www.elsevier.com/locate/toxinvit

Toxicology in Vitro 19 (2005) 957–962

Hydrophilic polymers—biocompatibility testing in vitro

K. Kejlova a,*, J. Labsky b, D. Jırova a, H. Bendova a

a National Reference Center for Cosmetics, National Institute of Public Health, Srobarova 48, 100 42 Prague 10, Czech Republicb Institute of Macromolecular Chemistry, Academy of Sciences of the Czech Republic, 162 06 Prague 6, Czech Republic

Available online 2 August 2005

Abstract

Biocompatibility is one of the main prerequisites for safe use of medical devices. Estimation of cytotoxicity is a part of the initialevaluation laid down in ISO standards on biological evaluation of medical devices. Hydrophilic polymers (based on 2-hydroxyethylmethacrylate HEMA) doped by addition of selected additives with antioxidant and/or free radical scavenging potential (vitamin Cand hindered amine stabilizer N-(2,2,6,6-tetramethylpiperidin-4-yl)methacrylamide) were tested in different in vitro systems (3T3Balb/c cell culture and a 3D human skin model) for biocompatibility and suitability for use as wound dressings. The results ofthe 3T3 NRU cytotoxicity test using both the direct and indirect contact approaches and a 3D skin model modified irritation test(EpiDermTM) confirmed high biocompatibility and good skin tolerance of both the basic polymers and those enriched with specificadditives up to a balanced level. HEMA polymer showed a beneficial effect against cytotoxicity of an irritant (sodium dodecyl sul-fate). The in vitro biocompatibility test results were confirmed by human local skin tolerance testing.� 2005 Elsevier Ltd. All rights reserved.

Keywords: Biocompatibility; Hydrophilic polymer; 3T3 NRU cytotoxicity test; 3D human skin model; Human patch test

1. Introduction

Biomaterials are developed in order to evaluate, treat,augment or replace human tissue, organ or function(Groth et al., 1995). Biocompatibility is the main prere-quisite for their safe use as medical devices. Local skincompatibility is one of the crucial parameters for theirpossible clinical application. A comprehensive, generalguideline on the testing of biocompatibility of materialsfor medical applications is specified in the InternationalStandard ISO 10993 (International Organization forStandardization, 1999), with priority being given to cell

0887-2333/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.doi:10.1016/j.tiv.2005.06.032

Abbreviations: HEMA, 2-hydroxyethyl methacrylate; HAS, hin-dered amine stabilizer; SDS, sodium dodecyl sulfate; 3T3 NRU, 3T3neutral red uptake.* Corresponding author. Tel.: +420 267082327; fax: +420

267082386.E-mail address: kejlova@szu.cz (K. Kejlova).

culture-based in vitro tests using both the direct andindirect contact approaches.Hydrophilic polymers are promising biomaterials in-

tended to replace conventional products in wound dress-ing systems. New films, foams and gels capable toregulate skin humectation and softness may promoteand accelerate wound healing (Labsky et al., 2003). Thispotential can be enhanced by addition of antioxidantssuch as vitamin C or hindered amine stabilizer (HAS),which reacts selectively with hydroxyl radical, hydrogenperoxide and other reactive oxygen species to create sta-ble and non-toxic nitroxide radical (Macleay and Lange,1994).A new generation of saccharide modified hydrophilic

polymers, based on 2-hydroxyethyl methacrylate(HEMA), enriched with antioxidants, were tested in dif-ferent in vitro systems for biocompatibility and benefi-cial potential. Human patch testing was employed toconfirm the in vitro biocompatibility results.

Copolymer HEMA-HAS

CH2

CH3

CO

O

CH2CH2OH

O CH2CH2O CO CH3

CH2

CH3

CO

CH2

CH3

CO

NH

N

H

HEMA polymer

CH2

CH3

CO

O

CH2CH2OH

O

CH2CH2O CO CH3

CH2

CH3

CO

CH2

CH3

CO

O

CH2CH2OH

HAS monomerN-(2,2,6,6-tetramethylpiperidin-4-yl)methacrylamide

CH2 CH3

OCNH

N

H

Fig. 1. Chemical structure of HEMA polymers: Copolymer HEMA–HAS, HEMA polymer, HAS monomer and N-(2,2,6,6-tetramethyl-piperidin-4-yl)methacrylamide.

958 K. Kejlova et al. / Toxicology in Vitro 19 (2005) 957–962

2. Materials and methods

Test samples were prepared by the Institute of Mac-romolecular Chemistry, Prague, in the form of filmsor gels, and sterilized by autoclaving (120 �C, 20 min).Considering the foreseen clinical application, the filmform was tested using cell culture, while the gelform convenient for epicutaneous application wastested in a 3D human skin model and on human vol-unteers.

2.1. Preparation of films and gels

Chemicals: HEMA (2-hydroxyethyl methacrylate),ethylene glycol dimethacrylate, 2,2 0-azobis(2-methylpropanenitrile), ascorbic acid, all Fluka products.Darocur 1173 (2-hydroxy-2-methyl-1-phenylpropan-

1-one), CIBA.

2.1.1. HEMA film

A mixture of 50 ml of HEMA (ethylene glycoldimethacrylate content 0.11%), 0.3 ml of ethylene glycoldimethacrylate and 0.5 ml Darocur 1173 was degassedwith a stream of argon for 10 min and transferred to atransparent mold (21 · 30 cm) enabling the preparationof a film (thickness 1.5 mm). The mold was irradiatedwith 6 UV lamps Philips (20 W each) for 10 min at a dis-tance of 20 cm. The film was exposed three times to 30%ethanol (24 h) to remove monomeric starting material.Discs of 21 mm in diameter were cut from the wet film,equilibrated in distilled water and dried.

2.1.2. HEMA–HAS film

HEMA–HAS copolymers were prepared using thesame procedure as for HEMA polymer; HAS monomerN-(2,2,6,6-tetramethylpiperidin-4-yl)methacrylamide(Macleay and Lange, 1994; Chmela and Hrdlovic, 1985)was added at concentrations of 3% and 6% (w/w). Thepurification of prepared films was done following thesame procedure as for pure HEMA polymer.

2.1.3. HEMA–HAS gels (3% gel)

A mixture of 80 g HEMA, 0.6 g ethylene glycoldimethacrylate, 2.4 g HAS monomer, and 0.5 g 2,2 0-azo-bis(2-methyl propanenitrile) was heated in 1000 ml ofbenzene at 70 �C for 12 h. The obtained copolymer(powder) was extracted three times with boiling benzene(1000 ml total volume) and dried. A gel was prepared bymixing dry HEMA–HAS copolymer with 50% aqueoussolution of polyethylene glycol 300 (BASF) to obtain agel of suitable consistency. The HEMA/HAS 6% andHEMA/vitamin C gels were prepared by the same pro-cedure, adding either 6% (w/w) of HAS or 1% (w/w) ofascorbic acid. The chemical structures of HEMA poly-mer, HAS monomer and HEMA–HAS copolymer areshown in Fig. 1.

2.2. In vitro tests

The 3T3 NRU cytotoxicity test was performedaccording to ISO 10993 (Part 5—tests for in vitro cyto-toxicity) and standard INVITTOX Protocol No. 64(http://ecvam-sis.jrc.it/invittox/static/index.html).For HEMA films, the direct contact test was used:

Balb/c 3T3 fibroblasts (L1, ECACC No. 86052701,seeding density 1 · 105 per well) were precultured for18 h in Dulbecco�s modified essential medium supple-mented with bovine serum (10%) in 24-well plates andexposed for 24 h to the films placed in the center ofeach well, covering 1/10 of the cell layer surface. Themorphological changes indicating cytotoxicity andcell growth characteristics were recorded using an Olym-pus IX50 microscope and attached Olympus C-5050camera.

K. Kejlova et al. / Toxicology in Vitro 19 (2005) 957–962 959

For HEMA gels, tests on extracts were performed.The extracts were prepared according to ISO 10993 Part12 (International Organization for Standardization,1996) at a ratio of 0.1 g of the test material to 1 ml ofextraction vehicle (DMEM without serum), extractiontime: 24 h, extraction temperature: 37 �C. Balb/c 3T3fibroblasts (seeding density 1 · 104 per well) were precul-tured for 18 h in 96-well plates and exposed in quadru-plicates to the tested gel extracts (undiluted, 50% and25% diluted in DMEM) for 24 h. After the treatmentthe cells were exposed to neutral red solution and theneutral red uptake was determined fluorimetrically(Denley-Wellfluor reader, UK). The mean fluorescenceof the negative control (DMEM) culture was set to rep-resent 100% viability and the viability of the treated cellculture was calculated.

2.3. 3D human skin model—EpiDermTM (MatTek, USA)

The test procedure used was a modified EpiDermSkin Corrosivity Test (INVITTOX Protocol No. 119).

2.3.1. Biocompatibility test of HEMA gels

Before starting the experiment, the tissues were prein-cubated for 1 h (37 �C, 5% CO2) and prior to dosing, theassay medium was renewed. Fifty microliter of the testedgels were applied atop the tissues in triplicates and incu-bated for 24 h. At the end of the exposure period, the tis-sue cultures were gently rinsed with phosphate-bufferedsaline (PBS), incubated in fresh medium for further 24 h,and finally placed into plates containing MTT solution.Following a 3 h incubation and a rinse with PBS, the re-duced MTT was extracted by submerging the tissues in2 ml of isopropanol and shaking for 2 h. The absor-bance (optical density; OD) of the extraction solutionswas measured with a Varian Cary 1E spectrophotometerat 540 nm. The mean OD and standard deviation werecalculated for each tested sample. The mean OD ofthe negative control (H2O) tissues was set to represent100% viability.

2.3.2. Protective effects of HEMA gels (HEMA,

HEMA + 3% HAS, HEMA + vitamin C) against sodium

dodecyl sulfate (SDS)

Before starting the experiment, the tissues were prein-cubated for 1 h (37 �C, 5% CO2) and prior to dosing, theassay medium was renewed. Twenty-five microliter ofSDS (SIGMA) diluted to 0.5% in distilled water weredispensed atop the tissues in triplicates and incubatedfor 15 min. At the end of the exposure period, the tissuecultures were gently rinsed with phosphate-buffered sal-ine (PBS) to eliminate any residual SDS and 50 ll of thetested gels were applied atop the tissues. The tissues wereincubated in fresh medium for further 42 h, rinsed withPBS, and finally placed into plates containing MTTsolution. Following a 3 h incubation and a rinse with

PBS, the reduced MTT was extracted by submergingthe tissues in 2 ml of isopropanol and shaking for 2 h.The absorbance (optical density; OD) of the extractionsolutions was measured with a Varian Cary 1E spectro-photometer at 540 nm. The mean OD and standarddeviation were calculated for each tested sample. Themean OD of the negative control (H2O) tissues was setto represent 100% viability.

2.4. In vivo test

2.4.1. Patch test in human volunteers (COLIPA, 1995)

The tested gels (50 ll) were applied in semiocclusion(Curatest, Lohman/Rauscher) on the back of 15 healthyvolunteers (13 females and 2 males, aged 26–62 years),the exposure time was 4 h. Visual scoring of skin reac-tions was performed 1 h, 24 h and 48 h after patch re-moval, with skin erythema graded on a four-point scale(0 no erythema, 1 very slight erythema (barely percepti-ble), 2 well defined erythema, 3 moderate erythema, 4 se-vere erythema). The selection of volunteers and the testmethods were carried out in accordance with the ethicalprinciples as set out in the Declaration of Helsinki andInternational Ethical Guidelines for Biomedical Re-search Involving Human Subjects (CIOMS, 1993). Thestudy was approved by the Ethical Review Committeeof the National Institute of Public Health, Prague.

3. Results

3.1. Direct contact cytotoxicity

Good biocompatibility, i.e. no cytotoxicity, was provedfor the films prepared from HEMA alone (Fig. 2b) andHEMA enriched with 3% HAS (Fig. 2c), when com-pared with intact, control cell culture (Fig. 2a). HEMAfilm enriched with 6% HAS appeared moderately cyto-toxic, causing changes in the cell morphology anddecreasing the density of viable cells (Fig. 2d).

3.2. Indirect contact cytotoxicity tested on extracts

Cytotoxicity results for extracts of HEMA gel,HEMA + vitamin C gel, HEMA + 3% HAS gel andHEMA + 6% HAS gel are shown in Fig. 3. The bestbiocompatibility was proved for HEMA + HAS 3%gel; the cell viability is high at 25% and 50% extract con-centrations, with a slight decrease detected at 100%. Theextracts of pure HEMA and HEMA enriched with 6%of HAS were only tolerated by the cell culture at 25% ex-tract concentration. The extract of HEMA + vitamin Cgel was tolerated by the cell culture at 50% and lowerconcentrations, with a slight elevation within the usualviability range at 25% concentration.

Fig. 2. Biocompatibilty of HEMA films in the 3T3 Balb/c cell culture (direct contact), no staining, magnification 100·. (a) Control cell culture,(b) HEMA film, (c) HEMA + 3% HAS film and (d) HEMA + 6% HAS film.

960 K. Kejlova et al. / Toxicology in Vitro 19 (2005) 957–962

3.3. 3D human skin model

All of the HEMA gels exhibited good biocompatibil-ity, with the tissue viability not falling below 80% of thenegative control treated with water. However, the lowestviability of EpiDerm tissues was detected forHEMA + 6% HAS (Fig. 4).

3.4. Patch test in human volunteers

Local skin tolerance was recorded for all test samplesin the form of gels. No erythema, oedema or other ad-verse effects were observed.

3.5. Protective effect against sodium dodecyl sulfate

(SDS) in the 3D human skin model

The possible protective effect of gels with the best bio-compatibility parameters was tested against SDS-in-duced tissue toxicity (Fig. 5). The toxic effect elicitedby SDS (0.5%) was reduced by subsequent applicationof pure HEMA gel and HEMA + 3% HAS gel. The via-

bility of skin tissues rose from 42% to 53% and 63%,respectively. No protective effect was found forHEMA + vitamin C.

4. Discussion

Hydrophilic HEMA polymers exhibit high biocom-patibility when tested in vitro on skin-derived tissues.This conclusion supports the long-term experience withthe use of different types of polymers ranging from softhydrogels for contact and intraocular lenses to highstrength polymers for orthopedic materials (Lim andWichterle, 1965; Horak, 1992). Hydrophilic polymershave a wide variety of applications including wounddressings (Burrow and Schmidt, 1995).The addition of antioxidants may even have further

beneficial effect (Samuni et al., 2002). However, wefound HEMA polymers enriched with 6% comparedto 3% HAS to exhibit higher toxicity which may beattributed to the very strong alkaline nature of hinderedamine stabilizers. The cytotoxicity test revealed a certain

0

20

40

60

80

100

120

Control SDS 0.5% SDS + HEMA SDS +HEMA/HAS

3%

SDS +HEMA/vit. C

MT

T v

iab

ility

(%

of

neg

ativ

e co

ntr

ol)

Fig. 5. Protective effects of HEMA gels against sodium dodecyl sulfate(SDS) in the 3D human skin model (EpiDermTM). Each columnrepresents the viability of 6 tissues ± SD (data from two separateexperiments).

0

20

40

60

80

100

120

140

0 0.25 0.5 0.75 1

concentration of extract

viab

ility

(%

of

con

tro

l)

HEMA

HEMA + 3% HAS

HEMA + 6% HAS

HEMA + vit.C

Fig. 3. Cytotoxicity of HEMA gel extracts in the 3T3 NRUcytotoxicity test. Each point represents the mean viability for 12 wells± SD (data from three separate experiments).

K. Kejlova et al. / Toxicology in Vitro 19 (2005) 957–962 961

degree of cytotoxicity for all but one of HEMA gel ex-tracts: HEMA + 3% HAS. Tests on the 3D skin modelEpiDermTM proved good compatibility of all HEMAgels, confirmed by in vivo patch testing in a group ofhuman volunteers. All of the tested HEMA gels werefound convenient for application on human skin, withno adverse skin effects recorded.The beneficial effect of some HEMA polymer gels

against toxicity elicited by an irritant (sodium dodecylsulfate) was demonstrated, namely for HEMA andHEMA + 3% HAS. The more pronounced protectiveeffect of HEMA enriched with 3% HAS is ascribablenot only to the known ability of HEMA polymers to re-tain considerable amounts of water, and thus to act asskin softeners and humectants (Mathur et al., 1996),but also to free radical scavenging potential of HAS,which became the basis of several patents (Labskyet al., 2003). For clinical purposes, the addition of upto 3% of HAS is recommended.The beneficial effect of another antioxidant, vitamin

C, was not demonstrated. Its antioxidant and freeradical scavenging potentials may have been impaired

0

20

40

60

80

100

120

control HEMA HEMA +3% HAS

HEMA +6% HAS

HEMA +vit. C

MT

T v

iab

ility

(%

of

neg

ativ

e co

ntr

ol)

Fig. 4. Biocompatibility of HEMA gels in the 3D human skin model(EpiDermTM). Each column represents the viability of 6 tissues ± SD(data from two separate experiments).

during the preparation for cell culturing. Moreover,vitamin C that shows problematic solubility in the mix-ture with HEMA gel was used at a relatively lowconcentration.Good biocompatibility, proved both in vitro and

in vivo, and the beneficial effect revealed in the 3D skinmodel system qualify the HEMA polymers as promisingbiomaterials for clinical use.

Acknowledgement

The study was supported by grant project no. IAA4050301 of the Academy of Sciences of the CzechRepublic.

References

Burrow, T.R., Schmidt, R., 1995. Wound dressings. GB Patent2288736.

Chmela, S., Hrdlovic, P., 1985. The inhibited photo-oxidation ofpolystyrene and styrene/HALS copolymers. Polymer Degradationand Stability 11, 339–348.

CIOMS, 1993. International Ethical Guidelines for BiomedicalResearch Involving Human Subjects. Council for InternationalOrganizations of Medical Sciences, Geneva.

COLIPA, 1995. The European Cosmetic Toiletry and PerfumeryAssociation Tests for Human Skin Compatibility. CosmeticProduct Test Guidelines for Assessment of Human Skin Compat-ibility, COLIPA, Brussels.

Groth, T., Falck, P., Miethke, R.R., 1995. Cytotoxicity of biomate-rials—basic mechanisms and in vitro test methods: A review.Alternatives to Laboratory Animals 23, 790–799.

Horak, D., 1992. The use of poly(2-hydroxyethyl methacrylate) inmedicine. Chemicke Listy 86 (9), 681–691.

<http://ecvam-sis.jrc.it/invittox/static/index.html>.International Organization for Standardization, 1996. Biological

evaluation of medical devices—Part 12: sample preparation andreference materials (ISO 10993-12:1996). International Organiza-tion for Standardization, Geneva.

962 K. Kejlova et al. / Toxicology in Vitro 19 (2005) 957–962

International Organization for Standardization, 1999. Biologicalevaluation of medical devices—Part 5: tests for in vitro cytotoxicity(ISO 10993-5:1999). International Organization for Standardiza-tion, Geneva.

Labsky, J., Vacık, J., Hosek, P., 2003. A preparation for preventionand healing of inflammation affections. US Patent 6610284; WO99/28359; PCT/CZ 98/00046.

Lim, D., Wichterle, O., 1965. Cross-linked hydrophilic polymers andarticles made therefrom. US Patent 3220960.

Macleay, R.E., Lange, H.C., 1994. Derivatives of N-HALS-substitutedamic acid hydrazides. US Patent 5338853.

Mathur, A.M., Moorjani, S.H., Scranton, A.C., 1996. Methods forsynthesis of hydrogel networks: A review. Journal of Macromo-lecular Science—Review. Macromolecular Chemistry and Physics,405–430.

Samuni, A.M., DeGraff, W., Krishna, M.C., Mitchell, J.B., 2002.Nitroxides as antioxidants: Tempol protects against EO9 cyto-toxicity. Molecular and Cellular Biochemistry 234/235, 327–333.