Mechanical Investigation of a Tandem Embolization-visualization System for Minimally Invasive Procedures

Abstract

Transcatheter arterial embolization is a minimally invasive intervention process in which the blood supply to a tumor or an abnormal area of tissue is blocked. One of the most commonly used embolic agents in clinics is microsphere (MS). In order to understand the flow behavior of microspheres in arteries, it is essential to study their mechanical properties systematically. In this work, calcium-alginate MSs with varying calcium concentrations were synthesized using a coaxial airflow method. Indocyanine green (ICG) was added as a fluorescent dye. The effect of ICG concentration change on microspheres was investigated by studying morphology, imageability, rheology, and swelling behavior. Then the effect of calcium chloride concentration change on microspheres was studied by conducting rheological tests, atomic force microscopy tests, hemolysis assay, and thrombogenicity assay. Results showed that microspheres with higher ICG concentrations have longer lasting fluorescence and lower storage modulus (G’). Higher concentrations of calcium chloride led to higher G’, while the local Young’s modulus obtained by AFM test was not significantly affected. The MSs with and without ICG showed good hemocompatibility and thrombogenicity.

Graphical Abstract

1. Introduction

Transcatheter arterial embolization (TAE) is a minimally invasive intervention technique that has been widely used in many vascular applications, such as hemorrhagic lesions, aneurysms, pseudoaneurysms, arteriovenous malformations (AVM), and solid tumors [1]. This is due to their lower damage and risk, less pain, shorter hospital stays and fewer complications in comparison to open surgeries [2, 3]. During the TAE process, embolic agents are injected into blood vessels through a microcatheter where they can occlude the blood flow for treatment purposes. Microsphere (MS) as one of the embolic agents has gained broad attention due to the controllable size, low risk of catheter occlusion [4], and excellent capability as therapeutic carriers [5].

MSs, among the most commonly utilized embolic agents in clinics, are available in various materials such as PVA and trisacryl gelatin, with sizes ranging from 50 to 1200 μm [6]. These variations allow for strategic selection tailored to treating diverse conditions, including AVMs and hypervascular tumors. The precise size of the particles is paramount for targeted embolization as the delivery relies on blood flow, and the in vivo movement and accumulation are size-dependent. Commercial MSs have certain limitations, including the risk of being washed away, lack of intrinsic radiopacity for X-ray visualization, and absence of therapeutic properties. Despite sharing similar morphologies, microspherical embolic agents exhibit differing physical and mechanical properties due to variances in chemical composition and manufacturing processes. These differences influence interactions between microspheres and tissues, consequently impacting clinical outcomes. Currently, no comprehensive platform exists to explore the relationship between these properties and embolic outcomes systematically.

Near-infrared (NIR) imaging with fluorescence dyes has been applied as an imaging technique due to its sensitive detectability with small amounts of agent dosage [7]. Indocyanine green (ICG), an FDA-approved fluorescence dye has been utilized as an indicator in cardiac, circulatory, hepatic, and ophthalmic applications [8]. A major limitation of ICG is that the retention time in human body is low. According to Kawaguchi et al. [9], the retention time of 95% of ICG in liver is 15 minutes after injection. The fluorescence can be detected for up to 6 hours in liver and bile [10]. Lee et al. [11] used ICG in alginate hydrogel as a marker for precise laparoscopic operations and the fluorescence was detectable up to 3 days. Fan et al. [12] encapsulated ICG in poly lactic acid-co-glycolic acid MSs to slow down photodegradation and increase photothermal stability.

This study aims to create an MS system with uniform size, imageability, tunable mechanical properties, and biocompatibility. Currently there are several types of materials to produce embolic MSs used in the clinic. Acrylamido-polyvinylalcohol hydrogel has been used as the main composition of LC Bead (BTG), LC Bead LUMI (BTG), Bead Block (BTG), LifePearl (Terumo), and DC Bead (BTG). Gelatin has been used to produce Glefoam (Pfizer), Gel-Bead (Vascular Solutions), Embosphere (Merit Medical). Embocept (Pharmacept) uses Amilomer (hydrolyzed potato starch). Acrylic acid has been utilized to synthesize Embozene (Boston Scientific), HepaSphere/QuadraSphere (Merit Medical), and Tandem/Oncozene (Boston Scientific) [6]. For this work, calcium alginate was selected due to its tunability, biocompatibility [13], degradability [14–17], drug deliverability, and low cost. Many techniques have been developed to fabricate MSs [18], such as spray-drying [19], extrusion [20], emulsification [21], etc. A one-step scalable coaxial airflow technique was employed in this study to produce imageable calcium-alginate MSs due to ease of operation, controllable and consistent sizes of MSs, avoidance of other material (i.e., oil), and high efficiency.

To further study the flow behavior of MSs as an embolic agent working in arteries, a systematic investigation of their mechanical properties is essential. However, there are only a few research focusing on the mechanical properties of MSs. Hidaka et al. [4, 22] evaluated the elasticity and viscoelasticity of MSs by performing compression tests on both single and a monolayer of trisacryl gelatin MSs. Yang, et al. [23] tested the elastic and deformable properties (using a microparticle strength tester) as well as surface adhesion forces (using an atomic force microscope (AFM)) of MSs made of poly(lactic-co-glycolic acid), chitosan, calcium alginate, and poly(vinyl alcohol).

In this work, MSs made with sodium alginate and varying concentrations of calcium chloride and ICG were synthesized using a coaxial airflow technique. Morphology and size distribution of MSs with changing ICG concentration were investigated. Fluorescence intensity was monitored over 28 days for optimized ICG concentration. Rheological studies were performed to evaluate the effect of calcium chloride and ICG concentrations on the mechanics of MSs. The local Young’s modulus of MSs was assessed using a force microscopy technique. Hemolysis and thrombogenicity assays were also conducted to evaluate the interaction of MSs with blood. This study highlights the imageability of ICG loaded Ca-alginate MSs with tunable mechanical properties and the systematical investigation of mechanical properties as guidelines for evaluation of MSs in the TAE process.

2. Experimental Section

2.1. Materials

Alginate acid sodium salt (medium viscosity), calcium chloride dihydrate, and phosphate buffered saline tablet were purchased from Sigma-Aldrich (St. Louis, MO). Indocyanine green was purchased from Pfaltz & Bauer (Waterbury, CT). Citric acid trisodium salt dihydrate was purchased from VWR Life Science (Solon, OH).

2.2. Synthesis of calcium-alginate MS

Calcium-alginate MS were synthesized using a co-axial needle technique, as shown in Figure 1. 1 wt.% sodium alginate solution with 0.00, 0.10, 0.25, 0.50, 0.75, and 1.00 mg/mL ICG were first prepared in deionized water. The solution was mixed using a speed mixer (DAC 330–100 SE, FlackTek, Landrum, SC.) at 2000 rpm until the sodium alginate was completely dissolved. Calcium chloride (CaCl₂) solutions of 0.5, 1, 2, and 4 wt.% were also prepared.

Figure 1.

Schematic of MS synthesis apparatus

For MS synthesis, sodium alginate-ICG solution was passed through the inner needle of a 21/30 G concentric needle using a syringe pump (LEGATO100, kdScientific, Holliston, MA.) at 0.05 mL/min. Air flowed through the outer needle at 0.1 L/min to blow the sodium alginate solution droplets down into the calcium chloride solution in the collecting tray below to form MSs. The distance between the needle tip and the surface of the sodium chloride solution was kept at 10 cm. Different MSs were produced based on varying reacting reagents as described above.

Produced MSs were named as x-Alg-y-Ca-z-ICG, where x, y, and z denote the concentrations of sodium alginate (in wt.%), concentrations of calcium chloride (in wt.%), and concentrations of ICG (in mg/mL). Table 1 shows all the types of MS synthesized in this work.

Table 1.

Concentrations of solutions used for the synthesis of MSs.

	Sodium alginate (wt.%)	Calcium chloride (wt.%)	Indocyanine green (mg/mL)
1Alg0.5Ca1ICG	1	0.5	1.00
1Alg1Ca1ICG	1	1	1.00
1Alg2Ca1ICG	1	2	1.00
1Alg4Ca1ICG	1	4	1.00
1Alg2Ca0ICG	1	2	0.00
1Alg2Ca0.1ICG	1	2	0.10
1Alg2Ca0.25ICG	1	2	0.25
1Alg2Ca0.5ICG	1	2	0.50
1Alg2Ca0.75ICG	1	2	0.75

2.2. Characterization of MSs

2.2.1. Size and morphology distribution

The morphology and sizes of MSs were characterized using an inverted microscope (Fisherbrand, Waltham, MA). The images were analyzed using ImageJ (NIH, Bethesda, MD) [24] for the diameter, circularity, and roundness. Roundness and circularity of the MS were obtained from the following equations [24] using ImageJ:

$$\text{Circularity}=\frac{4\pi \text{A}}{{\text{P}}^2}$$ Equation 1

$$\text{Roundness}=\frac{4\text{A}}{\pi \times \text{major}\_{\text{axis}}^2}$$ Equation 2

where A stands for area of the circle and P denotes its perimeter. The roundness and circularity of a perfect circle are both 1. Ten optical microscopic images of each type of MSs were taken at random locations. The total number of each type of MS analyzed was between 175 to 376.

2.2.2. Fluorescence study

To investigate the fluorescence intensity (FI) change with respect to the change of time and ICG concentrations, 1Alg2Ca0ICG, 1Alg2Ca0.1ICG, 1Alg2Ca0.25ICG, 1Alg2Ca0.5ICG, and 1Alg2Ca1ICG were tested using an In Vivo Imaging System (IVIS) (Xenogen, Caliper Life Sciences, Hopkinton, MA.) with an ICG filter. In particular, the excitation and emission wavelengths were 745 nm and 840 nm, respectively. The exposure time was 1 second. 100 μL MSs were added to the bottom of the 96 well plate (black plate) and 4 tests were run per each sample. The ICG signals were measured at room temperature (25 °C) on day 0, 1, 2, 3, 4, 5, 6, 7, 14, 21, and 28 after the MSs were synthesized. The FI was then analyzed using Living Imaging software (PerkinElmer, Waltham, MA).

2.2.3. Swelling

For swelling, 1Alg2Ca0.1ICG, 1Alg2Ca0.25ICG, 1Alg2Ca0.5ICG, 1Alg2Ca0.75ICG, and 1Alg2Ca1ICG were immersed in 1X PBS solution at room temperature. The sizes were measured at 0h, 24h, and 48h. 10 optical microscopic images of each type of MSs were taken at random locations. The total number of each type of MSs analyzed was between 42 to 178.

2.2.4. Rheological properties

Rheological properties of MSs were studied using a strain-controlled rheometer (MCR 302e, Anton Paar, Graz, Austria) with sandblasted upper (8 mm diameter) and lower plates at 25 °C. MSs were added onto the lower plate and any excess liquid was carefully removed to ensure the distribution of a monolayer on the lower plate. Due to slight variation in particles sizes, instead of using a fixed gap between the measuring geometries, the tests were conducted when the normal force between the upper plate and the monolayer was 0.10 N to ensure proper contact. Large amplitude shear oscillation (LAOS) tests were conducted to identify the linear viscoelastic regime. The angular frequency was set at 10 rad/s and the shear strains were set from 0.01 to 100%. Each test was run in triplicate. The viscoelastic characteristics were then studied using frequency sweep (FS) tests at constant shear strain of 0.04% (linear viscoelastic regime as determined from LAOS tests) and the angular frequency were set from 0.1 to 100 rad/s. All the rheological tests were run in triplicates.

2.3. Atomic Force Microscopy (AFM)

After testing the properties of MSs with different ICG contents, 1.00 mg/mL was selected as the optimal ICG concentration. To evaluate the impact of cross-linking density on the mechanical properties, 1Alg1Ca1ICG, 1Alg2Ca1ICG, and 1Alg4Ca1ICG were assessed using an atomic force microscopy (AFM) (Nanosurf FlexAFM, Liestal, Switzerland) coupled with an inverted microscope (Leica DMi8, Wetzlar, Germany). The Young’s Modulus, E, of MSs were probed (ContGD-G, Budget Sensors, Sofia, Bulgaria) in 1X PBS. The tip of the probe was manually moved with the x, y stage at the apex of the MS. A total of 25 points were tested within each MS and a total of 30 MSs were tested per condition. The set point force was 5 nN.

Force-distance curves obtained from the C3000 software (Nanosurf, Liestal, Switzerland) were then analyzed using Atomic J software [25] to obtain E values. The curves were fitted using the pyramid model (Equation 3) [26] in Atomic J.

$$P=\frac{1.4906\mathit{\text{Etan}}\left[\theta \right]}{2\left(1-v^2\right)}{\delta }^2$$ Equation 3

In Equation 3, P stands for the load force and δ is the indentation depth. E is the Young’s Modulus to be calculated. The inclination of polyhedral face of the probe θ (also known as half-angle) was 10°. The Poisson’s ratio ν was selected as 0.5 [27].

2.3.1. Hemolysis and thrombogenicity

Hemolysis and thrombogenicity assays were conducted to evaluate the interactions between MSs and blood. The hemolysis rate was assessed following the ISO 10993-4 standard [28]. Citrated fresh porcine blood (Lot # 2Ek52883, catalog # 7204906, LAMPIRE Biological Laboratories, Pipersville, PA) was mixed with 1X PBS at a ratio of 4:5. 0.5 g of each type of MSs (1Alg2Ca0ICG, 1Alg1Ca1ICG, 1Alg2Ca1ICG, and 1Alg4Ca1ICG) was added with PBS to 10 mL (concentration of MSs is 50 mg/mL) correspondingly in a centrifuge tube and incubated at 37 °C for 30 minutes. 0.2 mL of mixed blood was then added to the centrifuge tube and incubated again at 37 °C for 60 minutes and centrifuged at 3000 rpm for 5 minutes. The resulting supernatant was transferred to a 96-well plate and measured in a microplate reader (GENios, TECAN, Crailsheim, Germany) for absorbance (A) at a wavelength of 545 nm. The positive and negative controls were prepared by adding 0.2 mL of mixed blood to 10 mL of DI water and 10 mL of PBS respectively. Equation 4 [29] is used to calculate the hemolysis rate:

$$\text{Hemolysis}(\%)=\frac{{\text{A}}_{\text{sample}}-{\text{A}}_{\text{negative}\text{control}}}{{\text{A}}_{\text{postive}\text{control}}-{\text{A}}_{\text{negative}\text{control}}}\times 100$$ Equation 4

A thrombogenicity assay developed from our previous studies [30, 31] was conducted to assess the capability of MSs to promote clotting. 100 μL of each type of MSs (1Alg2Ca0ICG, 1Alg1Ca1ICG, 1Alg2Ca1ICG, and 1Alg4Ca1ICG) was added into a 96-well plate followed by addition of 100 μL of fresh porcine blood (activated by adding 10% (v/v) 0.1M CaCl₂ solution). 100 μL of 0.109 M sodium citrate was then added to the mixture at 2, 4, 5.5, 6.5, and 7.5 minutes to stop clotting. Clinically used coils (2D Helical—35, Boston Scientific, Ireland) and porcine blood alone were utilized as the positive and negative controls for comparison.

2.4. Statistical Analysis

One-way analysis of variance (ANOVA) in GraphPad Prism 10 (GraphPad Software, CA, USA) was utilized to calculate statistical differences between groups. P < 0.05 was regarded as significant. Data are presented as mean ± standard deviation (S.D.) or mean ± standard error of mean (S.E.M.), unless otherwise stated.

3. Results and Discussion

3.1. Physical, imaging, and mechanical properties of MSs with varying ICG concentrations

3.1.1. Size and morphology distribution

The spherical shape of MSs is important in TAE process, as non-spherical particulates tend to interlock, block the catheter, and are difficult to be carried by blood flow to distal locations [6]. Moreover, a narrow size distribution is critical because it can accurately determine the size of vessels that can be embolized [6]. The synthesis conditions were tuned so that the sizes of MSs were close to 300 μm, which is one of the clinically commonly used sizes[6]. Figure 2 showed the size and morphology distribution of MSs with respect to different ICG concentrations. In order to optimize the ICG concentration, MSs with ICG concentrations from 0 mg/mL up to 1.0 mg/mL (the solubility of ICG in water) were synthesized and tested. The average diameters of MSs with different amounts of ICG added ranged from 265.0 ± 10.7 μm to 286.7 ± 4.0 μm, showing excellent size consistency. For all produced particles, the upper bound (307.7 ± 7.0 μm for MSs with no ICG) and the lower bound (265.0 ± 10.7 μm for MSs with 0.25 mg/mL ICG) of the diameter differed by 16%. The standard deviation of the diameters for each type of MS was within 5%, indicating a narrow size distribution. The obtained roundness and circularity of MSs ranged from 0.93 ± 0.07 to 0.98 ± 0.02 and from 0.87 ± 0.07 to 0.90 ± 0.01, suggesting all produced MSs had good spherical shape (perfect spheres have both values at 1).

Figure 2.

Size and morphology distribution of MSs. A) Mean diameters; B) roundness, and C) circularity of 1Alg2Ca0ICG, 1Alg2Ca0.1ICG, 1Alg2Ca0.25ICG, 1Alg2Ca0.5ICG, 1Alg2Ca0.75ICG, and 1Alg2Ca1ICG MSs. Optical images of the obtained MSs with different ICG concentrations, D) 0.00 mg/mL, E) 0.10 mg/mL, F) 0.25 mg/mL, G) 0.50 mg/mL, H) 0.75 mg/mL, and I) 1.00 mg/mL. Data are presented as mean ± SD (n = 175, 267, 376, 188, 193, 246 for 0.00, 0.10, 0.25, 0.50, 0.75, 1.00 mg/mL ICG respectively for A, B, and C).

3.1.2. In vitro Fluorescence

To investigate the effect of ICG on imageability of MSs over time, FI were monitored for 4 weeks in vitro. For imaging consistency (Figure 3A), the upper and lower limits of the images were set at 0.5×10⁷ and $1.5\times {10}^7\frac{{\text{p}/\text{sec}/\text{cm}}^2/\text{sr}}{\mu {\text{W}/\text{cm}}^2}$, respectively. As shown in Figure 3B, all MSs with ICG had strong mean FIs at around $8\times {10}^6\frac{{\text{p}/\text{sec}/\text{cm}}^2/\text{sr}}{\mu {\text{W}/\text{cm}}^2}$ on day 0. The FIs decreased significantly on day 2 and then remained at certain levels for one week. After that, the FIs continued to decrease until all the MSs with ICG had comparable signal as 1Alg2Ca0ICG on day 28. This indicated that the ICG of MSs remained detectable for 3 weeks, which also meant that the retention time of ICG inside the body could be largely extended by encapsulating it into MSs and as a marker for embolization. Compared to 1Alg2Ca0.1ICG, 1Alg2Ca0.25ICG, and 1Alg2Ca0.5ICG, 1Alg2Ca1ICG exhibited the highest ICG concentration, and showed the highest FI on day 0, 4, 6, 7, 10, 14, 21. Initially (day 0), 1Alg2Ca1ICG had the same FI as 1Alg2Ca0.25ICG and 15% and 5% higher than 1Alg2Ca0.1ICG and 1Alg2Ca0.5ICG (FI of 1Alg2Ca0ICG as the control was deducted), respectively. On day 21, when the FI of all the MSs with ICG attenuated to around $3\times {10}^6\frac{{\text{p/sec/cm}}^2\text{/sr}}{\mu {\text{W/cm}}^2}$, the FI of 1Alg2Ca1ICG was still 63% higher than 1Alg2Ca0.1ICG and 1Alg2Ca0.5ICG, and 72% higher than 1Alg2Ca0.25ICG. Therefore, 1Alg2Ca1ICG exhibited the best imageability and 1.00 mg/mL ICG was selected as the optimal ICG concentration to be applied for the rest of the study.

Figure 3.

Fluorescence signal of MSs (1Alg2Ca0ICG, 1Alg2Ca0.1ICG, 1Alg2Ca0.25ICG, 1Alg2Ca0.5ICG, and 1Alg2Ca1ICG) at Day 0, 1, 2, 3, 4, 5, 6, 7, 10, 14, 21, 28. A) Representative fluorescence imaging of MSs obtained at each time point using IVIS; B) Quantitative analysis of fluorescence intensity of each MS per time point. Data are presented as mean ± S.E.M. (standard error of mean) (n = 4 for B).

3.1.3. Rheological tests on MSs with different ICG concentrations

Rheological properties were investigated on a monolayer of each type of MSs to assess mechanical behavior of MSs under shear, which will be experienced by MSs in the blood vessels. Amplitude sweep was conducted first to confirm the linear viscoelastic regime (LVE) of the MSs. Figure 4A showed stable storage modulus, G’, value between the shear strain of 0.01% and 0.2% as the LVE range for all MSs. Hence, shear strain at 0.04% was selected to further carry out frequency sweeps. The selection of a small value of shear strain was to ensure that the MSs would not be damaged due to the deformation during the frequency sweeps.

Figure 4.

Rheology tests on 1Alg2Ca0ICG, 1Alg2Ca0.1ICG, 1Alg2Ca0.25ICG, 1Alg2Ca0.5ICG, 1Alg2Ca0.75ICG, and 1Alg2Ca1ICG. A) Amplitude sweep; B) Summary of storage modulus of each MSs; C) Frequency sweep tests; D) Loss factor of MSs obtained from frequency sweep tests. Statistical significance was determined using one-way ANOVA with multiple comparison. ns = not significant; * p ≤ 0.05, ** p ≤ 0.01, **** p ≤ 0.0001. Data are presented as mean ± SD (n = 3 for B).

Figure 4B shows the G’ of MSs with varying ICG concentrations corresponding to the 0.04% shear strain. G’ represents the elastic component of the material. Higher G’ means the material can store more elastic energy under the shear force. The G’ of 1Alg2Ca0ICG (no ICG) was 32.2 ± 0.5 kPa while the G’ of MSs with ICG were between 15.9 ± 0.1 kPa (1Alg2Ca1ICG) and 20.5 ± 1.3 kPa (1Alg2Ca0.5ICG). It indicated that the addition of ICG diminished the ability of MSs to store energy elastically. This might be due to the negatively charged surfaces of both ICG and alginate which reduced the cross-linking density and consequently led to lower G’ values.

Frequency sweep was performed at a shear strain of 0.04% (pre-determined in amplitude sweep) to obtain the G’ and G” of MSs at varying angular frequency. As shown in Figure 4C, both G’ and G” increased slightly with increasing angular frequency. The G’ of 1Alg2Ca0ICG was slightly higher than the MSs with ICG at angular frequency ranged from 0.1 rad/s to 30 rad/s. Generally, the addition of ICG showed decreased G’. 1Alg2Ca0.5ICG, 1Alg2Ca0.75ICG, and 1Alg2Ca1ICG that had higher ICG concentrations showed lower G’ compared with 1Alg2Ca0ICG, 1Alg2Ca0.1ICG, and 1Alg2Ca0.25ICG which had lower ICG concentrations.

However, this trend was not significantly dependent on concentration. It might be because the ICG concentration could only reach 1.00 mg/mL, which was low and did not have a strong impact on the cross-linking structure. The lost factor (tan (δ) = G”/G’) shown in Figure 4D fell approximately within the range of 0.1 to 0.4 and exhibited almost no dependence on the frequency variation throughout the entire range of the frequency sweep. Since a loss factor of 1 stands for the transition between a solid state and a liquid state [32], the results indicated that the MSs were at solid state but not pure elastic solids (tan (δ) = 0). The similarities of loss factors also indicated that the ICG concentrations did not affect the solid/liquid behavior of the MSs significantly.

3.1.4. Swelling behavior of MSs

Swelling determines the size of the targeted blood vessel that the MSs can occlude. The sizes of 1Alg2Ca0.1ICG, 1Alg2Ca0.25ICG, 1Alg2Ca0.5ICG, 1Alg2Ca0.75ICG, and 1Alg2Ca1ICG in PBS solution were measured within 48 hours to obtain their swelling behavior. It was found that MSs with different ICG concentrations had different swelling ratios. The highest and lowest changes (ratio) of mean diameter occurred on 1Alg2Ca1ICG (from 270.5 μm to 384.3 μm, a 42% increase) and 1Alg2Ca0.25ICG (from 260.6 μm to 305.0 μm, a 17% increase), respectively. The mean diameters of 1Alg2Ca0.1ICG, 1Alg2Ca0.5ICG, and 1Alg2Ca0.75ICG increased by 18%, 38%, and 32%, respectively.

It can be extrapolated that the swelling ratio had a tendency to increase when ICG concentration increases. Similar trend can be found in the results of rheological tests (Figure 4C) that G’ and G” were relatively high with low ICG concentration. Based on Flory’s [41][40] theory of rubber elasticity, G’ is proportional to the crosslinking density and inverse proportional to the swelling capacity. In this case, MSs with high concentration of ICG showed both low G’ and high swelling ratio, which can be explained well by Flory’s theory that the crosslink density was reduced due to the addition of ICG. However, given that the G’ did not change significantly and that the fluorescence of MSs with 1 mg/mL ICG had the longest detectability, 1 mg/mL was still considered as the optimal concentration of ICG for further study.

3.2. Mechanical and biocompatible properties of MSs with different calcium concentrations

3.2.1. Rheological tests on MSs with different calcium concentrations

The tunable mechanical strength of MSs ensures that they can be adaptable for different TAE needs. This can be achieved by changing the crosslinking density. The calcium chloride concentrations were tuned from 0.5 wt.% to 4 wt.% during the synthesis to obtain MSs with similar sizes but varying crosslinking densities. Rheological tests were conducted to evaluate the impact of calcium concentrations on mechanical strength of MSs. As shown in Figure 6A, all MSs exhibited linear viscoelastic performance (plateau) within the shear strain range from 0.01% to 0.1%. Therefore, 0.04% shear strain was again selected for further frequency sweep tests. The G’ of MSs at 0.04% was summarized in Figure 6B, showing 1Alg4Ca1ICG with the highest G’ (27.7 ± 0.5 kPa). The other MSs have similar G’ ranging from 15.9 ± 0.1 kPa (1Alg2Ca1ICG) to 17.2 ± 1.5 kPa (1Alg0.5Ca1ICG). This indicates that 1Alg4Ca1ICG has the highest ability to recover its shape after being loaded during the amplitude sweep test. However, the coefficients of variation (0.089 for 1Alg0.5Ca1ICG, 0.060 for 1Alg1Ca1ICG, 0.006 for 1Alg2Ca1ICG, and 0.018 for 1Alg4Ca1ICG) show that 1Alg2Ca1ICG has the most stable behavior for multiple parallel tests.

Figure 6.

Rheology tests on 1Alg0.5Ca1ICG, 1Alg1Ca1ICG, 1Alg2Ca1ICG, and 1Alg4Ca1ICG. A) Representative amplitude sweep; B) Summary of storage modulus; C) Representative frequency sweep tests; and D) Representative loss factor of MSs obtained from frequency sweep tests. Statistical significance was determined using one-way ANOVA with multiple comparison. ns = not significant; **** p ≤ 0.0001. Data are presented as mean ± SD (n = 3 for B).

The G’ and G” obtained from frequency sweeps (Figure 6C) demonstrated the overall stability of all MSs over the tested frequency range. The loss factors (Figure 6D) did not have significant change with varying angular frequency. All the loss factors were less than 1 and higher than 0.1, indicating solid-like behavior.

3.2.2. AFM

At the nanoscale, an MS may exhibit unique local characteristics that differ from its bulk properties. Investigating the mechanical changes resulting from the deformation of individual MS when subjected to external forces becomes imperative. Therefore, AFM as a powerful tool for local property measurement was applied to find the E of MSs by conducting nanoindentation tests. AFM loading (approach) and unloading (withdraw) force-indentation curves were obtained for 1Alg1Ca1ICG, 1Alg2Ca1ICG, and 1Alg4Ca1ICG to investigate the local modulus change with respect to the change of calcium chloride concentration.

For most MSs, the loading and unloading curves were not identical. It was because the loading curves were affected by both plastic deformation and elastic deformation, while the unloading curves were only affected by elastic deformation [34]. Hence, the unloading curve was chosen to be fitted (Figure 7A) to calculate E. Figure 7B and 7C showed the E values of 25 points within a 10 μm * 10 μm region of a single 1Alg1Ca1ICG MS. The E values ranged from 62 kPa to 180 kPa with 40% of which fell in the range between 80 kPa and 100kPa. The fluctuation of Es might be caused by the variation of the local surfaces where calcium ions from the MSs and phosphate ions from the PBS solution reacted and the crosslinking density was slightly different.

Figure 7.

Young’s moduli of 1Alg1Ca1ICG, 1Alg2Ca1ICG, and 1Alg4Ca1ICG obtained from AFM tests. A) Representative fitting for force-indentation curve from one of 1Alg1Ca1ICG; B) Representative modulus distribution map for one of 1Alg1Ca1ICG; C) Histogram of E distribution for one MS from B; and D) Summary of E of MSs for each concentration of calcium chloride. Data are presented as median with 1.5 interquartile range (IQR) as error bar. Statistical significance was determined using one-way ANOVA with multiple comparison. ns = not significant. (n = 30 for D)

Figure 7D summarized the E values of 3 types of MSs synthesized with increasing CaCl₂ concentration of 1%, 2% and 4%. The E of 1Alg1Ca1ICG, 1Alg2Ca1ICG, and 1Alg4Ca1ICG were 42.51 ± 34.93 kPa, 54.13 ± 18.97 kPa, and 41.47 ± 23.72 kPa, respectively. These values were close to some of the clinically used MSs (48.0 ± 10.5 kPa for DC Bead +Iri [35] and 39.6 ± 5.05 kPa for EmboSphere [22]). However, in most of these studies they performed compression tests on a monolayer of the MS instead of testing the local E on a single MS. It is worth noting that the E of human blood vessels [36] at the early stage of radius expansion is around 20 kPa to 50 kPa, which indicates good flexibility and recoverability. These values are close to the E values of the MSs obtained in this research. The matching E of blood vessels and MSs ensures good embolization without destruction of either blood vessels or MSs. Statistical analysis showed no significant difference between the E of each type of MSs. This was partially caused by the wide distribution of E of the MSs. However, the E values of 1Alg1Ca1ICG and 1Alg4Ca1ICG had relatively higher variation and lower average value. For 1Alg1Ca1ICG, low calcium ions might lead to lower cross-linking density or incomplete cross-linking process. The swelling of 1Alg1Ca1ICG might further decrease the stability. For 1Alg4Ca1ICG, the high concentration of calcium ions might generate a quickly crosslinked shell whereas the sodium alginate inside the MSs was not completely reacted. Meanwhile, the reaction of phosphate ions in PBS with the high concentration of calcium ions might change the surface property and cause instability of the MSs. Theoretically, appropriate concentration calcium ions (as in 1Alg2Ca1ICG) can form a uniformly crosslinked core which exhibit relatively stable behavior in PBS. Therefore, 2% was considered the optimal concentration of calcium chloride used to make MSs based on AFM studies.

3.2.3. Hemolysis and Thrombogenicity

Study of blood compatibility and thrombogenicity is essential when evaluating new embolic materials due to their direct interaction with blood [37]. Hemolysis, the process of red blood cell destruction, can cause hemolytic anemia, which leads to serious problems for the human body [38]. The hemolysis test is regarded as a straightforward and dependable method for assessing hemocompatibility, with a hemolysis rate below 5% being deemed acceptable [28]. It can be observed in Figure 8A that the hemolysis rate of 1Alg2Ca0ICG was close to 0. The hemolysis rates of 1Alg1Ca1ICG (0.92% ± 0.1%), 1Alg2Ca1ICG (1.47% ± 0.07%), and 1Alg4Ca1ICG (1.67% ± 0.03%) were all lower than 5%, indicating good hemocompatibility [28].

Figure 8.

Hemolysis and thrombogenicity tests on 1Alg2Ca0ICG, 1Alg1Ca1ICG, 1Alg2Ca1ICG, and 1Alg4Ca1ICG. A) Hemolysis rate shows good hemocompatibility of MSs; B) Images of coagulated blood in thrombogenicity test in which activated porcine blood was in contact with MSs. Liquid residual was removed at 2, 4, 5.5, 6.5, and 7.5 minutes. Clinically used coils and porcine blood alone were used as controls for comparison. Data are presented as mean ± SD (n = 4 for A)

Thrombogenicity is the tendency of a material to generate blood clotting and/or thrombus [39]. The formation of blood clot can aid in minimizing the risk of inadequate volumetric packing and organization of a thrombus during TAE, thereby lowering the possibility of recanalization [40]. The effect of 1Alg2Ca0ICG, 1Alg1Ca1ICG, 1Alg2Ca1ICG, and 1Alg4Ca1ICG on the coagulation of porcine blood was compared to clinically used coils (Figure 8B). At 2 minutes, 1Alg2Ca0ICG showed stronger thrombogenicity than 1Alg2Ca1ICG, indicating the addition of ICG might have a negative effect on blood clotting. The thrombogenicity of 1Alg4Ca1ICG was the highest among all the three types of MSs with ICG, meaning that the increasing calcium concentration (or crosslinking density) might enhance the thrombogenicity. All the embolic agents exhibited capability of thrombogenicity, while the negative control (plate alone) showed almost no blood clot. At 4 minutes, the coagulation of blood mixed with MSs were able to take over 90% of the wells area, which was significantly higher than that with coils (50%), and obviously much higher than the negative control. After 5 minutes, all the materials including the negative control showed 100% blood clotting in the wells. This result suggested that MSs have good enhancement on thrombogenicity compared with the clinically used embolic agent.

4. Conclusion

Ca-alginate MSs with varying calcium chloride and ICG concentrations were synthesized. Microscopic images showed high uniformity of morphology and size distribution. Fluorescence test indicated that higher ICG concentrations increased fluorescence intensity and exhibited longer retention time. Results of rheological and swelling experiments demonstrated that higher ICG concentrations and lower calcium chloride concentrations both led to lower crosslinking strength and yield lower G’. The local young’s modulus was not significantly affected by calcium chloride concentrations based on the results of AFM test. Blood tests showed good hemocompatibility and enhancement of blood clotting. Based on the stable behavior shown in both rheological test and AFM tests, our study suggested that the optimal concentrations of ICG and calcium chloride for embolic calcium alginate MSs were 1 mg/mL and 2 wt.%, respectively.

Figure 5.

Swelling behavior of 1Alg2Ca0.1ICG, 1Alg2Ca0.25ICG, 1Alg2Ca0.5ICG, 1Alg2Ca0.75ICG, and 1Alg2Ca1ICG in 1X PBS. A) Mean diameters of MSs at 0, 24, 48 hours; B), C), and D) Representative microscopic images of 1Alg2Ca1ICG in PBS at 0, 24, 48 hours respectively. Data are presented as mean ± SD (42 ≤ n ≤ 178 for A).

Highlights:

A series of calcium-alginate microspheres were developed as embolic agents.

Indocyanine green was incorporated into the microspheres for imageability.

Tunable mechanical properties were obtained with varying Ca²⁺ concentrations.

Optimized microspheres showed good imageability, stability, and blood compatibility.

Abbreviations:

TAE

Transcatheter arterial embolization

AVM

Arteriovenous malformations

MS

Microsphere

NIR

Near-infrared

ICG

Indocyanine green

AFM

Atomic force microscopy

PBS

Phosphate buffered saline

FI

Fluorescence intensity

LAOS

Large amplitude oscillatory shear

FS

Frequency sweep

E

Young’s modulus

G’

Storage modulus

G”

Loss modulus