MA), and 1% penicillin/streptomycin (Thermo Fisher Scientific) at 37°C and 5% CO₂. HUVECs were grown in EGM™-2 Endothelial Cell Growth Medium-2 BulletKit™ at 37°C and 5% CO₂ (Lonza, Morristown, NJ). The L929 and HUVEC cells were seeded in their respective culture media in 96-well plates at a density of 5000 cells per wall followed by incubation at 37°C and 5% CO₂ for 24 hours. Then the growth medium was aspirated out and replaced with extraction media of embolic gel at 37°C for another 24 hours. Different concentrations of extraction media of embolic gels were prepared such as 100%, 50%, 25% and 12.5% in respective media. Cell viability was analyzed using WST-1 (Cayman Chemical, Ann Arbor, MI, USA) according to manufacturer’s protocol. 10% DMSO and non-treated cells were served as positive and negative controls, respectively. Three independent experiments with four replicates in each experiment were conducted.

2.11. Hemolysis test

The hemolysis rate of 8NC100 and 8NC50 was tested according to ISO 10993-4.⁽³⁸⁾ Citrated fresh porcine blood (Lot # 23D57081, catalog # 7204906, LAMPIRE Biological Laboratories, Pipersville, PA, USA) was diluted with 1X PBS (P4417, Sigma Aldrich, USA) in 4:1 ratio. Approximately 1 mL of the embolic gels was taken in a centrifuge tube containing 9 mL PBS. After pre-warming the PBS containing gels at 37°C for 30 min, 0.2 mL of diluted blood was added. The samples were incubated for an additional 1 hr at 37°C followed by centrifugation at 3000 rpm for 5 minutes. The supernatant was carefully transferred into a 96-well plate and the absorbance (A) was measured using a microplate reader (GENios, TECAN, Crailsheim, Germany) at wavelength 545 nm. 10 mL of PBS and deionized water incubated with 0.2 mL of diluted blood were used as negative and positive controls, respectively. The hemolysis percentage was calculated using the following formula:

$$Hemolysis\left(\mathrm{\%}\right)=\frac{A_{sample}-A_{negativecontrol}}{A_{positivecontrol}-A_{negativecontrol}}\times 100$$ Equation 3

2.12. Thrombogenicity

The interaction of the engineered embolic gels and blood was studied through thrombogenicity assay. Clotting time was quantified using a previously developed protocol.^(2),(10) Briefly, 100 μL of 8NC100 and 8NC50 gels were deposited at the bottom of 96 well plate, which was then centrifuged at 1000 rpm to standardize the blood interaction surface. Uncoagulated citrated whole porcine blood was activated by adding 10% (v/v) 0.1 M CaCl₂. A 100 μL of activated blood was added to each sample and allowed to react for 2, 4, 6, 8 and 10 min. At each time point, clotting was stopped by the addition of 100 μL of 0.109 M sodium citrate solution. Residual liquid was removed to isolate the blood clot. Porcine blood alone and clinically used coils (2D Helical – 35, Boston Scientific, Ireland) were used as controls.

2.13. Sterility test

The sterility of the embolic gel was tested. Approximately 1 ml of embolic gel was placed into a 15 ml tube filled with 10 ml of Luria-Bertani (LB) broth (Miller, Sigma-Aldrich, St. Louis, MO, USA). Plain LB broth was used as negative control. Escherichia coli (E. coli, ATCC, 25922) was used as a positive control. The samples were then kept in an orbital shaker (New Brunswick Scientific, Edison, NJ, USA) at 100 rpm at 37°C for 24 hours. The optical density of the suspension was measured using a microplate reader at 600 nm (Ultrospec 10 cell density meter; Amersham Biosciences, Little Chalfont, UK). Three readings were recorded for each sample.

2.14. Sterilization

The sterilization process on the gels was investigated. 8NC100 and 8NC50 gels were autoclaved (Steris Prevac Steam Sterilizer, USA) under liquid cycle for 25 minutes. The autoclaved gels were then subjected to rheological testing (amplitude sweep) (n=3) to assess the change in appearance and storage modulus of the gel.

2.15. Imageability of the embolic gels

The imageability of the 8NC50 gels under X-ray was studied using AMI HTX (Spectral instruments imaging, USA). Tantalum microparticles (Thermo Scientific Chemicals) of varying concentration 10, 20 and 30 wt% were incorporated into 8NC50 gel. Two concentrations (100% and 50%) of commercially used contrast agent, Omnipaque (350 mgI mL⁻¹, GE Healthcare, MA, USA), were used as positive controls. PBS and pristine 8NC50 gel were used as negative controls.

2.16. Shelf life

Freshly prepared 8NC50 was stored at 4°C for 60 days to evaluate the shelf life of the gel. After 60 days, the gel was subjected to rheological testing (amplitude sweep) (n=3) and injectability through 5F (100 cm length) and 1.8F (150 cm length) catheters (n=3) following previously described testing protocols.

2.17. Statistical Analysis

Statistical differences between multiple groups were calculated using analysis of variance (ANOVA) with Tukey post-test using GraphPad Prism 10 (GraphPad Software, CA, USA). P < 0.05 was considered to be significant. Data is reported as average ± standard deviation (S.D.), unless otherwise stated.

3. RESULTS

3.1. Characterization of NC and CH

The hydrodynamic diameter of NC measured using dynamic light scattering (DLS) is 40.1 ± 1.4 nm. Fig 1(a) shows the size distribution of NC. Two peaks were observed for NC at approximately 6 nm and 50 nm. This suggests that NC has two diffusion coefficients and can be explained on the basis of NC nano-disc orientation in relation to the vector of velocity. Those particles moving with their large fore-front against the solvent display small diffusion rates, while those particles moving with their thin edges against the solvent exhibit higher diffusion rates, i.e., a smaller apparent hydrodynamic diameter.⁽³⁹⁾ The polydispersity index (PDI) values indicate the width of the overall size distribution, with a value ranging between 0 (monodisperse) to 1 (polydisperse). The PDI of NC dispersion in molecular grade water is 0.257 ± 0.009, suggesting that the NC is relatively monodisperse with a narrow size distribution.⁽⁴⁰⁾ The size and morphology of CH nanofibers were investigated using SEM, showing varying length ranging from 3 to 30 μm (Fig 1(b)).

Fig 1.

Size distribution of (a) NC obtained from DLS and (b) CH nanofibers obtained from SEM.

Zeta potential of NC and CH was determined to understand their surface charges, which in turn governs their electrostatic interactions. Zeta potentials of the NC and CH were −45.1 ± 2.5 and 26.3 ± 4.3 mV respectively. The high magnitude of zeta potential (greater than positive 30 mV or less than negative 30 mV) of NC results in greater repulsion between the particles resulting in high degree of colloidal stability.⁽⁴¹⁾ Whereas the zeta potential value of CH indicates the eventual agglomeration of CH nanofibers due to interparticle interactions.⁽⁴¹⁾ In addition, the opposite charges lead to electrostatics interactions, which contribute to the formation of CH-NC gel system.

3.2. Rheology of hydrogels

Three groups of gels were engineered with a total solid content of 4% (4NC), 6% (6NC) and 8% (8NC) with varying ratio between NC and CH, with NC content at 100, 75, 50, 25 and 0% in each system (Table S1). The gels were subjected to various tests such as flow curves, amplitude sweep, frequency sweep and thixotropic tests to assess their rheological behavior.

The flow curve of 4, 6 and 8NC gels (Fig 2(a) – (c)) shows a drastic reduction in the viscosity with increasing shear rate, suggesting the shear thinning property of the gel which in-turn facilitates the transcatheter injectability. The flow behavior of the gel can be described by the fluid power law:

$$\eta =\text{K}{\stackrel{\cdot }{\gamma }}^{(\text{n}-1)}$$ Equation 4

Fig 2.

(a) – (c) Representative flow curve of 4NC, 6NC and 8NC gels showing decrease in viscosity with increase in shear rate; (d) – (f) representative amplitude sweep curve of 4NC, 6NC and 8NC respectively indicating the change in storage and G” with respective to shear strain; (g) – (i) G’ of 4NC, 6NC and 8NC respectively at 0.1% shear strain; (j) – (l) yield stress and yield strain of 4NC, 6NC and 8NC respectively. Data are presented as mean ± standard deviation (s.d.) (n = 3 for (g) – (l)).

where n is the flow behavior index and K is the flow consistency index. The value of n determines the flow property of the fluid. When n equals 1, the material behaves as a Newtonian fluid. When n is less than 1, there is a viscosity reduction at higher shear stress making the fluid shear thinning or pseudo fluids.⁽⁴²⁾ The n values are less than 1 (Table S2), further confirming all gels are shear thinning.

Amplitude sweep tests of 4, 6 and 8NC were carried out to determine their storage (or elastic) moduli (G’) and loss (or viscous) moduli (G”). The G’ quantifies the elastic energy stored in the material whereas the G” quantifies the viscous component, i.e. the energy dissipated when the material is deformed.⁽⁴³⁾ The amplitude sweep curves of 4, 6 and 8NC gels (Fig 2(d) – (f)) show that the G’ is higher than the G” at lower shear strain which indicates the solid like behavior of the gel. However, the G” dominates the G’ at higher shear strain indicating the dominance of viscous component, i.e., the material starts to flow. It is observed that with the increase in solid content in the gel, there is a significant increase in the G’ as well. The G’ of 6NC50 (730.0 ± 43.2 Pa) and 8NC50 (4652.9 ± 394.6 Pa) are 32 and 206 times respectively the G’ of 4NC50 (22.6 ± 2.4 Pa). It is interesting to note that pristine NC and CH samples (xNC100 and xNC0) have high G’. A reduction in the G’ is observed for gels containing the mixture of NC and CH (xNC25, xNC50 and xNC75), and the least is being observed for gel with xNC50 (50% NC- 50% CH) (Fig 2(g) – (i)).

The yield stress (i.e., stress at yield strain), a measure of the stress that has to be overcome to initiate steady material flow, was studied. It is reported that the mean shear stress experienced by the endothelial cell in the artery walls varies between 0.3 Pa in femoral arteries and 1.3 Pa in the carotid artery. Meanwhile, the shear stress experienced by arterioles walls are approximately around 1 to 5 Pa.^(44),(45) Fig 2(j) – (l) shows that the yield stress of all 6NC and 8NC samples are 4 to 62 times greater than the yield stress experienced by the artery walls. On the other hand, the yield stress of 4NC75, 4NC50, 4NC25 and 4NC0 are 6.3 ± 0.5, 0.3 ± 0.07, 0.5 ± 0.08 and 7.4 ± 0.3 Pa respectively, suggesting that the blood flow can induce a shear stress above their yield stress and leads to the failure of these gels. The general trend in all the cases is xNC0 (all CH) have the highest yield stress followed by xNC100 (all NC). With the addition of CH (xNC25, xNC50 and xNC75) the yield stress decreases. Also, xNC50 (50% CH and 50% NC) have the least yield stress in each group (e.g.: 8NC50 has the least yield stress among all gels in the 8NC group), similar to the trend observed in the amplitude sweep curve. The same trend was observed for yield strain.

The loss factor (tan δ = G’/G”) of the gels is also estimated from the amplitude sweep test. The value of tan δ is below 1 for all the samples of 4NC, 6NC and 8NC (Fig S2(a) – (c)) at lower shear strains indicating that the G’ is dominant over G” and the gel behaves elastically. Whereas, at higher shear strains (> 10%) the value of tan δ increases to a value greater than 1 indicating the dominance of G” over G’ and hence the gel exhibit fluid properties.

Frequency sweep tests of 4NC, 6NC and 8NC gels were carried out at 0.1% shear strain (LVE region) to predict the structural integrity and mechanical strength of the material with respect to the frequency of oscillations. The G’ of all 4NC, 6NC and 8NC gels shows predominance over the G”, indicating elastic behavior at all tested angular frequencies (0.1 – 100 rad/s) (Fig 3(a) – (c)).

Fig 3.

(a) – (c) Representative frequency sweeps of 4NC, 6NC and 8NC respectively; (d) – (f) Representative thixotropic tests of 4NC, 6NC and 8NC respectively.

Thixotropic studies were performed to understand the recoverability and stability of the gels when subjected to oscillatory shearing, mimicking intermittent injections. The thixotropic tests (Fig 3(d) –(f)) indicate that the gels recover to its initial state instantaneously when the high shear is removed. A slight loss of G’ is observed for all samples after the first shearing cycle. However, the recovery from the second shear cycle for all the samples are above 90% indicating a reconstruction of gel after the shear cycle. Furthermore, the recovery of G’ after the third and fourth shear cycle are also above 90% and instantaneous, indicating the stability and excellent recoverability of the gel after shearing (Table S3). The rheological tests indicate 8NC50 has the optimum mechanical properties compared to the other gel systems.

3.3. Morphology and Zeta potential of gel

The SEM images of 8NC100 gel revel flake like morphology of NC as shown in Fig S3. With the addition of CH into the NC matrix, the morphology changes to porous morphology as evident from Fig S3. In addition, the zeta potential of the final gel, as a mixture of NC and CH, was measured. Since 100% NC and 100% CH are negatively and positively charged respectively as reported in 3.1, 8NC50 gel was selected due to the one-to-one ratio of NC to CH. Its zeta potential was measured to be −29.3 ± 2.14 mV, which is in between of pure NC and pure CH.

3.4. Transcatheter injectability

The force required to inject the gel through a 5F 100 cm long catheter is measured to simulate the experience of physicians during clinical practices. For all the samples of 4NC, 6NC, and 8NC, the forces build up and plateau after a few seconds. The average of the constant force is considered as injection force (IF), which sustains the movement of plunger. In a few cases, the force builds up and attains a peak value and thereafter reduces and plateaus out to a constant injection force. The peak force attained is called break loose force (BLF), which is the force required to achieve the movement of syringe plunger. (Fig 4(a) – (c)). A maximum injection force of 79.8 N is considered optimum for the physicians to comfortably inject the gel through the catheter.⁽⁴⁶⁾ The general trend indicates that the increase in solid content in the gel system increases the injection force. The maximum force recorded for 4NC, 6NC (except 6NC0) and 8NC samples are 30, 40 and 85 N respectively (Fig 4(d) – (e)). However, the BLF and IF of 8NC50 gel are 53.1 ± 1.9 and 44.8 ± 1.7 N respectively (Fig 4(f)), which is considered as injectable. The injection force of 8NC0 was beyond the maximum load capacity of the instrument (200N) and was therefore excluded for further testing. It is also noted that not all gels exhibit BLF. In addition, the force required to push the embolic gels directly out of the syringe without any catheter attached was measured as shown in Fig 4(g) – (i). The overall required forces, ranging from 0.38 N to 1.7 N, are low (Fig 4(j) – (l)).

Fig 4.

(a) – (c) Representative injection force of 4NC, 6NC and 8NC respectively through a 5F 100cm long catheter; (d) – (f) The break-loose force and injection force of 4NC, 6NC and 8NC respectively through a 5F 100cm long catheter; (g) – (i) Representative injection force of 4NC, 6NC and 8NC respectively through a syringe; (j) – (l) The injection force of 4NC, 6NC and 8NC respectively through a syringe (no catheter attached). Data are presented as mean ± standard deviation (s.d.) (n = 3 for (d) – (f) and (j) – (l)).

Hence, based on the values of measured injection force (<79.8N) and the yield stress, which indicates the stability of the gel to withstand the shear stress induced by the blood, we selected 8NC50 for further characterization, together with 8NC100 as its pristine NC gel to understand the effect of CH.

3.5. Prediction of injection force

A correlation between the injection force and the rheological parameters of the gels (i.e., η, n and K) can facilitate the prediction of their transcatheter deliverability. The predicted injection force was calculated using two models by firstly studying the gel using a simplified Newtonian fluid model (Equation 5) and secondly as a power-law governing shear thinning material (Equation 6), using the following equations.⁽⁴⁷⁾

$$F=8.\eta {.l}_{catheter}.\frac{R_{syringe}^2}{R_{catheter}^2}+F_{friction}$$ Equation 5

$$F=2^{n+2}{.\pi }^{1-n}.l_{catheter}.R_{syringe}^2.K.{\left(\frac{\partial V}{\partial t}\right)}^n.R_{catheter}^{-\left(3n+1\right)}.{\left(\frac{3n+1}{2n+1}\right)}^{n-1}+F_{friction}$$ Equation 6

In particular, l_catheter is the length of the catheter, R_syringe is the radius of the syringe plunger and R_catheter is the radius of the catheter. The second term F_friction is the friction force experienced by the gel due to the syringe. It is obtained from the injection force required to push the gel out of a syringe without catheter connection. In Equation 5, η is the viscosity of the gel in the syringe or catheter at a given flow rate (2 mL/min) based on its flow curve. For Equation 6, (∂V/∂t) is the flow of hydrogel through the syringe. n and K are the parameters obtained from the fitting of linear region of the flow curve.

For both models, the predicted and measured injection forces are close (Fig 5 (a) – (b)) at lower values (below 60 N). In most cases, xNC100, xNC75, xNC50 and xNC25 exhibited good correlation between measured and predicted injection forces. However, predicted injection force of pristine CH gels (4NC0 and 6NC0) shows a greater deviation (designated with arrows in Fig 5) from the measured injection force. A deviation is also exhibited by 8NC100 where the predicted injection force is higher compared to the measured injection force (Fig 5 (b)).

Fig 5.

Predicted injection force using (a) Newtonian and (b) non-Newtonian models. The dashed line represents 100% agreement between measured and predicted injection force; (c) Representative injection force curve of 8NC100 and 8NC50 through a 1.8F 150 cm long microcatheter; (d) Break loose force and injection force of 8NC100 and 8NC50 through a 1.8F 150 cm long microcatheter. Data are presented as mean ± standard deviation (s.d.) (n = 3 for (d)).

3.6. Injection force through microcatheter

The flow potential of selected 8NC50 and 8NC100 formulations was further investigated through a clinically used 150 cm 1.8F catheter which has an approximate 500 μm inner diameter. This catheter was selected since it is commonly used to maneuver into finer neurovasculature for neurointerventional angiography. Similar to the injection profile observed in 5F catheter, the force builds up to a peak value, represented by BLF, and then plateaus to IF in 1.8F microcatheter. A high BLF of 81.3 ± 1.5 N is observed for 8NC100 with an IF of 66.1 ± 4.6 N (Fig 5(c) – (d)). A much less BLF and IF of 56.8 ± 1.4 N and 52.0 ± 1.3 N respectively is recorded for 8NC50 gel (Fig 5(c) – (d)).

3.7. In-vitro pressure test

The in-vitro pressure test was performed to study gel’s ability to withstand blood pressure. The selected gels, 8NC100 and 8NC50, were placed in a tube of internal diameter (ID) 2.3 mm (3/32”) and 4.8 mm (3/16”), which is equivalent to average ID of coronary artery⁽⁴⁸⁾ and femoral artery⁽⁴⁹⁾ respectively. It mimics the pressure experienced by the gels in the arteries. A maximum pressure of 0.08 ± 0.02 and 1.46 ± 0.01kPa was observed when porcine blood was passed through the tube of ID 2.3 and 4.8 mm respectively, without any gel (control) (Fig 6(a) – (b)). The pressure required to displace the 8NC100 and 8NC50 gels through a 2.3 mm ID tube was 62.9 ± 9.4 and 43.3 ± 2.1 kPa respectively. These results suggested that 8NC100 and 8NC50 can withstand approximately 4 times and 2.7 times the physiological blood pressure (16 kPa – systolic pressure). Similarly, a pressure of 25.4 ± 1.7 and 25.3 ± 6.4 kPa is required to displace 8NC100 and 8NC50 respectively through a 4.8 mm ID tube indicating that both gels can withstand approximately 1.5 times the physiological blood pressure respectively (Fig 6(a) – (b)), suggesting excellent mechanical stability to prevent distal migration.

Fig 6.

(a) The in-vitro pressure profile of 8NC100 and 8NC50 through a tube of 2.3 mm (3/32”) and 4.8 mm (3/16”) inner diameter using blood; (b) The maximum pressure required by the blood to displace the 8NC100 and 8NC50 gel through a tube of 2.3 mm (3/32”) and 4.8 mm (3/16”) inner diameter; (c) The degradation percentage of 8NC100 and 8NC50 in PBS solution at varying pH (4, 7 and 9); (d) The degree of swelling of 8NC100 and 8NC50 in PBS solution at varying pH (4, 7 and 9) . Data are presented as mean ± standard error of mean (s.e.m.) (n = 3 for (b) and n = 4 for (c) and (d)). Statistical significance was determined using one-way analysis of variance (ANOVA) with multiple comparisons against respective controls. ns – not significant; ***P< 0.001.

3.8. Degradation and Swelling

The degradation rate of 8NC50 and 8NC100 at three different pH values (4, 7 and 9) at 37°C was demonstrated in Fig 6(c). The results showed that both materials had minimal weight change over 24 hours, and no significant degradation occurred at any pH with respect to time. At 24 hours, no statistical significance was observed for 8NC100 (p>0.999), whereas a significant difference was reported for 8NC50 between pH value of 4 and 9 pH (***p<0.001).

The swelling kinetics and ratio of both formulas in different pH (4, 7 and 9) are shown in Fig 6(d). The results indicate that there was a continuous weight gain for 8NC50 and 8NC100 up to 10 minutes for all pH values. This was followed by a weight reduction up to 40 minutes, followed by minimal swelling change until 24 hours. The difference within the group of 8NC100 and 8NC50 of different pH at 24 hours was not significant at all pH values (p>0.999).

3.9. Cell viability

The biocompatibility of 8NC100 and 8NC50 was investigated using L929 fibroblast cells and HUVEC following the ISO 10993-5 protocol.⁽³⁷⁾ L929 is an FDA standard for testing cytotoxicity of biomedical devices. The L929 viability for 100, 50, 25 and 12.5% concentration of 8NC100 extracts was 92.9 ± 3.4, 95.0 ± 5.9, 104.0 ± 7.4 and 102.9 ± 9.3% respectively. For 100, 50, 25 and 12.5% concentration of 8NC50 extracts, the viability was measured to be 94.2 ± 12.0, 94.6 ± 10.9, 82.9 ± 1.2 and 88.0 ± 4.9% respectively (Fig 7(a)). Both gels are therefore considered to be biocompatible since the cell viability greater than 70%.⁽³⁷⁾ There is also no significant statistical difference within or between the two groups (p>0.05). This suggests that the cell viability is independent on the dilution of the material, further confirming the biocompatibility.

Fig 7.

(a) Cell viability of L-929 fibroblast cells suggesting biocompatibility of the material; (b) Cell viability of HUVEC cells showing reduced viability; (c) Hemolysis rate of 8NC100 and 8NC50 suggesting hemocompatibility of the materials; (d) Images of thrombogenicity assay of showing comparable clotting time when porcine blood was in contact with 8NC100 and 8NC50 compared to clinically used embolic coils; (e) Sterility of 8NC100 and 8NC50 gels; (f) X-ray intensity of 8NC50 with tantalum microparticle content increasing from 0 to 30 wt% compared to 50% and 100% concentration of Omnipaque 350. Data are presented as mean ± standard deviation (s.d) (n = 3 for (a), (b), (c), (e) and (f)). Statistical significance was determined using one-way analysis of variance (ANOVA) with multiple comparisons against respective controls. ns – not significant; *P<0.05, **P<0.01, ***P< 0.001, ****P<0.0001.

HUVEC was selected since it represents vascular endothelial cells that embolic agents will be in direct contact during embolization. The HUVEC viability for 100, 50, 25 and 12.5% concentration of 8NC100 extracts was 56.7 ± 3.9, 85.2 ± 14.4, 86.1 ± 6.2 and 89.2 ± 7.1% respectively. For 100, 50, 25 and 12.5% concentration of 8NC50 extracts, its relative cell viability was measured to be 50.4 ± 4.9, 77.6 ± 4.4, 85.1 ± 7.1 and 85.9 ± 5.5% respectively (Fig 7(b)). These results suggested that the gels could lead to reduced viability, or potentially diminished metabolic activity at the acute phase after seeding.

3.10. Hemocompatibility

The hemocompatibility of gels was investigated since it is an important factor as any embolic gel will come into direct contact with the blood. Following ISO standard 10993-4, the hemolysis rate was 0.58% ± 1.5% and 0.00% ± 0.51% for 8NC100 and 8NC50 respectively (Fig 7(c)). Since a hemolysis rate less than 5% indicates hemocompatibility,⁽³⁸⁾ both gels are blood compatible. A statistical significance is observed between 8NC100 and 8NC50 (*p<0.05).

3.11. Thrombogenicity

The thrombogenicity of 8NC50 and 8NC100 was evaluated using porcine blood that was loaded into a 96-well plate by estimating the clotting time of blood. Fig 7(d) shows that blood alone (control) coagulated at 6 min, and the blood in contact with 8NC50 and 8NC100 fully coagulated at 8 minutes, which is comparable with the clinically used coils (fully coagulated at 8 minutes).

3.12. Sterility Test

The sterility of engineered gels was tested. The OD₆₀₀ reading of growth media (LB as negative control) and E. coli (positive control) were 0.001 ± 0.001 and 1.107 ± 0.008 respectively. 8NC100 and 8NC50 showed OD₆₀₀ values of 0.002 ± 0.003 and 0.012 ± 0.006 respectively (Fig 7(e)), suggesting that both gels were inherently sterile. A statistical significance is observed for control LB broth, 8NC100 and 8NC50 against E-coli (****p<0.0001)

3.13. Sterilization

The sterilizability of the 8NC100 and 8NC50 gels were studied by monitoring the properties before and after autoclave. As shown in Fig S4 (a), both 8NC100 and 8NC50 gels broke down after autoclaving whereas no color change was observed. Rheological testing indicated an increase in the strength of the gels post sterilization (Fig S4 (b) – (c)). The storage modulus of 8NC100 before and after autoclave was 4586.7 ± 203.6 and 13857.0 ± 687.8 Pa respectively, and the values were 4652.9 ± 394.6 and 14665 ± 2124.7 Pa for 8NC50. An approximate 3-fold increase of storage modulus was observed for both gels after autoclaving.

3.14. X-ray Imageability

Clinically, the ability to visualize embolic agents on X-ray based imaging modality is critical for precise and targeted delivery. Hence, tantalum particles were included in 8NC50 as an example to demonstrate its tunable radiopacity for translational capability under X-ray. Fig 7(f) compares the grey scale intensity of 8NC50 loaded with 10, 20 and 30 wt% tantalum microparticles (a contrast medium in Onyx, an FDA approved liquid embolic agent) against FDA approved aqueous contrast agent Omnipaque 350 (OMNI, 350 mgI/mL). The grey scale intensity of 8NC50 −30% Ta and 8NC50 −20% Ta (239.1 ± 1.0 and 235.0 ± 1.0 respectively) was comparable to 100% OMNI (235.3 ± 0.4). In addition, comparable grey scale intensities are observed between 8NC50-10% Ta and 50% OMNI (207.8 ± 0.8 and 208.8 ± 1.0 respectively), which mimics Omnipaque 140 (140 mgI/mL, the lower end of iodine concentration in the Omnipaque series).

3.15. Shelf life

The properties of 8NC50 gel were monitored after storing the gel at 4°C for 60 days (2 months) on the evaluation of the shelf life. Fig S5 (a) – (b) indicated an increase in the storage modulus of 8NC50 from 4652.9 ± 394.5 Pa (D0) to 8312.4 ± 240.8 Pa (D60). The injection force of the gel through 5F and 1.8F catheter were also studied and found to be injectable. An increase in the injection force through 1.8F was observed for 8NC50 from 51.9 ± 1.3 (D0) to 68.9 ± 1.5 N (D60) (Fig S5 (c) – (d)). Interestingly, the injection force decreased in the case of 5F catheter from 44.81 ± 1.7 (D0) to 21.1 ± 0.7 N (D60) (Fig S5 (c) – (d)).

4. DISCUSSION

In this study we developed a novel shear thinning hydrogel comprising of NC and CH nanofibers for endovascular embolization, which can be easily delivered through a clinical catheter. We hypothesized that the negative zeta potential of NC and positive zeta potential of CH nanofibers can result in electrostatic force of interaction between the gel components resulting in a stable gel. Also, the well-established gelling property of NC⁽¹⁵⁾ along with the bioactive property of CH⁽²⁷⁾ results in a hydrogel with optimum mechanical properties and good biocompatibility for the successful occlusion of blood vessels. To achieve this, we first prepared hydrogels with 4, 6 and 8% total solid content (TSC) with ranging NC content of 100, 75, 50, 25 and 0% in each system. Our results suggested that 8NC50 has the optimal mechanical and injectable properties and is therefore selected for further investigation. 8NC100 was also included as the corresponding pristine NC gel to understand the effect of CH.

The surface morphology of the lyophilized gel obtained from SEM shows flake like structures for pure NC sample (8NC100) whereas with the addition of CH nanofibers into the system, the structure became porous. Since the gels are water based, SEM was taken on lyophilized specimen showing the preserved structure and volume after the removal of water content.

All engineered CH-NC gels exhibited shear thinning property, a key characteristic for transcatheter injection since it allows smooth delivery of gel through the catheter as a liquid and solidifies back in the bloodstream upon the removal of force for blood vessel occlusion. Next, storage modulus, the elastic energy stored in the material or in other words the solid-like nature of the gel, serves as an important factor to indicate gel’s strength and stability. It has been suggested that the initial G’ of the embolic agent needs to be higher than 800 Pa to achieve a pressure resistance up to 200 mmHg.⁽⁵⁰⁾ In addition, the yield stress of the gel should be greater than the stress induced by the blood on the gel to prevent deformation and migration in the body. The high yield stress values of 6NC and 8NC gels ensure the reliability of the gel to stay in the desired location and thereby reducing the chance of recanalization and increasing the embolic efficiency. 4NC group is therefore excluded.

It is important to note that all 50% CH samples, i.e., 4NC50, 6NC50 and 8NC50 exhibit a lower G’ compared to other samples in the same group. We hypothesize that this might be due to the electrostatic interaction between the anisotropically charged NC and positively charged CH nanofibers. For 100% NC gels, i.e., 4NC100, 6NC100 and 8NC100, the disc shaped NC orient themselves what is called as “house of cards” (Fig 8(a)). This structure can deform and then re-form upon oscillating shear. As a result, the high level of electrostatic interaction between the NC discs results in a high G’ of the xNC100 gels. On the other hand, 100% CH samples, i.e., 4NC0, 6NC0 and 8NC0, form an entangled chain structure (Fig 8(e)) which results in high G’ for xNC0 gels. For xNC25, it is hypothesized that the “house of cards” structure of the NC is still preserved to a large extent, whereas the positively charged CH nanofibers interact with the negative faces of the nanodiscs and attempt to intercalate the NC structure (Fig 8(b)). The xNC50 samples are likely to have an exfoliated nanoclay structure due to the presence of CH nanofibers. The CH nanofibers penetrate through the structured orientation of nanoclay thereby reducing the interaction in between. Furthermore, the entanglement of the CH nanofibers may be compromised for xNC50 samples (Fig 8(c)), leading to the low G’ of the xNC50. Lastly, the xNC25 gels are believed to have entangled chains of CH nanofiber whereas the inherent “house of cards” structure of NC is highly interrupted (Fig 8(d)). The dominated structure from entangled chains of CH nanofiber, therefore, recovered G’ of xNC25 as compared to xNC50. While the detailed mechanisms will be further investigated, the implication of these results is significant. The reduction in G’ even for high total solid content CH-NC system helps to engineer an embolic gel with low viscosity, making the transcatheter injectability smooth and comfortable for the physicians. The unique property of 50% CH samples with balanced active components and sites in the gel is beneficial for the drugs/contrast agents carrying for therapeutic activities, which will be investigated in future studies.

Fig 8.

Schematic representation of NC and CH nanofibers in (a) xNC100, (b) xNC75, (c) xNC50, (d) xNC25, and (e) xNC0.

In practice, physicians inject the gel in multiple intermittent steps which can result in a recurring stress on the gel. It is important to ensure that the gel doesn’t lose its integral properties after being subjected to a series of injection activities. The thixotropic test mimics this scenario and measures the change in G’ after each cyclic stress. No significant reduction in the G’ was observed for any of the gels confirming the mechanical properties of the gel was not compromised under interrupted and repeated injections.

Although the rheological properties affect materials’ injectability, the injection force, a key parameter that is directly related to physicians’ experience, suggests whether embolics are suitable for clinical use.

Studies have shown that the average maximum force which can be applied for injection through a syringe is estimated to be 79.8 N; the value decreases for females (64.1 N) and increases for males (95.4 N).⁽⁴⁶⁾ As expected, we also observed an increase in the injection force values with increases in TSC and xNC50 showed the lowest injection force within the group. Moreover, 8NC50 formulation that can be delivered via 1.8F catheter (~55N) for neurointervention can also be injected through other microsized microcatheters, such as 2.0-2.8F, which are commonly used by non-neuro-interventional radiologists. Its NC counterpart, 8NC100, was unable to pass 1.8F catheter without overcoming the high BLF (>81N) that may become not injectable. The reduced viscosity and injection force of xNC50 gels aids in the flow of the gel through small blood vessels and occluding them. This unique property of the engineered embolic gel can be utilized for occluding the fine vasculatures feeding tumors, benign uterine fibroids or arteriovenous malformation (AVM).

The ability to predict injectability is of great importance for the rational design of embolics. To our knowledge, only few models have been developed so far to predict the injection force through a syringe and all of them focus on needle-based injection rather than through a catheter. A catheter, being a long narrow tube, generates more resistance to flow and hence greater will be the injection force compared to a needle. For both Newtonian and power-law models, the predicted injection force values were in good agreement with measured injection force at values below 60N (i.e., desired injectability). A perfect correlation between the measured and predicted injection force was not achieved in our study due to several factors. The power law and Hagen-Poiseuille law, used to predict the injection force in our study, do not address the complexities of the CH-NC system completely. The power law is for ideal shear thinning fluids. However, the gels exhibited a slight deviation in the slope (K) of the flow curve (Fig 2(a) – (c)) which in turn resulted in a significant change in the value of viscosity due to the logarithmic nature of the power law equation. Such deviation, therefore, affects the predicted injection force value, as it is a critical parameter in the equation (Equation 6). An R² value of 0.75 (Fig 5 (b)) between the predicted and measured injection force indicates an overall good correlation when the system was predicted using a non-Newtonian model. However, an improved model capable of capturing the complexities of CH-NC needs to be developed for better prediction of the injection force based on the rheological parameters.

The ability to withstand blood pressure and at the same time to provide successful occlusion is a key parameter for any embolic agent. Distal migration can lead the embolic agents ending in either brain or lungs, causing stroke. We tried to mimic the embolization of coronary (1.7 – 2.3 mm internal diameter) and femoral arteries (2.7 – 9.6 mm internal diameter) by occluding tubes with similar internal diameters.^(48),(49) Our results showed that the pressure required to displace the embolic gels (8NC100 and 8NC50) is much higher than the systolic pressure (16 kPa). It is anticipated that the decreased diameter of the tube increases the required pressure to displace the same amount of gel. The in-vitro pressure test suggests the in-vitro mechanical stability of engineered embolic gels, reducing the chance of recanalization.

Degradation and swelling rates were investigated. No considerable degradation of the 8NC50 or 8NC100 at three different pH values (4, 7 and 9) was observed, suggesting the stability of the gels in acidic, basic and neutral environments. The swelling studies showed a maximum swelling ratio at 10 minutes for both materials at all pH conditions due to osmotic pressure driven water penetration from the surrounding media. The difference between swelling ratios at 24 hours for both gels at three pH values was insignificant.

Following ISO protocols on hemocompatibility and cell viability, our results demonstrated the overall in vitro biocompatibility. A less than 70% viability of HUVEC may suggest potential reduced cell viability at the acute phase after seeding. However, it is important to acknowledge that endovascular procedures involving manipulation of stents and catheters are known to induce vascular injury for both transient and long-term endothelial denudation.^(51)–(53) In addition, clinically used liquid embolic agents such as Onyx^® contain toxic organic solvents DMSO, which served as the positive control in our cell viability study. Therefore, the precise interactions between CH-NC and tissues need to be further investigated in vivo, which is planned in the future scope of this work.

Several considerations on the translational applicability on the CH-NC system were presented, including sterilization, shelf-life and imageability. Besides the inherent sterility of 8NC50 and 8NC100, terminal sterilization is necessary to ensure patient safety to use embolics in a clinical setting. Autoclaving was used to sterilize the materials though had negative consequences on CH-NC gels.⁽⁵⁴⁾ The water loss after autoclaving led to the disruption of the gel structure, which translates to a significant change in the mechanical properties (i.e., enhanced storage modulus). Alternative postprocessing sterilization techniques, including gamma-irradiation, ethanol washing, and ultraviolet light,⁽⁵⁵⁾ will be investigated to maintain the structure, physical and biological properties of CH-NC for their use in biomedical applications. The ability to track the material under X-ray based imaging modality is crucial for physicians on the precise targeting and tracking of deployed embolics. A contrast agent, tantalum (Ta) microparticles, was mixed with 8NC50. A 20% Ta inclusion exhibited comparable grey scale intensity as that of 350 mgI/mL Omnipaque, a clinically used aqueous contrast agent, demonstrating the potential imaging capability of the engineered embolics. Initial shelf-life studies were carried out on 8NC50 which has been stored for two months. The material exhibited slightly higher storage modulus and injection force (but within the comfortable range) compared to the freshly prepared gel. Despite the main focus of this work being on the development and characterization of the new CH-NC embolic platform, a systemic shelf-life study needs to be conducted to guarantee its features during storage, making it feasible for medical industry applications.

5. LIMITATION AND FUTURE WORK

Before closing, we acknowledge that this work has several limitations. The flow rate for the migration study was 50 mL/min, whereas the actual flow rate of blood inside the body can be greater than 250 mL/min.⁽⁵⁶⁾ For fluids the faster it flows (higher flow rate), higher shear stress (friction) presents against the wall, and hence higher pressure is needed. The ability of the gel exhibiting adequate mechanical strength at tested lower flow rates suggested its capability to withstand higher pressure at elevated flow rates. Additional experimentation needs to be performed under flow conditions that more accurately simulate those present in vivo. Biomimetic phantoms (i.e., tumor vasculature and AVM), integrated with dynamic flow conditions within the vasculature, will further serve as a valuable platform to simulate various physiological conditions and real-time embolization delivery process, providing a more comprehensive assessment of CH-NC performance under different scenarios.

In summary, the innovative CH-NC system was generated to serve as a potential candidate for transcatheter embolization. As the medical field transition from conventional open surgery to minimally invasive procedure, it is of utmost importance to diverge the application of embolic gels to therapeutic, real-time imaging capability, and flexibility in addition to occlusion. Clinically relevant in-vivo embolization study (i.e., porcine rete mirabile as an AVM model) will be carried out to investigate the embolic efficacy and tissue remodeling process, as well as uncovering the mode of excretion of the degraded CH-NC by-products.

CONCLUSIONS:

In this work, we developed a CH-NC nanofiber-based system as an embolic agent for endovascular embolization. The embolic gel was shown to exhibit excellent shear thinning, mechanical strength, and recoverable properties. Integrated theoretical and experimental injectability approaches, for the first time, provide a platform for the prediction of practical deliverability of newly engineered gel embolics through clinically used catheters. In addition, the in-vitro pressure test confirms the strength and stability of engineered gels, suggesting that the chances of recanalization are low. The gels also demonstrated outstanding biocompatibility, sterility, thrombogenicity, hemocompatibility, and imageability making them excellent potential candidates for embolic agents for transcatheter based minimally invasive interventions.