February 2012
Volume 53, Issue 2
Free
Retina  |   February 2012
The Cross-linked Biopolymer Hyaluronic Acid as an Artificial Vitreous Substitute
Author Affiliations & Notes
  • Charlotte Schramm
    From the Department of Ophthalmology, Section Experimental Ophthalmic Surgery,
  • Martin S. Spitzer
    From the Department of Ophthalmology, Section Experimental Ophthalmic Surgery,
  • Sigrid Henke-Fahle
    From the Department of Ophthalmology, Section Experimental Ophthalmic Surgery,
  • Gabriele Steinmetz
    From the Department of Ophthalmology, Section Experimental Ophthalmic Surgery,
  • Kai Januschowski
    From the Department of Ophthalmology, Section Experimental Ophthalmic Surgery,
  • Peter Heiduschka
    From the Department of Ophthalmology, Section Experimental Ophthalmic Surgery,
  • Jürgen Geis-Gerstorfer
    the Department of Prosthodontics, Center of Dentistry, Oral Medicine and Maxillofacial Surgery, and
  • Tilo Biedermann
    the Department of Dermatology, Eberhard-Karls University Tübingen, Tübingen, Germany.
  • Karl U. Bartz-Schmidt
    From the Department of Ophthalmology, Section Experimental Ophthalmic Surgery,
  • Peter Szurman
    From the Department of Ophthalmology, Section Experimental Ophthalmic Surgery,
  • Corresponding author: Martin S. Spitzer, FEBO, Department of Ophthalmology, University Eye Hospital Tübingen, 72076 Tübingen, Germany; martin.spitzer@med.uni-tuebingen.de
Investigative Ophthalmology & Visual Science February 2012, Vol.53, 613-621. doi:10.1167/iovs.11-7322
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Charlotte Schramm, Martin S. Spitzer, Sigrid Henke-Fahle, Gabriele Steinmetz, Kai Januschowski, Peter Heiduschka, Jürgen Geis-Gerstorfer, Tilo Biedermann, Karl U. Bartz-Schmidt, Peter Szurman; The Cross-linked Biopolymer Hyaluronic Acid as an Artificial Vitreous Substitute. Invest. Ophthalmol. Vis. Sci. 2012;53(2):613-621. doi: 10.1167/iovs.11-7322.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: Biopolymers are promising substances in the development of a new vitreous substitute to overcome the drawbacks associated with current hydrophobic tamponade materials.

Methods.: Different hydrogels were assembled by cross-linking hyaluronic acid either with adipic dihydrazide (ADH) by carboxylation with N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDCI) after hydrazation or by photocrosslinking with UV-light and N-vinyl-pyrrolidinone. The refractive index and rheologic properties of the obtained gels were investigated. To quantify the degradation of the hydrogels over time, free hyaluronic acid was measured photometrically by means of the degradation product uronic acid. For biocompatibility testing, the hydrogels were applied on top of cultured retinal pigment epithelial (RPE) cells and analyzed by the cell viability, MTT, and alamar blue viability cytotoxicity assays and flow cytometry, with Annexin V-FITC and propidium iodide co-staining. The in vivo biocompatibility of the hydrogels was tested in vitrectomized rabbit eyes for up to 6 weeks.

Results.: The synthesized hydrogels were all clear and transparent and had a refractive index similar to human vitreous. The rheologic measurements suggested sufficient viscosity and elasticity for intraocular use. Quantification of the degradation products revealed only a small decay of the gels over 1 month. However, the ADH cross-linked hydrogels induced mild cytotoxicity in the RPE cells. The UV cross-linked hydrogels showed no toxicity or induction of apoptosis. In vivo the UV cross-linked biogels remained in place for 6 weeks, and electrophysiology and histology showed excellent tissue biocompatibility.

Conclusions.: Biopolymers based on UV cross-linked hyaluronic acid may be promising vitreous substitutes.

Vitreous substitutes are needed for tamponading the reattached retina after vitrectomy for retinal detachments. An ideal vitreous substitute has to fulfill several requirements. Apart from crucial physical properties, such as optical transparency, high viscosity, and stability, it must possess excellent biocompatibility because it is in intimate contact with the neuroretina. However, several drawbacks are associated with currently available vitreous substitutes. The most commonly used ones include air, expansive gases, and silicone oil, but their clinical application is limited by emulsification, inflammatory reaction and secondary glaucoma. 1 All commercially available tamponades have in common that they are hydrophobic and act through buoyancy and interfacial tension. However, the hydrophobic nature of these tamponades is responsible for the major disadvantage that a complete filling of the vitreous cavity is not possible. With all hydrophobic tamponades a fluidic space remains, allowing growth factors to accumulate. This often results in reproliferation and redetachment of the lower retina. Newly developed materials that are heavier than water, such as heavy silicone oil, unfortunately did not solve this problem but rather shifted it to the upper part of the retina. 1 4  
To address these problems, several natural hydrophilic polymers such as collagen and hyaluronic acid (HA) have been proposed. 5,6 In recent years, the use of synthetic polymers has become a focus of research. 7 Chirila et al. 8 and Hong et al. 9,10 performed extensive studies on the synthesis, characterization, and application of hydrophilic polymers as vitreous substitutes. They greatly improved biocompatibility, but the recurrent problems were the long-term stability and the inability of these compounds to be injected through a small-gauge needle. 7,8,10,11  
HA is a part of the human vitreous, and so it is predestined to be a basic material for the development of a three-dimensional vitreous substitute. 12,13 HA is a glycosaminoglycan co-polymer of d-glucuronic acid and N-acetyl-d-glucosamine, which are connected through alternating β-1,4 and β-1,3 glycosidic bonds. The axial hydrogen atoms form a nonpolar, hydrophilic face, thereby creating a twisting ribbon structure. Consequently, a hyaluronan molecule assumes an expanded random coil structure in physiological solutions which occupies a very large domain, and so HA forms a viscous compound. Furthermore, this allows an exceptionally high swelling factor of up to 10,000. 14,15  
HA is commonly used in ophthalmic surgery, but the native form is not suitable as a vitreous substitute because it degrades rapidly and may lead to an undesired increase in intraocular pressure. 
This study demonstrates and compares two ways to create a three-dimensional hydrogel with long-term biostability for a vitreous substitute based on cross-linked hyaluronic acid (CHA). 
Materials and Methods
In the methods described herein, HA from rooster comb with a molecular weight of 1 to 4 × 106 was used for the synthesis of the cross-linked hydrogels. In both methods we tested different concentrations of the cross-linking components in preceding studies. Then, the best variant was selected. 
Synthesis of a Cross-linked Hydrogel Using Dihydrazides
HA may be cross-linked by means of dihydrazides, which creat stable bonds between the HA molecules. Here, the carboxyl groups of HA play a crucial role. Dihydrazide cross-linked hydrogels were manufactured by using a modified method described by Prestwich et al. 16 and Vercruysse et al. 17  
For cross-linking, 0.08 g HA (Sigma-Aldrich, Munich, Germany) was dissolved in 8 mL distilled water to get a solution of 1% HA. Forty microliters of adipic dihydrazide (ADH) dissolved in distilled water (Sigma-Aldrich) and 40 μL n′(3-dimethylaminopropyl)-n′-ethylcarbodiimide hydrochloride (EDCI) 98% (Sigma-Aldrich), also dissolved in distilled water, were added separately and injected through a 20-gauge needle. The solution was incubated under constant stirring for 4 hours at 4°C, followed by an incubation period of 24 to 48 hours. The value was reduced to approximately 4.0 by adding 46 μL 1% HCl. This mixture was incubated for another 24 hours while the cross-linking reaction proceeded and a cross-linked ADH-HA hydrogel was formed. The hydrogels were dialyzed in PBS for 24 to 48 hours to remove remnants of the cross-linkers. 
Synthesis of a Cross-linked Hydrogel by Photocrosslinking
Cross-linked hydrogels based on photocrosslinking were synthesized according to a protocol modified from Leach et al. 18,19 creating bonds between glycidyl methacrylate-hyaluronic acid (GMHA) conjugates and N-vinyl-pyrrolidinone. 
To form GMHA conjugates, we dissolved 0.05 g HA in 5 mL phosphate-buffered saline (PBS) to get a solution of 1% HA. Then 110 μL triethylamine, 110 μL glycidyl methacrylate, and 110 mg tetrabutylammonium bromide (Sigma-Aldrich) were gradually added while stirring. After the solution was stirred overnight at room temperature, it was incubated for 1 hour at 60°C and then precipitated in acetone. The GMHA conjugates obtained were dissolved in 5 mL distilled water and stored at 4°C. To synthesize cross-linked UV-HA hydrogels, 250 μL N-vinyl-pyrrolidinone and 0.05 g of a photoinitiator (Irgacure 2959; Ciba Specialty Chemicals, Basel, Switzerland) were added to the 5 mL of dissolved GMHAs and exposed to UV light for 5 to 10 minutes (365 nm, ∼20 mW/cm2). The hydrogels were dialyzed in PBS for 24 to 48 hours to remove remnants of the cross-linkers. 
Biodegradation of the Hydrogels
Four milliliters of the respective hydrogels were placed in PBS. Over 4 weeks, the PBS was changed every 48 hours and the HA concentration was measured. HA degradation was quantified by a photometric uronic acid assay. 20 Briefly, 191 mg sodium borate was dissolved in 20 mL concentrated sulfuric acid to obtain solution 1. To produce solution 2, 0.0125 g carbazole was dissolved in 9.9875 mL absolute ethanol. Then, the samples were transferred into a cooled 15-mL tube containing 3 mL of solution 1 and carefully mixed. After 10 minutes of incubation in boiling water, the mixture was cooled to room temperature and then 0.5 mL of solution 2 was added. The well-mixed solution was incubated for 15 minutes in boiling water. It was kept at room temperature for 2 hours and measured photometrically at 540 nm with a photometer (Ultrospec 1000; Pharmacia Biotech, Munich, Germany). 
Each sample was measured four times and the mean value was calculated. HA concentration was determined by comparison with the previously obtained calibration curve of pure HA. 
Biocompatibility
Cell Viability Staining.
Cell viability was assessed with a commercial assay (Live/Dead Viability/Cytotoxicity Kit 1, L-7013; Invitrogen-Molecular Probes, Eugene, OR). The hydrogels and medium containing 5% fetal calf serum were added to the cells. Staining was performed according to the manufacturer's instructions. Adequate negative (cells without the respective hydrogel) and positive (cells treated with Triton X-100 [Serva, Heidelberg, Germany] 0.3% in PBS) controls were run for each set of experiments. The results were analyzed by fluorescence microscopy after 24 hours of incubation from the ARPE19-cells (ATCC, Manassas, VA) with the CHA hydrogels. 
MTT Assay.
To assess the cytotoxicity of the hydrogels on proliferating ARPE-cells, we performed the 3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazoliumbromide (MTT; Sigma-Aldrich) assay. The CHA hydrogels and medium containing 5% fetal calf serum were added to the ARPE19 cells. After 72 hours, the cells were washed with PBS, and fresh medium with MTT at 0.5 mg/mL was added to the cells. After 2 hours of incubation, formazan extraction was performed, and the quantity was measured photometrically with an ELISA plate reader (SLT Spectra 400; ATX, Salzburg, Austria) at a wavelength of 570 nm. 
Alamar Blue Cytotoxicity Assay.
The Alamar blue assay (BioSource, Solingen, Germany) is designed to measure quantitatively the cytotoxicity and the proliferation of human cells. The CHA hydrogels were incubated with ARPE19-cells and the nonfluorescent form of Alamar blue for 48 or 72 hours. The reduced fluorescent form was monitored photometrically (SLT Spectra 400; ATX) at 590 nm. 
Apoptosis Assay.
Flow cytometry for apoptosis was performed, to investigate cellular responses after exposure to the hydrogels. In this test the cross-linked hydrogels were placed in a Boyden chamber on top of filter inserts with the ARPE19-cells in the bottom chamber, to avoid mechanical interaction of the hydrogel with the cells. After 24 hours of incubation, an Annexin V FITC apoptosis detection kit (Calbiochem, Darmstadt, Germany) was used to identify early apoptotic cells and to discriminate between late apoptotic/necrotic and living cells. For simultaneous detection a co-staining technique with fluorochrome-conjugated annexin V in tandem with the DNA-binding dye propidium iodide (PI) was used according to the manufacturer's instructions. For simultaneous scoring of the differential cellular response, aliquots of 104 cells were each immediately analyzed by flow cytometry (FACSCalibur; BD Biosciences, Mountain View, CA) and the data were analyzed (CellQuest software; BD Biosciences). 
Measurement of the Refraction Index and Rheologic Properties
Refraction Index.
The refractive index of the hydrogels was obtained with a refractometer (AR4/AR4D Abbe; Krüss Optronic, Hamburg, Germany). The samples were prepared as previously described, sliced, transferred to the refractometer, and measured at room temperature with a wavelength of 590 nm. 
Rheologic Measurements.
The rheologic analysis was performed using with a rheometer (UDS 200; Paar-Physica GmbH, Radeburg, Germany) with a titanium cone, at 37°C. The measurements were performed with the CRS (controlled shear rate) test. The shear rate at the beginning of the measurements was 0.1/s and steadily increased to 100/s. The viscosity was measured over time in Pascals per second. The cross-linked hydrogels were measured in sliced pieces (diameter, 2 cm; thickness, 1.5 mm) or injected onto the plate through needles with an inner diameter of 0.6 (23-gauge) and 0.7 or 0.9 mm (20-gauge). 
In Vivo Experiments.
The vitreous was removed by standard three-port pars plana vitrectomy in 12 rabbit eyes and replaced by injection of the UV-CHA hydrogels through 20-gauge cannulas. In each case the right eye was operated on and the left eye served as the control. Animals were cared for and handled in accordance with the ARVO Statement for the Use of Animals in Vision and Ophthalmic Research and in accordance with institutionally approved protocols. Electroretinogram (ERG), slit lamp examination, funduscopy, and measurement of intraocular pressure with a Schiötz tonometer and a handheld tonometer (Tono-Pen; Reichert Technologies, Depew, NY) were performed at regular intervals after surgery. The eyes were enucleated 6 weeks after surgery, and histology was assessed for signs of cytotoxicity. 
Results
With both methods three-dimensional hydrogels in variable forms could be produced. The hydrogels for the following experiments had a volume of 5 or 8 mL and seemed to achieve the desired physical properties, such as stability, elastic or rubberlike structure, and optical transparency (Fig. 1). 
Figure 1.
 
With both methods, either cross-linking HA with adipic dihydrazides (a) or cross-linking HA with N-vinyl-pyrrolidinone and photopolymerization (b), optical transparent, rubber-like and three-dimensional hydrogels were produced.
Figure 1.
 
With both methods, either cross-linking HA with adipic dihydrazides (a) or cross-linking HA with N-vinyl-pyrrolidinone and photopolymerization (b), optical transparent, rubber-like and three-dimensional hydrogels were produced.
In Vitro Degradation
The cross-linked hydrogels remained stable over months. Consistency, transparency, and three-dimensional structure were maintained throughout. 
The ADH-HA hydrogels showed a degradation of 7.43 ± 0.62 mg HA during the first 30 days (80 mg cross-linked HA on day 1; Fig. 2a). Measurements of UV-HA displayed a similar degradation of 5.81 ± 0.22 mg HA during the first 15 days, and the degradation continued at a very slow rate for the next 15 days (Fig. 2b). The release occurred in two phases until it reached steady rate. To specify the degradation, the following formula was developed for both hydrogels (Table 1):    
Figure 2.
 
HA-degradation from the ADH-HA (a) and the UV-HA hydrogels (b). The degradation was separated into two phases. The first showed a faster degradation from 7.487 mg HA from 80 mg cross-linked ADH-HA (a) in 30 days and 5.634 mg HA from 50 mg cross-linked UV-HA (b) in 15 days. Thereafter, in both cases, the rate of HA-degradation slowed down.
Figure 2.
 
HA-degradation from the ADH-HA (a) and the UV-HA hydrogels (b). The degradation was separated into two phases. The first showed a faster degradation from 7.487 mg HA from 80 mg cross-linked ADH-HA (a) in 30 days and 5.634 mg HA from 50 mg cross-linked UV-HA (b) in 15 days. Thereafter, in both cases, the rate of HA-degradation slowed down.
Table 1.
 
Factors from the Regression Formula of HA Degradation from Cross-linked Hydrogels
Table 1.
 
Factors from the Regression Formula of HA Degradation from Cross-linked Hydrogels
ADH-HA UV-HA
c (t = 0) 80000 80000
k1 4845.476667 4974.295
t (50,1) 7.097 1.4288
n1 0.892 0.98665
k2 5952.04 1186.15
t (50,2) 27.39 11.63
n2 3.798333333 10.522
In Vitro Biocompatibility of the CHA hydrogels
After incubation for 24 hours of RPE cells with ADH-HA cross-linked hydrogels (Fig. 3), an increased number of red fluorescent dead cells were found (Fig. 3.3c). Even when the ADH-HA cross-linked hydrogels were dialyzed over 24 to 48 hours in PBS, the number of dead cells remained higher than in the controls (Fig. 3.5c). 
Figure 3.
 
After an incubation period of 24 hours with the CHA hydrogels, the cell viability of the RPE cells was assessed by staining the live and dead cells with calcein AM (b, live cells) and EthD-1 (c, dead cells). To exemplify the RPE cell structure, images were obtained with reflected-light microscopy or after staining the nucleus with 4′6′-diamidino-2-phenylindole (DAPI) (a). Compared with controls of live (1) and dead (2) cells, an intense red fluorescence as a signal for dead cells in the group under incubation with the ADH-HA hydrogels was initially observed (3). However, after dialysis over 24 to 48 hours, the RPE cells incubated with UV-HA hydrogels (4) showed no increased signs of cytotoxicity compared with the control. However, RPE cells incubated with ADH-HA hydrogels (5), showed an increased number of red fluorescent dead cells and the cell groups appeared to be disrupted in some areas even after dialyzing the gels for 24 to 48 hours.
Figure 3.
 
After an incubation period of 24 hours with the CHA hydrogels, the cell viability of the RPE cells was assessed by staining the live and dead cells with calcein AM (b, live cells) and EthD-1 (c, dead cells). To exemplify the RPE cell structure, images were obtained with reflected-light microscopy or after staining the nucleus with 4′6′-diamidino-2-phenylindole (DAPI) (a). Compared with controls of live (1) and dead (2) cells, an intense red fluorescence as a signal for dead cells in the group under incubation with the ADH-HA hydrogels was initially observed (3). However, after dialysis over 24 to 48 hours, the RPE cells incubated with UV-HA hydrogels (4) showed no increased signs of cytotoxicity compared with the control. However, RPE cells incubated with ADH-HA hydrogels (5), showed an increased number of red fluorescent dead cells and the cell groups appeared to be disrupted in some areas even after dialyzing the gels for 24 to 48 hours.
In contrast, the fluorescence staining of RPE cells, which were incubated with UV-HA hydrogels that had been dialyzed for 24 to 48 hours, showed no difference compared with the control (Fig. 3.4c). 
The cellular metabolic activity assessed by the MTT test of the RPE cells incubated for 48 hours with the ADH-HA hydrogels was not significantly different from the control group (Fig. 4, P > 0.05). However, in the Alamar blue assay, a significant difference was observed after 72 hours of incubation (Fig. 5b; P < 0.05). 
Figure 4.
 
MTT test: RPE cells after incubation for 48 hours with CHA hydrogels. No significant differences were found between the mitochondrial activity of the RPE cells incubated with UV-HA or ADH-HA hydrogels compared with the controls (P > 0.05).
Figure 4.
 
MTT test: RPE cells after incubation for 48 hours with CHA hydrogels. No significant differences were found between the mitochondrial activity of the RPE cells incubated with UV-HA or ADH-HA hydrogels compared with the controls (P > 0.05).
Figure 5.
 
To quantify the cytotoxicity of the CHA hydrogels on RPE cells and their effects on cell proliferation, RPE cells were incubated for 72 hours with the CHA hydrogels and Alamar blue. The mitochondrial activity of RPE cells incubated with the UV-HA hydrogels (a) was not significantly different from that in the control group (P > 0.05). The mitochondrial activity of RPE cells incubated with the ADH-HA hydrogels (b) was significantly higher than that in the control group (*P < 0.05).
Figure 5.
 
To quantify the cytotoxicity of the CHA hydrogels on RPE cells and their effects on cell proliferation, RPE cells were incubated for 72 hours with the CHA hydrogels and Alamar blue. The mitochondrial activity of RPE cells incubated with the UV-HA hydrogels (a) was not significantly different from that in the control group (P > 0.05). The mitochondrial activity of RPE cells incubated with the ADH-HA hydrogels (b) was significantly higher than that in the control group (*P < 0.05).
In contrast, the metabolic activity of the RPE cells incubated with the UV-HA hydrogels was not significantly different from that in the control group after 48 hours, in the MTT test (Fig. 4, P > 0.05), or after 72 hours in the Alamar blue assay (Fig. 5a; P > 0.05). Thus, it is reasonable to assume that the UV-HA hydrogels have no toxic or anti- or pro-proliferative influence on RPE cells. 
In all samples tested with ADH-HA and UV-HA, FACS analysis showed a high rate of living cells (range, 95%–99%), which did not differ significantly from the control (Fig. 6, P > 0.05). The same was true for the low rate of apoptotic or necrotic cells. 
Figure 6.
 
Representative Annexin V and PI dot plots for the control groups with live (a) and dead (b) cells and UV-HA (c) and ADH-HA (d) hydrogels. (Q1, Q2) Dead and dying cells (PI positive). (Q3) Living cells without signs of apoptosis (Annexin V and PI negative). (Q4) Early stages of apoptosis (Annexin V positive and PI negative).
Figure 6.
 
Representative Annexin V and PI dot plots for the control groups with live (a) and dead (b) cells and UV-HA (c) and ADH-HA (d) hydrogels. (Q1, Q2) Dead and dying cells (PI positive). (Q3) Living cells without signs of apoptosis (Annexin V and PI negative). (Q4) Early stages of apoptosis (Annexin V positive and PI negative).
However, the biocompatibility tests showed that the ADH-HA and the UV-HA hydrogels had to be dialyzed to remove remnants of the cross-linkers and thus achieve good biocompatibility. Nevertheless, RPE cells that were incubated with the ADH-HA cross-linked hydrogels still showed a slightly increased rate of cell death and broken cell structures in the cell viability staining. Thus, the dialyzed ADH-HA cross-linked hydrogels were not tested further. 
Refraction Index and Rheologic Experiments
The cross-linked hydrogels consistently revealed a refractive index of 1.338 in all experimental series. This result is almost identical with that in human vitreous (1.336) or water. Interestingly, the refractive index did not differ when large or small pieces of the hydrogels (after injection through small gauge needles) were measured. 
However, the dynamic viscosity of the hydrogels was dependent on the diameter of the needles that were used for injection (Fig. 7). With a 0.9-mm-diameter (20-gauge) needle at a low shear rate of approximately 0.1/s, the dynamic viscosity was 93.65 ± 28.91 Pa s. The dynamic viscosity at this starting point was 60.64 ± 28.9 Pa · s, with a 0.7-mm-diameter needle, and 76.89 ± 30.52 Pa · s, with the UV-HA as a hole piece. The dynamic viscosity of hydrogels injected through a 0.6-mm (23-gauge) needle was even lower, and the measurement of the dynamic viscosity was feasible only at a shear rate of 3448/s or higher. With higher shear rate, the viscosity decreased exponentially. 
Figure 7.
 
The dynamic viscosity was measured in reference to the shear rate. UV-HA was measured in a whole piece and injected through needles with a diameter of 0.6, 0.7, or 0.9 mm.
Figure 7.
 
The dynamic viscosity was measured in reference to the shear rate. UV-HA was measured in a whole piece and injected through needles with a diameter of 0.6, 0.7, or 0.9 mm.
In Vivo Biocompatibility in the Rabbit Eye
After pars plana vitrectomy the UV-CHA hydrogels were easily injected via the pars plana into the vitreous cavity of the rabbit eye. However, in four eyes, iatrogenic retinal detachment occurred during the preceding vitrectomy. Nevertheless, these eyes were also filled with UV-CHA and included in the follow-up. On slit lamp examination and funduscopy no signs of inflammation or disintegration of the hydrogels was visible. Intraocular pressure was normal during the entire follow-up. Electrophysiology was consistent with an undisturbed retinal function (Fig. 8). After 6 weeks, when the eyes were enucleated, the vitreous cavity of the operated eyes was still completely filled with hydrogel (Fig. 9A). The transparency and viscoelasticity of the injected UV-CHA was grossly unchanged since injection 6 weeks earlier. Good biocompatibility was further confirmed by histology (Fig. 9B). Interestingly, the retina remained attached in three of four cases with iatrogenic retinal detachments although the retinal detachments were not specifically treated (e.g., no laser treatment of retinal breaks or intraoperative use of heavy fluids to flatten the retina) except by the instillation of UV-CHA. The lens remained clear throughout the entire follow-up period, apart from two cases of iatrogenic intraoperative lens contact. 
Figure 8.
 
ERG responses in the hydrogel vitreous substitute group (right eye) in comparison to the control (left eye). Top: bar graphs represent the ratio of the B-wave amplitudes between the operated eye (UV-CHA filled) and the nonoperated eye. Bottom: the ratio of the B-wave latencies between the operated eye (UV-CHA filled) and the nonoperated eye. There was no statistically significant difference between the surgery and the control groups. Data the mean ± SD; P > 0.05.
Figure 8.
 
ERG responses in the hydrogel vitreous substitute group (right eye) in comparison to the control (left eye). Top: bar graphs represent the ratio of the B-wave amplitudes between the operated eye (UV-CHA filled) and the nonoperated eye. Bottom: the ratio of the B-wave latencies between the operated eye (UV-CHA filled) and the nonoperated eye. There was no statistically significant difference between the surgery and the control groups. Data the mean ± SD; P > 0.05.
Figure 9.
 
Macroscopic and histologic findings 6 weeks after implantation of UV-CHA. (A) The vitreous cavity of the operated eyes was still completely filled with the hydrogels. The transparency and viscoelasticity of the injected UV-CHA was grossly unchanged since injection 6 weeks ago. The lens was clear. (B) Histology revealed unaltered retinal structure and cell morphology (H&E; ×200).
Figure 9.
 
Macroscopic and histologic findings 6 weeks after implantation of UV-CHA. (A) The vitreous cavity of the operated eyes was still completely filled with the hydrogels. The transparency and viscoelasticity of the injected UV-CHA was grossly unchanged since injection 6 weeks ago. The lens was clear. (B) Histology revealed unaltered retinal structure and cell morphology (H&E; ×200).
Discussion
A hydrogel vitreous substitute based on HA has many theoretical advantages in terms of biocompatibility and tamponade properties, because it mimics the characteristics of the natural human vitreous body. The CHA three-dimensional hydrogels combine many of the vitreous properties such as optical transparency, high viscosity, and long-term stability as well as a good biocompatibility. Lee et al. 21,22 found the viscosity of healthy human vitreous to be 300 to 2000 cP (0.300–2 Pa · s). However, the viscosity of our CHA hydrogels was many times higher, thus making it an ideal vitreous replacement in situations in which a strong tamponade effect is desired. Apart from this tamponade effect, a high swelling factor for good retinal reattachment after surgery has been mentioned as advantageous in the literature. 23 However, pronounced swelling of a material inside the vitreous cavity may also cause undesired increases in intraocular pressure. Only mild swelling of the CHA used in this study was noted. 
This study demonstrated that hydrophilic hydrogels like CHA may replace in principle the hydrophobic silicone oils as a long-term vitreous substitute. Currently, only silicone oil is available as a long-term vitreous replacement. Silicone oils have multiple disadvantages when they are employed as a vitreous substitute. These include the risk of secondary glaucoma, cataract formation, hyperopic shift, and the need for additional surgery to remove the silicone oil. 11,24 26 Apart from transparency, stability, sufficient viscosity, and biocompatibility, another practical advantage of the UV-CHA hydrogels is their hydrophilic properties in contrast to the hydrophobic silicone oils. These properties facilitate a complete filling of the vitreous cavity and a good tamponade effect. 
Because of their hydrophilic characteristics, the CHA hydrogels may be used as delivery systems for sustained drug release, as described for various other hydrogels and biopolymers. 26 29 Hydrophilic vitreous substitutes such as CHA can be loaded with drugs for prolonged release into the vitreous cavity. 
Previous experiments demonstrated that HA itself is not toxic. It is widely used in ophthalmology and has shown good clinical biocompatibility. 5,24,3031 The experiments with RPE cells demonstrated a difference with respect to biocompatibility between the hydrogels cross-linked with dihydrazides and the UV-CHA hydrogels. Whereas the UV-CHA hydrogels did not show any signs of cytotoxicity or adverse effects on cell proliferation, the experiments with the ADH-HA cross-linked hydrogels did not entirely prohibit harmful biological effects, especially in the cell viability staining. Another finding with the ADH-HA cross-linked gel was the slightly increased cellular mitochondrial activity observed in the MTT assay, which was confirmed by the Alamar blue assay. This finding may be critical with respect to a potentially enhancing effect on proliferative vitreoretinopathy. These results differ slightly from those obtained by Su et al., 32 who showed no toxic effect of similar cross-linked ADH-HA hydrogels on RPE cells. Indeed they added more steps to the reactions before cross-linking HA, so the toxic substances could be removed earlier. However, in our case it seems likely that after cross-linking, the ADH-HA hydrogels retained some of the cross-linkers that induced the observed cytotoxic effects. This emphasizes the importance of meticulous washing. 
The UV-CHA hydrogels showed improved biocompatibility after dialysis. This finding is supported by previous studies with variable photocrosslinkable hydrogels also displaying good biocompatibility. 33 According to these results, we regarded the UV-CHA hydrogels as a more suitable vitreous substitute and proceeded with further experiments using only these gels. 
The refractive index of CHA hydrogels is similar to human vitreous or water. This seems to be a common trait for hydrophilic hydrogels. 32,34 This characteristic may be helpful in visual rehabilitation. In contrast, silicone oils have a refractive index of approximately 1.4 and induce a hyperopic shift of approximately 4 to 6 D after surgery. 35  
For application during vitreoretinal surgery the cross-linked hydrogel should be injectable through a needle. Our rheologic experiments showed that after an injection through a 20-gauge needle, the viscosity at low shear rate remained high compared with that of silicone oil with a kinetic viscosity of 5000 cSt, which is approximately equivalent to a dynamic viscosity of 4.85 Pa · s. The dynamic viscosity of the cross-linked hydrogels was much higher (61–94 Pa · s). The results are supported by a study by D'Errico et al., 36 who analyzed UV cross-linked N-vinyl-pyrrolidinone hydrogels. N-vinyl-pyrrolidinone is also a component of UV-CHA. The high viscosity of the cross-linked hydrogel is essential for the tamponade effect and may provide mechanical resistance against proliferative vitreoretinopathy. 37 Furthermore, a high viscosity may avoid turbulent flow in the area around a retinal hole and prevent redetachment of the retina as well as cause a secondary hemorrhage into the vitreous cavity. 27  
HA in its native form is rapidly degraded in the human eye. Cross-linking has been successfully used to markedly slow degradation. 38 The in vitro degradation rate in our study was only 10% during the first 4 weeks and slowed even more during the following months. The hydrogels remained stable without losing their characteristics. However, one has to keep in mind that the environment in the human vitreous is different from the in vitro situation in these experiments. Hong et al. 10 showed that real degradation may be faster. Nevertheless, the UV-CHA biogels remained stable in the vitreous cavity of rabbits over 6 weeks. Stability and persistence of a vitreous substitute within the vitreous cavity over a long period is of great importance, especially in the treatment of retinal detachments with pronounced proliferative vitreoretinopathy. A past experience with MIRAgel (MIRA, Inc., Waltham, MA) showed that unexpected complications can occur in vivo. 39  
UV-CHA seems to have good stability and transparency in the vitreous cavity for at least 6 weeks. The injection of UV-CHA alone (without the use of endolaser or heavy fluids) seemed to have successfully treated three of four cases of iatrogenic retinal detachments in our animal experiments. 
Good biocompatibility of the UV-CHA hydrogels was demonstrated in vitro and in vivo. Hydrophilic biopolymers such as UV-CHA hydrogels may be a valuable alternative for long-term vitreous replacement. Moreover, a drug may be added to the hydrogels during the cross-linking process, providing clinicians with a slow-release system. This approach may open new options for the prevention and treatment of retinal detachment. 
Footnotes
 Disclosure: C. Schramm, None; M.S. Spitzer, None; S. Henke-Fahle, None; G. Steinmetz, None; K. Januschowski, None; P. Heiduschka, None; J. Geis-Gerstorfer, None; T. Biedermann, None; K.U. Bartz-Schmidt, None; P. Szurman, None
References
Foster WJ . Vitreous substitutes. Expert Rev Ophthalmol. 2008;3(2):211–218. [CrossRef] [PubMed]
Spitzer MS Kaczmarek RT Yoeruek E . The distribution, release kinetics, and biocompatibility of triamcinolone injected and dispersed in silicone oil. Invest Ophthalmol Vis Sci. 2009;50(5):2337–2343. [CrossRef] [PubMed]
Lewis H Aaberg TM Abrams GW . Cause of failure after initial vitreoretinal surgery for severe proliferative vitreoretinopathy. AMJ Ophthalmol. 1991;111:8–14. [CrossRef]
Mackiewics J Mühling B Hiebl W . In vivo retinal tolerances of various heavy silicone oils. Invest Ophthalmol Vis Sci. 2007;48(4):1873–1883. [CrossRef] [PubMed]
Nakagawa M Tanaka M Miyata T . Evaluation of collagen gel and hyaluronic acid as vitreous substitutes. Ophthalmic Res. 1997;29:409–420. [CrossRef] [PubMed]
Pruett RC Schepens CL Swann DA . Hyaluronic acid vitreous substitute; a six-year clinical evaluation. Arch Ophthalmol. 1979;97(12):2325–2330. [CrossRef] [PubMed]
Swindle KE Ravi N . Recent advances in polymeric vitreous substitutes. Expert Rev Ophthalmol. 2007;2:255–265. [CrossRef]
Chirila TV Tahija S Hong Y . Synthetic polymers as materials for artificial vitreous body: review and recent advances. J Biomater Appl. 1994;9(2):121–137. [PubMed]
Hong Y Chirila TV Vijayasekaran S . Crosslinked poly(1-vinyl-2-pyrrolidinone) as a vitreous substitute. J Biomed Mater Res. 1996;30(4):441–448. [CrossRef] [PubMed]
Hong Y Chirila TV Vijayasekaran S . Biodegradation in vitro and retention in the rabbit eye of crosslinked poly(1-vinyl-2-pyrrolidinone) hydrogel as a vitreous substitute. J Biomed Mater Res. 1998;39(4):650–659. [CrossRef] [PubMed]
Soman N Banerjee R . Artificial vitreous replacements. Biomed Mater Eng. 2003;13(1):59–74. [PubMed]
Balazs EA . Fine structure and function of ocular tissues. the vitreous. Int Ophthalmol Clin. 1974;13(3):169–187.
Eisner G . Zur Anatomie des Glaskörpers. Graefes Arch. 1975;193:33–56. [CrossRef]
Hascall VC . Hyaluronan, a common thread. Glycoconj J. 2000;17 (7–9):607–616. [CrossRef] [PubMed]
Kakehi K Kinoshita M Yasueda S . Hyaluronic acid: separation and biological implications. J Chromatogr B. 2003;797:347–355. [CrossRef]
Prestwich GD Marecak DM Marecek JF . Controlled chemical modification of hyaluronic acid: synthesis, applications, and biodegradation of hydrazide derivatives. J Control Release. 1998;53(1–3):93–103. [CrossRef] [PubMed]
Vercruysse KP Marecak DM Marecek JF Prestwich GD . Synthesis and in vitro degradation of new polyvalent hydrazide cross-linked hydrogels of hyaluronic acid. Biokonjugate Chem. 1997;8:686–694. [CrossRef]
Leach JB Bivens KA Patrick CWJr Schmidt CE . Photocrosslinked hyaluronic acid hydrogels: natural, biodegradable tissue engineering scaffolds. Biotechnol Bioeng. 2003;82(5):578–589. [CrossRef] [PubMed]
Leach JB Bivens KA Collins CN Schmidt CE . Development of photocrosslinkable hyaluronic acid-polyethylene glycol-peptide composite hydrogels for soft tissue engineering. J Biomed Mater Res A. 2004;70(1):74–82. [CrossRef] [PubMed]
Bitter T Muir HM . A modified uronic acid carbazole reaction, Anal Biochem. 1962;4:330–334. [CrossRef] [PubMed]
Lee B Litt M Buchsbaum G . Rheology of the vitreous body, Part 1: viscoelasticity of human vitreous. Biorheology. 1992;29:521–533. [PubMed]
Lee B Litt M Buchsbaum G . Rheology of the vitreous body, Part 2: viscoelasticity of bovine and porcine vitreous. Biorheology. 1994;31:327–338. [PubMed]
Foster WJ Aliyar HA Hamilton P . Internal osmotic pressure as a mechanism of retinal attachment in a vitreous substitute. J Bioact Compat Polym. 2006;21:221–235. [CrossRef]
Silver FH LiBrizzi J Benedetto D . Use of viscoelastic solutions in ophthalmology: a review of physical properties and long-term effects. J Long Term Eff Med Implants. 1992;2(1):49–66. [PubMed]
Theelen T Tilanus MAD Klevering BJ . Intraocular inflammation following endotamponade with high-density silicone oil. Graefes Arch Clin Exp Ophthalmol. 2004;242:617–620. [CrossRef] [PubMed]
Kirchhof B Tavakolian U Paulmann H . Histopathological findings in eyes after silicon oil injection. Graefes Arch Clin Exp Ophthalmol. 1986;224:34–37. [CrossRef] [PubMed]
Szurman P Frank C Kaczmarek RT . Vitreous substitutes as drug release systems (in German). Klin Monbl Augenheilkd. 2009;226:718–724. [CrossRef] [PubMed]
Cho KY Chung TW Kim BC . Release of ciprofloxacin from poloxamer-graft-hyaluronic acid hydrogels in vitro. Int J Pharm. 2003;260(1):83–91. [CrossRef] [PubMed]
Kelly RM Meyer JD Matsuura JE . In vitro release kinetics of gentamycin from a sodium hyaluronate gel delivery system suitable for the treatment of peripheral vestibular disease. Drug Dev Ind Pharm. 1999;25(1):15–20. [CrossRef] [PubMed]
Liesegang TJ . Viscoelastic substances in ophthalmology. Surv Ophthalmol. 1990;34(4):268–293. [CrossRef] [PubMed]
Koster R Stilma JS . Healon as intravitreal substitute in retinal detachment surgery in 40 patients. Doc Ophthalmol. 1986;64(1):13–17. [CrossRef] [PubMed]
Su WY Chen KH Chen YC . An injectable oxidated hyaluronic acid/adipic acid dihydrazide hydrogel as a vitreous substitute. J Biomater Sci Polym Ed. 2011;22:1777–1797. [CrossRef] [PubMed]
Nguyen KT West JL . Photopolymerizable hydrogels for tissue engineering applications. Biomaterials. 2002;23;4307–4314. [CrossRef] [PubMed]
Swindle-Reilly KE Shah M Hamilton PD . Rabbit study of an in situ formin hydrogel vitreous substitute. Invest Ophthalmol Vis Sci. 2009;50:4840–4846. [CrossRef] [PubMed]
Hotta K Sugitani A . Refractive changes in silicone oil-filled pseudophakic eyes. Retina. 2005;25(2):167–170. [CrossRef] [PubMed]
D'Errico G De Lellis M Mangiapia G . Structural and mechanical properties of UV-photo-crosslinked poly(n-vinyl-2-pyrrolidone) hydrogels. Biomacromolecules. 2008;9:231–240. [CrossRef] [PubMed]
Pritchard CD Crafoord S Andréasson S . Evaluation of viscoelastic poly(ethylene glycol) sols as vitreous substitutes in an experimental vitrectomy model in rabbits. Acta Biomater. 2011;7:936–943. [CrossRef] [PubMed]
Segura T Anderson BC Chung PH . Crosslinked hyaluronic acid hydrogels: a strategy to functionalize and pattern. Biomaterials. 2005;26(4):359–371. [CrossRef] [PubMed]
Roldán-Pallarés M Hernández-Montero J Llanes F . MIRAgel: hydrolytic degradation and long-term observations. Arch Ophthalmol. 2007;125:511–514. [CrossRef] [PubMed]
Figure 1.
 
With both methods, either cross-linking HA with adipic dihydrazides (a) or cross-linking HA with N-vinyl-pyrrolidinone and photopolymerization (b), optical transparent, rubber-like and three-dimensional hydrogels were produced.
Figure 1.
 
With both methods, either cross-linking HA with adipic dihydrazides (a) or cross-linking HA with N-vinyl-pyrrolidinone and photopolymerization (b), optical transparent, rubber-like and three-dimensional hydrogels were produced.
Figure 2.
 
HA-degradation from the ADH-HA (a) and the UV-HA hydrogels (b). The degradation was separated into two phases. The first showed a faster degradation from 7.487 mg HA from 80 mg cross-linked ADH-HA (a) in 30 days and 5.634 mg HA from 50 mg cross-linked UV-HA (b) in 15 days. Thereafter, in both cases, the rate of HA-degradation slowed down.
Figure 2.
 
HA-degradation from the ADH-HA (a) and the UV-HA hydrogels (b). The degradation was separated into two phases. The first showed a faster degradation from 7.487 mg HA from 80 mg cross-linked ADH-HA (a) in 30 days and 5.634 mg HA from 50 mg cross-linked UV-HA (b) in 15 days. Thereafter, in both cases, the rate of HA-degradation slowed down.
Figure 3.
 
After an incubation period of 24 hours with the CHA hydrogels, the cell viability of the RPE cells was assessed by staining the live and dead cells with calcein AM (b, live cells) and EthD-1 (c, dead cells). To exemplify the RPE cell structure, images were obtained with reflected-light microscopy or after staining the nucleus with 4′6′-diamidino-2-phenylindole (DAPI) (a). Compared with controls of live (1) and dead (2) cells, an intense red fluorescence as a signal for dead cells in the group under incubation with the ADH-HA hydrogels was initially observed (3). However, after dialysis over 24 to 48 hours, the RPE cells incubated with UV-HA hydrogels (4) showed no increased signs of cytotoxicity compared with the control. However, RPE cells incubated with ADH-HA hydrogels (5), showed an increased number of red fluorescent dead cells and the cell groups appeared to be disrupted in some areas even after dialyzing the gels for 24 to 48 hours.
Figure 3.
 
After an incubation period of 24 hours with the CHA hydrogels, the cell viability of the RPE cells was assessed by staining the live and dead cells with calcein AM (b, live cells) and EthD-1 (c, dead cells). To exemplify the RPE cell structure, images were obtained with reflected-light microscopy or after staining the nucleus with 4′6′-diamidino-2-phenylindole (DAPI) (a). Compared with controls of live (1) and dead (2) cells, an intense red fluorescence as a signal for dead cells in the group under incubation with the ADH-HA hydrogels was initially observed (3). However, after dialysis over 24 to 48 hours, the RPE cells incubated with UV-HA hydrogels (4) showed no increased signs of cytotoxicity compared with the control. However, RPE cells incubated with ADH-HA hydrogels (5), showed an increased number of red fluorescent dead cells and the cell groups appeared to be disrupted in some areas even after dialyzing the gels for 24 to 48 hours.
Figure 4.
 
MTT test: RPE cells after incubation for 48 hours with CHA hydrogels. No significant differences were found between the mitochondrial activity of the RPE cells incubated with UV-HA or ADH-HA hydrogels compared with the controls (P > 0.05).
Figure 4.
 
MTT test: RPE cells after incubation for 48 hours with CHA hydrogels. No significant differences were found between the mitochondrial activity of the RPE cells incubated with UV-HA or ADH-HA hydrogels compared with the controls (P > 0.05).
Figure 5.
 
To quantify the cytotoxicity of the CHA hydrogels on RPE cells and their effects on cell proliferation, RPE cells were incubated for 72 hours with the CHA hydrogels and Alamar blue. The mitochondrial activity of RPE cells incubated with the UV-HA hydrogels (a) was not significantly different from that in the control group (P > 0.05). The mitochondrial activity of RPE cells incubated with the ADH-HA hydrogels (b) was significantly higher than that in the control group (*P < 0.05).
Figure 5.
 
To quantify the cytotoxicity of the CHA hydrogels on RPE cells and their effects on cell proliferation, RPE cells were incubated for 72 hours with the CHA hydrogels and Alamar blue. The mitochondrial activity of RPE cells incubated with the UV-HA hydrogels (a) was not significantly different from that in the control group (P > 0.05). The mitochondrial activity of RPE cells incubated with the ADH-HA hydrogels (b) was significantly higher than that in the control group (*P < 0.05).
Figure 6.
 
Representative Annexin V and PI dot plots for the control groups with live (a) and dead (b) cells and UV-HA (c) and ADH-HA (d) hydrogels. (Q1, Q2) Dead and dying cells (PI positive). (Q3) Living cells without signs of apoptosis (Annexin V and PI negative). (Q4) Early stages of apoptosis (Annexin V positive and PI negative).
Figure 6.
 
Representative Annexin V and PI dot plots for the control groups with live (a) and dead (b) cells and UV-HA (c) and ADH-HA (d) hydrogels. (Q1, Q2) Dead and dying cells (PI positive). (Q3) Living cells without signs of apoptosis (Annexin V and PI negative). (Q4) Early stages of apoptosis (Annexin V positive and PI negative).
Figure 7.
 
The dynamic viscosity was measured in reference to the shear rate. UV-HA was measured in a whole piece and injected through needles with a diameter of 0.6, 0.7, or 0.9 mm.
Figure 7.
 
The dynamic viscosity was measured in reference to the shear rate. UV-HA was measured in a whole piece and injected through needles with a diameter of 0.6, 0.7, or 0.9 mm.
Figure 8.
 
ERG responses in the hydrogel vitreous substitute group (right eye) in comparison to the control (left eye). Top: bar graphs represent the ratio of the B-wave amplitudes between the operated eye (UV-CHA filled) and the nonoperated eye. Bottom: the ratio of the B-wave latencies between the operated eye (UV-CHA filled) and the nonoperated eye. There was no statistically significant difference between the surgery and the control groups. Data the mean ± SD; P > 0.05.
Figure 8.
 
ERG responses in the hydrogel vitreous substitute group (right eye) in comparison to the control (left eye). Top: bar graphs represent the ratio of the B-wave amplitudes between the operated eye (UV-CHA filled) and the nonoperated eye. Bottom: the ratio of the B-wave latencies between the operated eye (UV-CHA filled) and the nonoperated eye. There was no statistically significant difference between the surgery and the control groups. Data the mean ± SD; P > 0.05.
Figure 9.
 
Macroscopic and histologic findings 6 weeks after implantation of UV-CHA. (A) The vitreous cavity of the operated eyes was still completely filled with the hydrogels. The transparency and viscoelasticity of the injected UV-CHA was grossly unchanged since injection 6 weeks ago. The lens was clear. (B) Histology revealed unaltered retinal structure and cell morphology (H&E; ×200).
Figure 9.
 
Macroscopic and histologic findings 6 weeks after implantation of UV-CHA. (A) The vitreous cavity of the operated eyes was still completely filled with the hydrogels. The transparency and viscoelasticity of the injected UV-CHA was grossly unchanged since injection 6 weeks ago. The lens was clear. (B) Histology revealed unaltered retinal structure and cell morphology (H&E; ×200).
Table 1.
 
Factors from the Regression Formula of HA Degradation from Cross-linked Hydrogels
Table 1.
 
Factors from the Regression Formula of HA Degradation from Cross-linked Hydrogels
ADH-HA UV-HA
c (t = 0) 80000 80000
k1 4845.476667 4974.295
t (50,1) 7.097 1.4288
n1 0.892 0.98665
k2 5952.04 1186.15
t (50,2) 27.39 11.63
n2 3.798333333 10.522
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×