Professional Documents
Culture Documents
Weng2013 PDF
Weng2013 PDF
Weng2013 PDF
PII: S1742-7061(13)00304-8
DOI: http://dx.doi.org/10.1016/j.actbio.2013.06.020
Reference: ACTBIO 2775
Please cite this article as: Weng, L., Rostambeigi, N., Zantek, N.D., Rostamzadeh, P., Bravo, M., Carey, J., Golzarian,
J., An in situ forming biodegradable hydrogel-based embolic agent for interventional therapies, Acta
Biomaterialia (2013), doi: http://dx.doi.org/10.1016/j.actbio.2013.06.020
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
An in situ forming biodegradable hydrogel-based embolic agent for
interventional therapies
*
Corresponding author. Address: Department of Radiology, University of Minnesota. Mayo B228, 420 Delaware Street SE,
Minneapolis, MN 55455. Phone: 612-625-5147. Fax: 612-626-5580. Email: jafar@umn.edu.
1
An in situ forming biodegradable hydrogel-based embolic agent for
interventional therapies
ABSTRACT
We present herein characteristics of an in-situ forming hydrogel prepared from carboxymethyl chitosan
and oxidized carboxymethyl cellulose for interventional therapies. The gelation, owing to the
formation of Schiff base, occurred both with and without the presence of a radiographic contrast. The
hydrogel exhibited a highly porous internal structure (pore diameter: 17 ± 4 µm), non-cytotoxicity to
human umbilical vein endothelial cells, hemocompatibility with human blood, and degradability in
lysozyme solutions. Drug release from hydrogel loaded with a sclerosant, tetracycline, was measured
at pH 7.4, 6 and 2 at 37 oC, and showed that tetracycline was more stable at acidic conditions with a
lower release rate observed at pH 6. Moreover, an anticancer drug, doxorubicin, was loaded into the
hydrogel and a cumulative release of 30% was observed over 78 h in PBS at 37 oC. Injection of the
hydrogel precursor through a 5-F catheter into a fusiform aneurysmal model was feasible, leading to
complete filling of the aneurysmal sac, which was visualized under fluoroscopy. Level of occlusion
was compared between hydrogel precursors (1.8% and 2.1%) and standard microspheres (100-300 µm)
in a rabbit renal model. Embolization with hydrogel precursors was performed without clogging and
the hydrogel achieved effective occlusion in more distal arteries than calibrated microspheres. In
conclusion, this hydrogel possesses promising characteristics potentially beneficial for a wide range of
2
1. Introduction
Interventional vascular therapies offer minimally invasive methods to treat patients, allowing for
faster healing processes compared with open surgeries. Currently, the treatment of endoleaks after
rapidly gaining interest due to their safety and efficacy [1-3]. Several embolic materials are now on
the market, such as coils, liquid embolics, and particles. Coils have been extensively used to treat
aneurysms, but only 30% of the sac can be occupied by the coils and a significant portion of the sac
will be occluded by unorganized thrombi [4, 5]. Hence, recurrence of endoleak has been observed
frequently. In contrast, liquid embolic agents such as cyanoacrylates and Onyx® (ev3, Irvine, CA) are
shown to achieve more complete filling [6, 7]; however, both types of these liquid agents have proven
to be problematic with regard to the migration of embolic agents out of the aneurysmal sac and
cytotoxicity due to the presence of organic solvents [5, 8]. Moreover, these materials are not resorbable
or loadable. Another shortcoming of currently available embolic agents lies in the flexibility in usage
for various clinical implications. For example, coils cannot be used to treat hypervascular tumors while
particles cannot be used to treat large aneurysms. Thus, it is very appealing to explore a new multiple-
potent embolic agent formulated with biocompatible components that can fill the aneurysm sac,
In situ forming hydrogels are a class of hydrogel materials that solidify from liquid forms in the
application sites. Among all the materials used to fabricate in situ forming hydrogels, polysaccharides
are of great interest due to their biocompatibility and biodegradability compared to synthetic materials
[9]. Strategies used to achieve in situ forming hydrogels generally involve physical crosslinking or
chemical crosslinking. Physical crosslinking usually takes a longer time with a weaker gel formed,
formation [11]. In contrast, chemical crosslinking may have a faster gelation with a higher gel
3
mechanical strength where the gelation is driven by Schiff base reactions [12, 13], Michael addition
In situ forming hydrogel can be potentially administered straightforward through a catheter prior to
gelation. Different in-situ-forming hydrogels have been investigated for the treatment of cerebral
aneurysms [18], cerebral arteriovenous malformations [15], hemorrhage [19], endoleaks after
endovascular repair [20], and as tissue sealants [14]. However, none of them has been well adapted as a
multi-potent embolic agent that has the potential to be used in a variety of conditions. The constellation
features of in situ forming hydrogels with the potential to deliver drugs, exert embolic properties, and
eventually resorb in tissue milieu without any toxicity or hemostasis alteration may enhance efficacy,
and reduce complications or reintervention rates reported with the use of previous agents [2, 21].
The purpose of this study was to comprehensively evaluate the physicochemical properties of a new
in situ forming embolizing hydrogel prepared from water-soluble natural polysaccharides, namely
carboxymethyl chitosan (CCN) and carboxymethyl cellulose (CMC). Although the safety of using
these two polymers for biomaterials has been established individually, the combination of two low-cost
polymers for fabricating an in-situ forming hydrogel has not been reported. We aimed to demonstrate
the non-cytotoxicity, degradability, impact on hemostasis, and drug release characteristics of this new
material and to demonstrate its embolization and immediate occlusive effects in a rabbit renal model.
Sodium carboxymethyl cellulose (CMC) (Mw = 700,000 Da), and chitosan (Mw = 190,000~310,000
Da) were purchased from Sigma-Aldrich Corp. (St Louis, MO). All the other chemicals and solvents
were of highest purity commercially available. Oxidized CMC was prepared by oxidizing CMC with
sodium periodate. Briefly, in a 250 mL flask, 1 g of sodium CMC and 80 mL distilled water were
added. After CMC dissolved completely, 25% molar equivalent of sodium periodate in 20 mL distilled
4
water was added into the flask. The reaction was allowed to proceed for 24 h at 25 oC. Then 0.21 mL
ethylene glycol was added into the flask to stop the reaction. After 30 min the mixture was poured into
a dialysis tube (MWCO 3500, Fisher Scientific, Pittsburgh, PA) to dialyze thoroughly against distilled
water. Dry product was then obtained by lyophilizing the solution. The oxidation degree of OCMC was
defined as the percentage of CMC structural units that were oxidized and it was determined by
measuring the aldehyde content in OCMC by iodometry [22], and the molecular weight was measured
with GPC (Waters, Milford, MA). CCN was synthesized with Chen’s method [23]. The substitution
described [24].
The hydrogel was prepared by blending OCMC and CCN aqueous solutions (1.5%, 1.8%, 2%, or
2.1%, w/v) in 1:1 ratio. Then the well mixed precursors were added into the 24-well non-tissue culture
The gelation time of the CCN/OCMC precursor in different solvents (water, 0.9% saline, 0.01M
PBS, and citrated human plasma) with or without radiographic contrast (Optiray 320, Mallinckrodt
Inc., Hazelwood, MO) was determined by the following protocol [25]. Human plasma was collected by
centrifuging the citrated whole blood donated by one study investigator who was known to have
normal results for the assays tested. In a Petri dish (BD, Franklin Lakes, NJ), 100 µL of OCMC
solution and 100 µL CCN solution were mixed with a magnetic stir bar at 155 rpm using a
hotplate/stirrer (Isotemp 11-100, Fisher Scientific, Pittsburgh, PA). The gelation time was determined
when the mixture formed a globule. The experiments were repeated four times per sample.
The morphology of lyophilized swollen hydrogel was evaluated by scanning electron microscopy
(SEM) (JEOL 6700, Japan) after spray-coated with gold. Swollen gel pieces were snap-frozen in liquid
5
nitrogen and then lyophilized. A section of the dried gel was mounted on a metal stub coating a layer of
conductive adhesive. SEM images were obtained at a 2.0 KV acceleration voltage in a deceleration
2.4. Cytotoxicity
Primary human umbilical vein endothelial cells and medium (MCDB–131 Complete medium) were
purchased from VEC Technologies, Inc. (Resselaer, NY). The endothelial cells were seeded in 35 mm
plastic dishes (an area of 8.55 cm2, BD, NJ) at a density of 200,000 cells per well. In order to avoid the
cell damage merely caused by weight of the hydrogel on top of the cells, one circular plastic was put at
the center of each dish before seeding. Hydrogels were prepared with 2% OCMC and 2% CCN
combination. The resulting hydrogels were further conditioned with PBS solution (10×, pH 7.4, Life
Technologies, NY) while shaking at 50 stroke/min for 30 min. Thereafter they were sterilized with
70% ethanol followed by thorough PBS rinsing. Then they were conditioned with standard endothelial
cell medium to remove PBS trapped inside the hydrogel. The circular plastic in the study group was
removed after 24 h and the area was replaced with a piece of hydrogel. After 24 h of culture of cells
with and without hydrogel, cells were rinsed twice with HBSS (Hank's Balanced Salt Solution, Life
Technologies Corp., NY). Then cells were incubated for 30 min at 37 °C in 1 mL 1 mg/mL MTT
reagent (Sigma-Aldrich Co., St. Louis, MO) dissolved in phenol free 20% Roswell Park Memorial
Institute medium (RPMI). After that, the MTT solution was removed from the cell culture dish, and the
dish was air dried. Subsequently, 1 mL of DMSO was added into the dish to dissolve all the MTT. The
absorbance of the MTT/DMSO solution was measured via a microplate reader (Gen5, BioTek®
Eighteen pieces of hydrogel were prepared from 0.5 mL 2 w/v% CMC solution and 0.5 mL 2 w/v%
CCN solution in a 24-well plate for a 21-day period. The hydrogels were incubated with 1 mL of
6
lysozyme PBS solutions (10 µg/mL and 4 mg/mL) at 37 oC and the medium was changed every other
day. At predetermined intervals, three samples were retrieved and rinsed thoroughly with distilled
water. The resulting hydrogel residues were then freeze-dried. The percentage of weight loss was
calculated based on the following equation: Percent weight loss = [(Wt – W0)/ W0]×100%, where Wt
and W0 were the dry weight of the hydrogel at day t and day 0, respectively.
Human whole blood was collected from one study investigator who was known to have normal
results for the assays tested. Blood was collected into a 3.2% sodium citrate vacutainer (Becton
The blood was kept at room temperature prior to use and testing was initiated within 30 minutes to 2
hours after collection. Testing was performed on a TEG® Hemostasis System 5000 (Haemoscope
Corporation, Niles, IL) following a reported protocol [19] with slight modifications. All reagents used
were from Haemoscope Corporation (Stoughton, MA) unless noted. Clear disposable TEG cups and
pins were pre-warmed to 37 °C. One mL of blood was added to a vial of kaolin and mixed gently by
inverting the vial. The reagents for the reaction were placed into the cup in the following order: 320
µL blood from the kaolin vial, 20 µL saline (0.9% sodium chloride injection USP, Baxter Healthcare
Corp., Deerfield, IL) or hydrogel precursor, and 20 µL 0.2 M calcium chloride. The hydrogel precursor
was added using a positive-displacement pipette (Microman®, Gilson, WI). Complete gelation of the
hydrogel precursor did not occur prior to its addition to the reaction. The reaction was allowed to run
For tetracycline loading and release study, 3 mL of 2% w/v CCN aqueous solution, 1 mL 1.2% (w/v)
OCMC aqueous solution, were added into a vial to give a drug payload of 10%. After the solutions
7
were mixed well, 2 mL aliquots were removed from the vial and introduced into 20 mL glass vials in
triplicates. Then the vials were kept at 37 oC for 2 h to ensure complete gelation. The release studies
were performed at three different pH conditions: 2, 6, and 7.4. Fifteen milliliter of the solution at
desired pH was added and then the vials were incubated at 37 ± 0.5 oC with mild shaking (100
strokes/min). At predetermined time-points, aliquots of 200 μL were withdrawn from each vial and
replaced with an equal volume of fresh medium of similar pH. Absorbance of the released tetracycline
was read at a predetermined λmax nm with UV-Vis spectrophotometer (DU650, Beckman Coulter Inc.,
Brea, CA) for each pH condition. The calibration curves were prepared for a series of tetracycline
solutions with known concentrations for each pH prior to the conduct of each release study. For
doxorubicin loading and release study, 1 mL 2 % CCN aqueous solution, 0.5 mL 2 mg/mL doxorubicin,
and 1 mL 2% OCMC aqueous solution were mixed in a 20 mL glass vial. The vial was then placed in
the refrigerator (5 oC) for 20 h for full gelation of the mixture. Into each vial containing the
doxorubicin loaded hydrogel, 15 mL PBS (0.01M) was added. The vial was then placed on a shaker
(50 stroke/min) in the incubator (37 oC). At predetermined time points, 0.5 mL aliquot was taken out
from the vial and replaced with 0.5 mL fresh PBS. The doxorubicin concentration was measured with
the UV-Vis spectrophotometer at a wavelength of 482 nm. All the releasing studies were done in
triplicate.
The injectability of the hydrogel precursor (2%) was tested with a 5-F catheter (Terumo, Glidecath,
ID=0.038") after mixing 5 times through a 3-way stopcock. The hydrogel solution containing 20% (v/v)
radiographic contrast was injected into an aneurysmal sac of a fusiform aneurysmal model mimicking
the endovascular treatment. Fluoroscopy was performed during the procedure depicting the filling of
8
The study protocol was approved by the Institutional Animal Care and Use committee. Four adult
rabbits (New Zealand, white) that weighed 3-4 kg were used. Anesthesia was induced by administering
ketamine hydrochloride (35 mg/kg), with xylazine (5 mg/kg) and acepromazine (0.75 mg/kg)
intramuscularly. Rabbits were intubated and general anesthesia was maintained with 1.5%-2%
isoflurane and 1 L/min oxygen, with approximately 100-110 mL tidal volume and 10-15 breaths/min.
Each animal was continually monitored with an electrocardiogram, oxygen saturation levels, body
temperature and assessment of depth of anesthesia during the procedure. The right common femoral
artery was surgically exposed and a 4-F sheath was placed. Using fluoroscopy, the renal artery was
catheterized with a 4-F Cobra catheter (ID = 0.038", Length: 65 cm, GlidecathTM, Terumo Medical
Corp., Somerset, NJ). A 2.8-F microcatheter (ID=0.027", ProgreatTM, Terumo Medical Corp., Somerset,
NJ) was placed in the renal artery and the 4-F catheter was pulled back into the aorta for free flow
delivery of microspheres at a constant flow rate of 2 mL/min. For each kidney, embolization was
performed until the total occlusion was achieved. The right kidneys were embolized with the hydrogels
while the left kidneys received calibrated bioresorbable microspheres (100-300 µm, 10 mg/mL in
normal saline) for comparison. The calibrated bioreosorbable microspheres were prepared according to
an inverse emulsion method [26]. Two of the four rabbits received the 1.8 w/v% hydrogel precursor
while the other two were embolized with the 2.1 w/v% hydrogel precursor. For each kidney,
angiography was performed before, immediately after, and 10 min after embolic procedure. The
volumes of the hydrogel precursor and microsphere suspension required were recorded. The ease of
injection was subjectively evaluated by an interventional radiologist (J. G.) with greater than 20 years
of experience.
Immediately after the final angiogram, these rabbits were given heparin at 150 units/kg and 5 min
later euthanized with an overdose of Euthasol®. The kidneys were surgically removed and fixed in
10% buffered formalin. 4-µm coronal sections were obtained from each kidney and stained with H&E.
9
The size and location of all embolized vessels (i.e., vessels containing embolic agent) were determined
as previously shown [27]. For the distribution of hydrogel/microspheres in kidneys, the following
parameters were assessed: 1) The arterial location of the hydrogels/microspheres. 2) The numbers of
glomeruli were counted for each kidney to obtain the percentage of glomeruli containing gels from 6
representative slides. 3) The vessel diameter occluded by microspheres. To measure the vessel sizes,
the histological slides were scanned with a whole slide scanner (Aperio ScanScope XT) using a
tracking feature. Then the vessel size was measured with a calibrated ruler in Aperio ImageScope
Data were summarized using mean and standard deviation for numeric variables, and counts and
proportions for categorical variables. Student’s t test and ANOVA were used for the comparison of
Bioresorbable materials are currently of increasing interest in interventional radiology with respect
to their temporary occlusion characteristics and biodegradable nature that overcomes the persistency in
the body of the permanent materials [26]. The first report on the use of bio-resorbable materials
appeared in 1980s [28] with subsequent reports on the use of various resorbable materials such as
living tissue. Injectable in situ forming bioresorbable hydrogels have received extra attention because
they can be used as an embolic agent through a coaxial system. Nonetheless the loco-regional impact
in the homeostasis of the organs, drug delivery and embolization features are largely unknown. We
presented a new in situ forming bioresorbable hydrogel composed of CCN and OCMC. Different
characteristics of this material indicated that its application in a variety of clinical conditions is feasible.
The indications such as the treatment of tumors, arteriovenous malformations, and arterial
hemorrhages as well as adjunctive use with stent graft for endovascular treatment of aortic aneurysm
10
are very likely applications that this hydrogel can be successfully used for.
CMC was partially oxidized with sodium periodate. Periodate turned the vicinal hydroxyl groups
into two aldehyde groups, and a number of the sugar rings on CMC were thus open, leading to a more
flexible polymer backbone [29] with functional groups. The oxidation degrees of OCMCs were 17.01
± 0.78% as determined by the iodometry titration method. The corresponding weight average
molecular weights were 109370 Da, as measured by GPC. Although chitosan is a biocompatible and
biodegradable polymer, it is not soluble at physiological pH that is necessary for materials designated
for biological applications. Carboxymethyl groups were thus introduced into the structure of chitosan.
The substitution degree of chitosan (carboxymethylation) was 1.03 ± 0.03 as verified by the back-
titration method. The modified chitosan was readily soluble at physiological pH.
The mixture of CN and OCMC solutions underwent a sol-gel transition and gelled into a clear
hydrogel at ambient temperature (23 oC) or 37 oC. The gelation is due to a self-crosslinking process
based on Schiff base formation [26]. The gelation time was adjustable via both the concentration of the
individual polymers and composition of the precursors as shown in Fig. 1. Incorporation of iodine
contrast delayed the gelation process, but less effect was observed at higher polymer concentrations.
The presence of 20% (v/v) contrast in a 2% hydrogel precursor produced a radiopaque gel within 2.55
± 0.05 min (mean ± SD). Moreover, the gelation time was both temperature and solvent dependent.
Elevation in temperature from 23 oC to 37 oC accelerated the gelation process, where at 23 oC, the
gelation time decreased from 4.81 ± 0.35 to 1.92 ± 0.22 min (mean ± SD) with an increase in the
polymer concentration from 1.5% to 2% (Fig. 1a). The gelation can occur in different media, such as
distilled water (DI), 0.01M PBS, 0.9% normal saline and human plasma (Fig.1b). At 37 oC, gelation in
PBS (1.30 ± 0.06 min) occurred slightly earlier than those in DI (1.53 ± 0.02 min) or saline (1.35 ±
11
0.09 min). In comparison, the gelation was slower in citrated human plasma (4.28 ± 0.36 min). The
slower gelation process is likely linked to the interactions between the proteins in the plasma and
individual polymers in the hydrogel precursor. Meanwhile, it is worthwhile mentioning that gelation in
this situation occurred with the plasma as the solvent for both polymers, whereas in vivo studies were
performed with hydrogel polymers dissolved in saline and then the precursor was introduced to blood
flow at approximately 1 min after mixing. In the latter situation gels formed inside the arteries within 1
min, which was evidenced by the rapid disappearance of the distal arteries during fluoroscopy in
SEM showed interconnecting pores in the hydrogel prepared from a 2% OCMC/CCN precursor (Fig.
2) and pore size analysis gave an average pore diameter of 17 ± 4 µm. In addition, it was also observed
that the pore size distribution depends on the swollen state of the hydrogel and polymer concentration
(data not shown). The porous structure of the hydrogel may facilitate the drug release, which has been
3.4. Cytotoxicity
This aqueous in situ gel-forming matrix is completely devoid of residual toxic solvents as compared
with many preformed bulk gels. Both CCN and CMC are biocompatible polymers [11, 26] and
cytotoxicity is not expected from the combination of these two polymers. Endothelial cell line was
selected for co-culture since they bear similarity to the endothelial layering of the arterial surface. Up
to 24 h, there was no significant difference in the cell density, morphology and distribution between the
cells co-cultured with hydrogels (Fig. 3a) and control (Fig. 3b). In addition, the standard MTT assay
based on the number of viable cells in the culture flask showed no significant difference between
endothelial cells cultured with and without hydrogels (optical density 1.17 vs. 1.19, p = 0.8, n = 4) (Fig.
3c), indicating the hydrogels were nontoxic to endothelial cells. In contrast to the conventional liquid
12
embolics such as Onyx that have angiotoxicity [20], this hydrogel therefore appears to be more
This hydrogel underwent degradation under physiological pH in the presence of lysozyme. The dry
weight of the hydrogel showed a consistent decrease over the degradation period of incubation with
lysozyme (Fig. 4). In the 4 mg/mL lysozyme solution, there was a significant drop in the hydrogel
weight between day 0 and day 10 (p = 0.01). The starting dry weight of the hydrogel was 20 mg on day
0, and decreased to 15.2 mg (day 3), 7.4 mg (day 5), 5.1 mg (day 7), and 1.2 mg (day 12)
demonstrating a consistent decrease in the weight of hydrogel with greater than 20% decline every two
days. The change after the tenth day was less remarkable although remained significant (p = 0.01). In
comparison, the degradation in the 10 µg/mL lysozyme solution occurred at a slower pace. On day 7,
the weight percentage of the hydrogel remained in this solution was 57.8%, which was over 2 times
higher than that (25.5%) in the 4 mg/mL lysozyme solution (p = 0.004). Moreover, there was still
22.1% of the hydrogel left on day 21, indicating that decreasing the concentration of the lysozyme
solution delayed the degradation process. The degradation can be attributed to two mechanisms: 1) The
main crosslinks in the hydrogel, Schiff bases, are well known to be subject to hydrolysis; 2) Chitosan,
part of the hydrogel, can be degraded by lysozyme, an enzyme that is ubiquitously found in the body
[30].
Hypothetically, the aneurysmal sac or an arteriovenous malformation vascular bed will be ideally
filled over time with fibrotic tissue while hydrogel is being resorbed and delivery of sclerosing agent
(i.e. tetracycline) happens. Therefore, by having a biodegradable embolic material there will be no
persisting foreign body in the vessel. There will be, instead, an organized solidified tissue structure as
may potentially reduce hypoxia-induced angiogenesis in the malignant tumors [31]. Moreover, the
13
vessels would be more accessible for potentially later interventions if the embolic agent is resorbed.
Also there is a chance that the resulting small pieces produced during the degradation of the hydrogel
might enter the main stream circulation. Embolization in the case of transcatheter arterial
pieces to more distal arterioles would not be troublesome. Indeed, resorbable materials (e.g. Gelfoam)
have been used in many embolization cases and distal migration has not been an issue. However, the
potential untoward effect of this biodegradable material could be that the detached hydrogel materials
during degradation embolize into undesired areas. Thrombus formation would help to prevent the
migration of the degraded material. In vivo studies are needed to test these potential effects.
3.6. Thromboelastography
For biomaterials used for vascular applications, their interactions with blood determined their
intended functions in vivo. A number of tests have been reported for assessing the hemocompatibility
of biomaterials, such as partial thromboplastin time (PTT), prothrombin time (PT), thrombin clotting
time (TT) [32, 33]. However, thromboelastography (TEG) is a global assay of blood coagulation based
on quantitative measurement of the elasticity of whole blood, from the beginning of coagulation
through clot formation towards the final fibrinolysis [34]. It provides additional information compared
to traditional lab tests (PT or PTT) which focus only on the initiation of the coagulation cascade. Fig. 5
presents representative thromboelastographs for citrated whole human blood with normal saline (black)
or hydrogel precursor (green) to elucidate their gelation effects on blood coagulation, and the data was
summarized in Table 1. The presence of hydrogel did not inhibit or enhance the onset of clot formation
or coagulation (Fig. 5 and Table 1) (i.e., R time, 7.1 min vs. 6.7 min, p = 0.34; K, 2.4 min vs 2.3 min, p
= 0.8; α angle, 59.1 degree vs 59.4 degree, p = 0.9). In addition, there was no significant difference in
the clot strength (as measured by the maximum amplitude, MA, 70.2 mm vs 68.5 mm, p = 0.1), and
fibrinolysis at 30 min (both were 0%) and 60 min (0.1% vs 0.2%, p = 0.2). Peng et al [21] investigated
14
the blood compatibility of a hydrogel made of polyethylene glycol and polyallamine hydrochloride
designated for hemorrhage control and found that the presence of their hydrogel enhanced the overall
procoagulant effects. In our study, the presence of the OCMC/CCN hydrogel did not show significant
impact on the blood clotting time, clot strength, or fibrinolysis as compared to normal saline control.
The difference in observations between our study and Peng’s [21] may be due to the addition of kaolin
in the reaction or the components of the hydrogel. Despite the addition of kaolin, a marked inhibitory
or prothrombotic effect by the hydrogel would be expected to be detected by the method used here, as
kaolin is routinely used in the test performed in clinical laboratories where it is specifically used to
detect abnormalities in hemostasis. However, kaolin is a strong activator of coagulation and a weak
inhibitory or prothrombotic effect of the hydrogel may not be detected with significant power due to
the short reaction time of the assay. A material for embolization that does not alter hemostasis may be
ideal for embolization purposes. A product that had an anticoagulant effect, i.e. inhibits thrombus
formation, would have the potential to worsen bleeding. While a product that has a prothrombotic
effect may locally promote thrombus formation and decrease local bleeding, if any of the product
enters into the more general circulation may set up an undesirable prothrombotic state elsewhere
perhaps leading to a more systemic effect such as disseminated intravascular coagulation. This
OCMC/CCN hydrogel does not appear to inhibit or enhance coagulation which allows its general use
Tetracycline is an antibiotic that is used as a sclerosing agent and is reported to provoke sclerosis and
tissue fibrosis, which will in turn lead to shrinkage of cavities such as lymphoceles or pleural space
[35]. This effect can also be utilized in the aortic sac via injection of tetracycline loaded hydrogels after
stent graft implantation to treat abdominal aortic aneurysm. Loading tetracycline hydrochloride into
the OCMC/CCN hydrogel was performed as the hydrogel matrix was negatively charged (-COO-)
15
while tetracycline was positively charged (-+NH(CH3)2). In addition to neutral pH, we tested
tetracycline release in two other conditions: PH = 6 was chosen to replicate more acidic conditions in
clinical situations such as infectious milieu; PH = 2 was chosen merely for experimental purposes to
show the trend of release and stability of tetracycline in extreme conditions. The tetracycline release
followed a consistent increasing pattern over the first 24 h, where the initial drug release speed at pH =
6 was lower than those at both pH = 7.4 and pH = 2 (Fig. 6a). Within the first hour, the cumulative
tetracycline released was 29.2%, 8.6%, and 23.4% in the medium of pH = 2, pH = 6 and pH = 7.4,
respectively. In the medium of pH = 7.4, the highest cumulative tetracycline release was 65.8% at 24 h.
Subsequently, a decline in the tetracycline concentration was observed after 24 h, which was attributed
to the instability of tetracycline in the medium of pH = 7.4 [37]. On the contrary, in the pH 6 solution
and more so with the pH 2 solution, the amounts of drug release leveled off after 24 h without a decline
until the end of the study period at 72 h since tetracycline was more stable in acidic conditions [36].
The final cumulative tetracycline release was 77.0% and 58.6% in the medium of pH = 2 and pH = 6,
pH = 7.4 and pH = 2. This suggested that this hydrogel would have some unique in vivo releasing
profiles because the extracellular pH is slightly acidic (about 6 ~ 7) under ischemic conditions [37]
To further explore the possibility of using this hydrogel as a drug eluting vehicle for TACE,
doxorubicin, was incorporated into the hydrogel before the gelation occurred. Fig. 6b shows the 78 h
release profiles of doxorubicin from the hydrogels in 0.1 M PBS at 37 oC. The hydrogel released
approximately 6% of its drug content within 1 h. The release leveled off around 50 h with a final
cumulative percentage of approximately 31.3%. Compared to the tetracycline release, the cumulative
release of doxorubicin was lower at pH = 7.4. This difference can be possibly explained as follows:
First, the initial payload of tetracycline was 10% while that of doxorubicin was 2.5%, and thus there
16
was likely a portion of tetracycline that remained in the hydrogel by simply physically trapping inside
the hydrogel network instead of ionic binding, and during releasing, this trapped portion of tetracycline
was released first. Secondly, the ionic interaction between doxorubicin and hydrogel (-+NH3 and -
COO-) [38] was stronger than that between tetracycline and hydrogel (-+NH(CH3)2 and -COO-). Thus,
relatively higher percentage of doxorubicin was retained inside the hydrogel through ionic binding.
TACE using drug-loaded microspheres has shown less systemic toxicity and drug-related side-
effects compared to the conventional TACE. However, the only commercially available drug-eluting
bead (DC bead, Biocompatibles, UK) in USA is not biodegradable, and it is reported that long-term
presence of these microspheres containing potentially harmful drug in the body provoked chronic
inflammation and thus caused more tissue damage [39]. In contrast, the biodegradable hydrogel may
be resorbed and the loaded drug completely released, reducing the chances of protracted tissue damage
Endovascular repair of aortic aneurysms (EVAR) may fail if it leads to the development of
endoleaks [2, 40]. It is reported that aneurysmal sac expansion and endoleak formation may lead to
reintervention in up to 27% of procedures [21]. Thus the use of preventive measures to reduce
endoleaks appears to be critical in the treatment of these patients. In this study, the injection of
hydrogel precursors was attempted before the sol transformation into gel in a static model without
considering the blood flow. However, in a real clinical situation, a stent graft will be deployed before
filling the sac, and the rapid pulsatile flow in the aorta will have less effect on the separated sac area as
compared to an open system without the stent graft. Thus, the gelation of the precursor after injection
is more important. This process was performed under fluoroscopic guidance (Fig.7). The contrast
media (Ioversol) incorporated in the hydrogel precursor allowed visualization of the infusion process.
The OCMC/CCN sol traversed through the 5-F catheter easily without any clogging, and formed a gel
17
after infusing into the occluded aneurysm sac in the presence of Ioversol as expected from results
mentioned previously (Fig.1). More importantly, upon gelation the gel was also able to effectively fill
the aneurysmal sac (Fig.7f), which was observed previously for Embogel and Ultragel [5]. With
additional drug loading/releasing ability demonstrated above, the hydrogel has the potential to be
mixed with sclerosing agents such as tetracycline and be used during the for EVAR procedure to
Another application of this hydrogel with its drug-delivery feature involved is tumor embolization.
Liquid and particulate embolic materials are effective in penetrating small vessels, and meticulous use
of these materials is essential in preventing unintended occlusions [41]. To address the level of
occlusion of the OCMC/CCN hydrogels in the vascular network, renal embolization was performed
with microspheres as comparison. In all kidneys of four rabbits, the injection of hydrogel and
microspheres through the microcatheters was performed and embolization was done without any
clogging or aggregation. When the concentration of the hydrogel increased from 1.8% to 2.1%, the
Before embolization, the renal arteriogram showed a normal vascular anatomy and homogeneous
renal parenchyma (Fig. 8a). Angiogram immediately after embolization with hydrogel revealed
occlusion of the renal artery (Fig. 8b). The volumes of microsphere suspension needed to achieve total
occlusion were between 5.5 and 7 mL, and the volumes of 1.8% hydrogel precursor needed were
comparable to those of microspheres (Table 2). However, the injection volume dramatically decreased
when the concentration of hydrogel precursor increased from 1.8% to 2.1% (7-8.5 mL vs.1-2 mL).
A follow-up angiogram performed 10 min after embolization confirmed the persistence of the renal
artery occlusion, suggesting the embolization was well maintained (image not shown).
Fig. 9 depicts the histological sections of the rabbit kidneys after embolization. The hydrogel was
18
found more distally than the microspheres in the renal artery. For both 1.8% and 2.1% hydrogel
precursors, hydrogels were observed in the glomeruli of the kidneys embolized. Owing to the
functional groups (-COO-) on the hydrogel matrix, hydrogels appeared to be dark pink under H&E
staining (Fig. 9a and 9c) surrounded with white areas. The white spaces were likely attributed to the
dehydration of the hydrogel (i.e. shrinking) during the staining process procedure [27]. Alternatively it
is also possible that the liquidity of hydrogel precursors allowed them to pass through the arteries to
that level. Composed of polysaccharides, the hydrogels exhibited a purple color under PAS staining
(Fig. 9b and 9d). This was the distinguishing feature of hydrogels from red blood cells which showed
similar pink color in H&E staining but negative response to PAS staining. For the kidneys embolized
with 1.8% hydrogel precursor, hydrogel was observed mainly at the glomerular level with a percentage
of 49.2% of 1869 counted glomeruli although the embolization endpoint was a total occlusion as
confirmed by angiogram (Fig. 8). In the kidneys embolized with 2.1% hydrogel precursor, hydrogels
were found at all levels from interlobar arteries to glomeruli (Fig. 9c, 9d, and 9e). Also 74.9% of the
1907 glomeruli counted contained gels. There were two possible reasons for the different percentages
of glomeruli with hydrogel content when we compared kidneys embolized with 2.1% and 1.8%
hydrogel: First, the 1.8% precursor was diluted by the blood and the gelation was not completed, i.e.,
the formed hydrogel was too weak so that it was partially flushed away by the blood flow. Second, the
weak hydrogel was washed out during the histological processing, as evidenced by lack of hydrogels in
large arteries although a total occlusion was performed. In all histological samples, the microspheres
were seen in round or oval homogenous structures with a pink-purple color at H&E staining in tissue.
They were found either completely blocking the arteries or surrounded by white spaces inside the
arteries (Fig. 9f). Apart from the dehydration of the microsphere, the cutting that was not through the
cross-section of the diameter could be another reason of the white areas, which has also been observed
with Embospheres [27]. The microspheres mainly landed in interlobar and arcuate arteries in 77.8%
19
and 22.2% of the arterioles. The mean (±SD) diameter of the vessel occluded by the microspheres was
224.0 ± 70 µm. There was no microsphere observed in glomeruli. These findings suggest that these
hydrogels can penetrate more distally than microspheres in the vascular network, allowing a better
4. Conclusion
A new in situ forming embolizing hydrogel based on water-soluble natural polysaccharides was
developed and evaluated in regard to its application as an embolizing agent for interventional therapies.
Composed of CMC and CN, this hydrogel is a non-cytotoxic, degradable, and highly porous material
which is capable of loading and releasing drugs. It does not interfere with hemostasis and can
effectively fill the aneurismal sac of the fusiform model with the procedure being fully visualized with
addition of contrast material. Furthermore, this hydrogel successfully embolized kidneys under
fluoroscopy guidance. In addition, various potential therapeutic agents, including cells, drugs and
growth factors, could be mixed into the matrix. Thus, it can be potentially used as an embolizing agent
for different interventional radiology therapies such as endovascular treatment of aortic aneurysm, or
chemotherapy of whole organ tumors, or embolizing bleeding arteries. Further research in the in-vivo
Acknowledgement
We thank Sean L Moen in the Fairview hospital, University of Minnesota Medical Center for his
References
[2] EVAR trial participants. Endovascular aneurysm repair and outcome in patients unfit for open
repair of abdominal aortic aneurysm (EVAR trial 2): randomised controlled trial. Lancet 2005;
20
365:2187–92.
[3] Banerjee S, Sen K, Pal TK, Guha SK. Poly(styrene-co-maleic acid)-based pH-sensitive liposomes
mediate cytosolic delivery of drugs for enhanced cancer chemotherapy. Int J Pharm 2012;
436:786–97.
[4] Golzarian J, Dussaussois L, Ait Said K, Abada HT, Dereume JP, Struyven J. Embolization of large
aneurysms with long wire coils. Cardiovasc Intervent Radiol 2002; 25:26–9.
[5] Barnett BP, Hughes AH, Lin S, Arepally A, Gailloud PH. In vitro assessment of EmboGel and
UltraGel radiopaque hydrogels for the endovascular treatment of aneurysms. J Vasc Interv Radiol
2009; 20:507–12.
[6] Teng MM, Chen CC, Lirng JF, Chen SS, Lee LS, Chang T. N-butyl-2-cyanoacrylate for
[7] Velat GJ, Reavey-Cantwell JF, Sistrom C, et al. Comparison of N-butyl cyanoacrylate and onyx for
[8] Pamuk AG, Saatci I, Cekirge HS, Aypar U. A contribution to the controversy over dimethyl
sulfoxide toxicity: anesthesia monitoring results in patients treated with Onyx embolization for
[9] Coviello T, Matricardi P, Marianecci C, Alhaique F. Polysaccharide hydrogels for modified release
[10] Gong C, Shi S, Wu L, Gou M, Yin Q, Guo Q, et al. Biodegradable in situ gel-forming controlled
[11] Tan L, Liu Y, Ha W, Ding LS, Peng SL, Zhang S, Li BJ. Stimuli-induced gel–sol transition of
21
hydrogel. Soft Matter 2012; 8:5746–9.
[13] Weng L, Romanov A, Rooney J, Chen W. Non-cytotoxic, in situ gelable hydrogels composed of
[14] Cho E, Lee JS, Webb K. Formulation and characterization of poloxamine-based hydrogels as
[15] Mclemore R, Preul MC, Vernon BL. Controlling delivery properties of a waterborne, In-Situ-
[16] S. Cohen, E. Lobel, A. Trevgoda, Y. Peled, A novel in situ-forming ophthalmic drug delivery
system from alginates undergoing gelation in the eye. J Control Release 1997; 44:201–8.
Injectable in situ forming drug delivery system based on poly(ε-caprolactone fumarate) for
tamoxifen citrate delivery: Gelation characteristics, in vitro drug release and anti-cancer evaluation.
[18] Brennecka CR, Preul MC, Bichard WD, Vernon BL. In vivo experimental aneurysm embolization
in a swine model with a liquid-to-solid gelling polymer system: Initial biocompatibility and
[19] Peng HT, Blostein MD, Shek PN. Experimental optimization of an in situ forming hydrogel for
[20] Fatimi A, Chabrot P, Berrahmoune S, Coutu JM, Soulez G, Lerouge S. A new injectable
[21] Halak M, McDonnell CO, Muhlmann MD, Baker SR. Open surgical treatment of aneurysmal sac
22
expansion following endovascular abdominal aneurysm repair: solution for an unresolved clinical
2012;87:457–60.
[23] Chen XG, Park HJ. Chemical charatersitics of O-carboxymethyl chitosan related to the preparation
[24] Poon YF, Zhu YB, Shen JY, Chan-Park MB, Ng SC. Cytocompatible hydrogels based on
[25] Ito T, Yeo Y, Highley CB, Bellas E, Benitez CA, Kohane DS. The prevention of peritoneal
[26] Weng L, Le HC, Talaie R, Golzarian J. Bioresorbable hydrogel microspheres for transcatheter
embolization: preparation and in vitro evaluation. J Vasc Interv Radiol 2011; 22:1464–70.
[27] Laurent A, Wassef M, Saint Maurice JP, et al. Arterial distribution of calibrated tris-acryl gelatin
and polyvinyl alcohol microspheres in a sheep kidney model. Invest Radiol 2006; 41:8–14.
[28] Sternlicht M, Sales SF, Daniels JR, Daniels A. Renal cisplatin chemoembolization with Angiostat,
Gelfoam, and Ethiodol in the rabbit: renal platinum distributions. Radiology 1989; 170:1073–5.
[29] Jia X, Burdick JA, Kobler J, Clifton RJ, Rosowski JJ, Zeitels SM, Langer R. Synthesis and
[30] Kwak BK, Shim HJ, Han SM, Park ES. Chitin-based embolic materials in the renal artery of
23
embolic agent: an experimental study in a rabbit renal model. Korean J Radiol 2007; 8:418–28.
[32] Shah HK, Conkie JA, Tait RC, Johnson JR, Wilson CG. A novel, biodegradable and reversible
polyelectrolyte platform for topical-colonic delivery of pentosan polysulphate. Int J Pharm 2011;
404(1-2):124–32.
[34] Salooja N, Perry DJ. Thrombelastography. Blood Coagul Fibrinolysis 2001; 12:327–37.
[35] Windsor PG, Como JA, Windsor KS. Sclerotherapy for malignant pleural effusions: alternatives to
[36] Wu Y, Fassihi R. Stability of metronidazole, tetracycline HCl and famotidine alone and in
[37] Nemoto EM, Frinak S. Brain tissue pH after global brain ischemia and barbiturate loading in rats.
Stroke 1981;12:77–82.
[38] L Weng, HC Le, J Lin, J Golzarian. Doxorubicin loading and eluting characteristics of
[39] Namur J, Citron SJ, Sellers MT. Embolization of hepatocellular carcinoma with drug –eluting
beads: doxorubicin tissue concentration and distribution in patient liver explants. J Hepatology
2011;55:1332–8.
[40] EVAR trial participants. Endovascular aneurysm repair and outcome in patients unfit for open
repair of abdominal aortic aneurysm (EVAR trial 2): randomised controlled trial. Lancet 2005;
365:2187–92.
[41] Lazzaro MA., Badruddin A, Zaidat OO, Darkhabani Z, Pandya DJ, Lynch JR. Endovascular
24
Figure Legends
Figure 1. Gelation time of the hydrogel precursors: (a) different polymer concentrations with and
without radiographic contrast; (b) different solvents and temperatures.
Figure 3. Cytotoxicity study of the hydrogels: (a) the cells cultured with hydrogel piece; (b) monolayer
cell control. (c) the cell viability (MTT) assay indicating comparing optical density (OD) of
cells’ reaction to MTT reagent with and without hydrogel.
Figure 4. In vitro degradation of the hydrogels in the lysozyme PBS solutions at 37 °C.
Figure 5. Thromboelastography showing the blood coagulation in the presence of hydrogel (green) as
compared to saline control (black).
Figure 6. Drug release from the hydrogel at (a) Tetracylcine release at pH=7.4 (0.01 M PBS solution),
pH=6 (PBS solution with HCl), and pH=2 (PBS solution with HCl) at 37 oC; (b) Doxorubicin
release in the PBS solution (pH=7.4) at 37 oC.
Figure 7. The stages of filling a fusiform aneurysmal model under fluoroscopic guidance with the
hydrogel showing clear visualization as well as injectability of the hydrogel material through
a 5-French catheter. (a) 0 s, catheter positioned; (b) 15 s, filling starting; (c) 31 s, half filling;
(d) 78 s, complete filling.
Figure 8. Angiogram images of the renal arteries before embolization (a) and after embolization (b)
with hydrogel.
Figure 9. Histological sections of the rabbit kidneys. (a) and (b) Hydrogels in the glomeruli of the
rabbit kidney embolized with 1.8% hydrogel precursor stained by H&E and PAS, respectively.
Green arrow: hydrogel; black arrow: red blood cell. (c) and (d) Hydrogels in the glomeruli of
the kidney embolized with 2.1% hydrogel precursor stained by H&E and PAS, respectively.
Arrow: hydrogel. (e) Hydrogels in the interlobular arteries, H&E staining. Arrow: hydrogel.
(f) A microsphere in the arcuate arteries, H&E staining. Arrow: microsphere.
25
Table 1 TEG measurements of citrated whole human blood in the presence of normal saline or
hydrogel precursor.
Table 2 Volume of embolic materials needed for total occlusion of the rabbit renal artery.
26
Figure 1-3
Figure 1
a 9
8
without contrast
Gelation Time (min) 7 with 20% (v/v) contrast
6
0
1.5% 1.8% 2.0%
Concentration of the hydrogel precursor
b 5
4
Gelation Time (min)
2% hydrogel precursor
3
0 o o o o o
DI, 2
3 C
DI, 3
7 C
ne, 37 C BS, 37 C a, 37 C
Sali m
P Plas
Solvent and temperature
A
Figure 2
Figure 3
a b
c
1.4
1.17 1.19
1.2
Absorbance at 570 nm
1.0
0.8
0.6
0.4
0.2
0.0
Hydrogel Control
Figure 4-6
Figure 4
100
4 mg/mL
Weight percentage (%)
80 10 g/mL
60
40
20
0 3 6 9 12 15 18 21
Time (day)
Figure 5
Figure 6
a 90
Cumulative released Tetracycline (%)
80
70
60
50
40
30
20
pH=2
pH=6
10 pH=7.4
-10 0 10 20 30 40 50 60 70 80
Time (h)
b 40
Cumulative released Doxorubicin (%)
35
30
25
20
15 pH=7.4
10
-5
-10 0 10 20 30 40 50 60 70 80 90
Time (h)
Figure 7-8
Figure 7
d
Figure 8
a b
Figure 9
Figure 9