Biomaterials 23 (2002) 4503–4513
`
`Hyaluronic acid hydrogel in the treatment of osteoarthritis
`R. Barbuccia,*, S. Lamponia, A. Borzacchiellob, L. Ambrosiob, M. Finic,d,
`P. Torricellic,d, R. Giardinoc
`a Department of Chemical and Biosystem Sciences and Technology, CRISMA, University of Siena, Via Aldo Moro n. 2, 53100 Siena, Italy
`b Interdisciplinary Research Center in Biomaterials (CRIB), Institute of Composite Materials Technology-CNR, University of Naples,
`‘‘Federico II’’, P.le Tecchio 80, 80125 Naples, Italy
`c Experimental Surgery Department, Research Institute Codivilla-Putti, Italy
`d Rizzoli Orthopaedic Institute, Via di Barbiano 1/10, 40136 Bologna, Italy
`
`Received 7 December 2001; accepted 16 May 2002
`
`Abstract
`
`In order to overcome the problem of rapid clearance of the polysaccharide hyaluronic acid (Hyal) in the treatment of
`osteoarthritis (OA), a 50% cross-linked Hyal hydrogel (Hyal 50%) was synthesised. The 50% refers to the amount of COOH groups
`of the polysaccharide involved in the cross-linking reaction, i.e. 50% of the total amount. The rheological behaviour of the Hyal
`50% hydrogel, and in particular the possibility to inject it through a needle, was studied. The results obtained demonstrated that the
`hydrogel injected through the needle still behaved like a gel, although it showed a reduction of the dynamic moduli. The most
`appropriate sterilisation technique for this kind of hydrogel was also evaluated. Liophilised Hyal 50% samples were sterilised by
`steam, Ethylene Oxide (EtO) and g-rays. EtO and g-rays did not modify the characteristics of the hydrogel in terms of swellability
`and morphology. Lastly, the in vivo effect of Hyal 50% hydrogel in the treatment of chondral defect in rabbit knee was also studied.
`The results obtained showed the Hyal 50% injections improved chondrocytes density and matrix appearance. Furthermore, the
`permanence in situ of the hydrogel was longer than that of the linear Hyal. r 2002 Published by Elsevier Science Ltd.
`
`Keywords: Hyaluronan hydrogel; Osteoarthritis; Sterilisation; Rheological properties; In vivo tests
`
`1. Introduction
`
`Osteoarthritis (OA), also called degenerative joint
`disease or iperthrophic-arthritis, is clinically characterised
`by joint pain and joint-instability. By the age of 40, about
`90% of humans show some degenerative changes in
`weight-bearing joints [1]. OA is frequently found in hip,
`knee, distal phalangeal and intervertebral joints. OA is
`termed as primary or secondary, and the end-stage of this
`disease is characterised by deterioration and detachment
`of the bearing of the joint and proliferation of new
`osteoarticular tissue at the margin [2].
`Current treatment of OA is conservative and aimed at
`the maintenance of mobility in the joint by physiother-
`apy and relief of pain. The problem that OA poses in
`terms of new systemic therapeutic agents is that it does
`not affect all, or even many joints in the patient, as well
`as in other pathologies. If a few joints are affected, the
`
`*Corresponding author. Tel/fax: +39-577-234382.
`E-mail address: barbucci@unisi.it (R. Barbucci).
`
`possibility of local treatment by intra-articular injection
`does exist, but the drug would have to be confined in its
`effects; in this case, however, the compound would also
`need to have sufficient long-term action at the site of
`injection to avoid frequent local treatments.
`The natural polysaccharide hyaluronic acid (Hyal)
`has been claimed to be effective in pain relief and
`improving joint-motion following intra-articular injec-
`tions in OA patients [3,4]. In fact, it has been reported
`that Hyal provides symptomatic relief for its lubricating
`and cushioning properties, showing also some in vitro
`anti-inflammatory activity, and a possible disease-
`modifying effect in patients and animals [5–8]. However,
`localisation of response is likely to be a problem, as Hyal
`clearance is extremely rapid, becoming quickly degraded
`in vivo by enzymatic or hydrolytic reactions, which take
`place in biological environments [9,10]. The evidence,
`from in vitro and in vivo models of OA and from clinical
`trials,
`suggests
`that
`efficacy permanency depends
`strongly upon Hyal molecular weight. There are many
`commercial preparations of Hyal available. These
`
`0142-9612/02/$ - see front matter r 2002 Published by Elsevier Science Ltd.
`PII: S 0 1 4 2 - 9 6 1 2 ( 0 2 ) 0 0 1 9 4 - 1
`
`ALL 2047
`PROLLENIUM V. ALLERGAN
`IPR2019-01505 et al.
`
`

`

`4504
`
`R. Barbucci et al. / Biomaterials 23 (2002) 4503–4513
`
`relatively low
`include hyaluronan preparations of
`molecular weight (Hyalgan and ARTZ), of intermediate
`molecular weight (ORTHOVISC) but lower than that of
`the hyaluronan in normal healthy synovial fluid, and
`cross-linked hyaluronan of high molecular weight
`(Hylan, SYNVISC) [11]. The existing evidence indicates
`that the only potentially serious adverse event connected
`with commercial preparations of Hyal during OA
`treatment are joint
`infection, aseptic arthritis, and
`inflammation, even if
`in rare cases, being directly
`dependent upon the number of
`injections [12–14].
`Treatment with low MW preparations of hyaluronan
`seems to be effective. However, visco-supplementation
`with hyaluronan preparations may have slightly higher
`risks and less benefits than visco-supplementation with
`hylan, because the relatively low MW hyaluronan
`preparations require more injections which may incur
`higher costs and chance of infection.
`In order to obtain a longer effect of Hyal in the action
`site (joint), and consequently to reduce the risk of
`infection due to numerous injections, we decided to
`make the polysaccharide water insoluble by cross-
`linking it. The result of this reaction was the formation
`of a hydrogel. We have recently synthesised a series of
`new hydrogels, based on linear polysaccharides [15]. The
`hydrogel
`synthesis procedure allows us
`to predict
`the final desired cross-linking degree by controlling the
`stoichiometry reaction and conditions, and thus to
`obtain products with highly reproducible characteristics
`[16]. Within the series of hyaluronic acid-based hydro-
`gels, the one with a 50% cross-linking degree (Hyal
`50%) (in which the ratio of free and reacted carboxylic
`groups is 1:1) was chosen in order to study its potential
`applications in the OA treatment. The rheological
`behaviour of
`these Hyal-based hydrogels, and in
`particular the possibility to inject them through a needle,
`was studied. Moreover, the most appropriate sterilisa-
`tion technique among steam, Ethylene Oxide (EtO) and
`g-rays for this kind of hydrogels was evaluated. The in
`vitro cytotoxicity after EtO sterilisation was carried out
`and an experimental model was realised to study the in
`vivo effect of Hyal 50% in the treatment of a surgically
`created chondral defect in the rabbit knee.
`
`2. Materials and methods
`
`2.1. Materials
`
`(Hyal-Na, MWE
`sodium hyaluronate
`The
`150–200 kDa) was kindly provided by Farmila S.p.A.
`(Milano, Italy).
`Hyaluronic acid-based hydrogels with a cross-linking
`degree of 50% (Hyal 50%) was obtained with the
`procedure previously described [15]. Shortly, a solution
`of Hyal-Na underwent a sodium–hydrogen ionic
`
`exchange using Dowex 50WX8 resin (Fluka, Switzer-
`land) and then added to a 5% tetrabutylammonium
`hydroxide solution (Fluka, Switzerland) until a pH level
`of 7–8 was reached. The solution was then liophilised
`and tetrabutylammonium salt of hyaluronic acid (Hyal-
`TBA) was obtained. The Hyal-TBA salt was dissolved in
`N,N’-Dimethylformamide (DMF) under stirring and
`nitrogen flow. The solution was kept at about 01C,
`adding activating agent 2-chloro-1-methylpyridinium
`iodide (CMPJ) estabilished for the activation of the
`50% number of carboxylate groups. The cross-linking
`di-amine agent 1,3 di-aminopropane (1,3 DAP) was
`then added in large excess and the reaction was left at
`room temperature for 3–4 h. The reaction was catalysed
`by a small amount of triethylamine as hydrogen iodide
`captor. The hydrogel
`formed was then washed in
`ethanol and water several times until no more solvents
`and reagents were found in the washing solutions as
`demonstrated by UV absorption measurements. The
`Hyal 50% hydrogel was dried through liophylisation
`and the cross-linking degree was assessed by potentio-
`metric analysis [16].
`All others reagents were commercially available and
`utilised without further purification.
`
`2.2. Methods
`
`2.2.1. Rheological analysis
`2.2.1.1. Viscoelastic properties and steady shear viscosi-
`ty. The rheological characterisation was performed on
`the cross-linked polymers by adding bi-distilled water to
`lyophilised Hyal 50% samples to obtain a concentration
`of 10 mg/ml. The tests were performed on the gels before
`and after injection through a 22-gauge needle. A Bohlin
`VOR Rheometer (Bohlin Reologi A B, Lund, Sweden)
`was used at a controlled temperature of 251C. The
`geometrical figures were cone and plate (CP 5/30 cell).
`The lower-plate was forced to rotate or oscillate,
`whereas the stress transferred from the fluid to the
`upper-plate was measured with a linear variable
`displacement transducer (LVDT) system.
`The small-amplitude oscillatory shear experiments
`were performed to measure the time-dependent response
`of the samples and hence, to determine their linear
`viscoelastic properties. The frequency range spanned
`from 0.05 to 10 Hz. This range includes the physiologi-
`cal frequencies of knee ranging from 0.5 Hz (walk) to
`3 Hz (running). The small-amplitude oscillatory shear
`test was used successfully to determine the structure–
`mechanical properties relationship of materials [17,18].
`In a dynamic experiment, the material is subjected to
`a sinusoidal shear–strain:
`g ¼ g0 sin ðotÞ;
`where g0
`the shear–strain amplitude, o is
`the
`is
`oscillation frequency (which can also be expressed as
`
`

`

`R. Barbucci et al. / Biomaterials 23 (2002) 4503–4513
`
`4505
`
`2pof where f is the frequency in Hz) and t the time. The
`mechanical response, expressed as shear stress t of
`viscoelastic materials, is intermediate between an ideal
`pure elastic solid (obeying Hooke’s law) and an ideal
`pure viscous fluid (obeying Newton’s law). Therefore, it
`is out of phase with respect to the imposed deformation
`as expressed by
`t ¼ G0ðoÞg0 sinðotÞ þ G00ðoÞg0 cosðotÞ;
`where G0ðoÞ is the shear storage modulus and G00ðoÞ is
`the shear loss modulus. G0 provides information about
`the elasticity, or the energy stored in the material during
`deformation, whereas G00 describes the viscous character
`or the energy dissipated as heat.
`The combined viscous and elastic behaviour is given
`by the absolute value of complex shear modulus G :
`p
`ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
`G ðoÞ ¼
`G02 þ G002
`or by the absolute value of complex viscosity Z defined
`ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
`p
`as
`
`Z ¼
`
`G02 þ G002
`o
`which is usually compared with the steady shear
`viscosity in order
`to evaluate the effect of
`large
`deformations and shear rates on the structure of the
`material.
`The ratio between the viscous modulus and the elastic
`modulus is expressed by the loss tangent:
`tan d ¼ G00
`G0 ;
`where d is the phase angle.
`The loss tangent is a measure of the ratio of energy
`lost to energy stored in the cyclic deformation [17].
`Strain sweep tests at a fixed oscillation frequency
`(consisting in monitoring the visco-elastic properties
`while logarithmically varying the strain amplitude g0)
`were performed on the materials to determine the strain
`amplitude at which linear visco-elasticity is valid.
`The viscosity Z as function of shear rate was evaluated
`through steady state shear measurements.
`
`2.2.2. Sterilisation
`Hyal 50% was sterilised with the following three
`different methods: steam, g-rays and EtO. Eventual
`modifications of Hyal hydrogel as a consequence of
`sterilisation were evaluated by analysing the morphol-
`ogy and the swelling properties of the hydrogels before
`and after sterilisation.
`Samples of freeze-dried liophilised Hyal 50% hydro-
`gel of known weight
`(100 mg) were sealed in the
`following containers: pyrex glass, polyethylene, poly-
`ethylene containers sealed in laminated bags. They were
`sterilised by steam, g-rays and EtO, respectively.
`Steam sterilisation was conducted at 1201C, 1 atm for
`20 min. g-rays sterilisation was performed at 25 kilograin
`
`for 4 h. EtO sterilisation was carried out in a cylindrical
`ethylene oxide sterilisation apparatus where the samples
`were exposed for 12 h at 42731C to a 20% EtO:
`80%CO2 atmosphere with a relative humidity ranging
`from 50% to 95%. Then they were aerated for 7 days in
`a desorption chamber at 40+3/21C, with a pressure of
`400 mm Hg in order to remove the residual EtO.
`EtO residuals in aerated Hyal 50% samples were
`detected and quantified by chromatographic analysis of
`eluates, obtained by putting the samples in distilled
`water at 371C for 5 h.
`
`to
`2.2.2.1. Cytotoxicity of EtO residues. In order
`evalute whether the level of EtO residues was toxic, a
`cytotoxicity test was performed on Hyal 50% after
`sterilisation and areation. The test was performed with
`both mouse fibroblasts (cell
`line 3T3) and coronary
`venular endothelial cells (CVEC).
`Fibroblasts 3T3 were maintained in culture in
`DMEM, supplemented with 10% foetal calf serum
`(FCS), 1.2% l-glutamine and 1% penicillin–streptomy-
`cin (Sigma, Germany) on polystyrene flasks. The
`fibroblast cultures were incubated at 371C in an atmo-
`sphere of 5% CO2 until cells were harvested with the aid
`of trypsin and suspended in fresh medium.
`The CVEC were isolated and characterised as
`previously described [19]. Cells were maintained in
`culture in DMEM, supplemented with 10% bovine calf
`serum (CS) and antibiotics (100 U/ml penicillin and
`100 mg/ml streptomycin) on gelatin-coated dishes. Cells
`were cloned and each clone was sub-cultured up to a
`maximum of 25 passages. Passages between 15 and 20
`were used in these experiments.
`towards
`The cytotoxicity of
`sterilised hydrogels
`mouse fibroblasts 3T3 was evaluated by the direct
`contact method. Briefly, 3T3 cells (4000 cells/ml) were
`suspended in DMEM containing 10% foetal calf serum
`and placed on the bottom of each well of a multi-well
`plate (24-well). The multi-well was then incubated in an
`humidified atmosphere containing 5% CO2 at 371C till
`the cells reached confluence (24 h). Afterwards, 5 mg of
`sterilised Hyal 50%, swollen for 24 h at 371C in 500 ml of
`DMEM, were added to each fibroblast monolayer and
`incubated at 371C for 24 h. The Hyal 50% samples were
`then removed, the cells were fixed with glutaraldehyde
`and stained with trypan blue, and counted by direct
`observation with an optical microscopy (BX40, Olym-
`pus). The bottom of the polystyrene wells was used as a
`negative control and disks (+=15 mm) of organo-tin
`stabilised poly(vinylchloride) (PVC) as a positive con-
`trol, as recommended by ISO-standard (ISO-10993-5).
`Four samples of each different materials were tested.
`The cytotoxic effect of hydrogels towards CVEC was
`studied by trypan blue exclusion. Briefly, endothelial
`cells
`(200 000 cells/ml) were suspended in DMEM
`containing 10% serum. 500 ml of cell suspension was
`
`

`

`4506
`
`R. Barbucci et al. / Biomaterials 23 (2002) 4503–4513
`
`added to 3 mg of the dried hydrogels in 15 ml tubes.
`Cells were incubated at 371C for 4–24 h. Cells were then
`counted in a haemocytometer and the percentage of
`dead cells over the total number of cells was calculated.
`In parallel experiments, the ability of the cells to
`adhere to cell culture plastic and to proliferate was
`monitored for 1 week after a 24 h period of incubation
`with the hydrogels.
`
`2.3. Morphological and physico-chemical
`characterisation of sterilised hydrogels
`
`2.3.1. Scanning electron microscopy (SEM) analysis
`The morphology of Hyal 50% samples before and
`after sterilisation was studied by Scanning Electron
`Microscopy (SEM, XL20 Philips, The Netherlands).
`Five milligrams of each sample were solubilised with
`1.5 ml of distilled water, then freezed in liquid nitrogen
`and liophylised. Samples were successively mounted on
`SEM stubs, gold-sputtered with an automatic sputter-
`coater (BAL-TEC SCD 050, Balzers, Germany) and
`analysed.
`
`2.3.2. Swelling measurement
`The Hyal 50% swelling degree (SD) in water was
`determined before and after sterilisation. The SD was
`calculated with the following formula:
`SD ¼Pw Pd
`Pd
`
` 100;
`
`where Pw and Pd are, respectively, the weight of the
`swollen and dried hydrogels. In practice, established
`amounts of the liophilised Hyal 50% were enclosed in
`small bags made of a hydrophobic water-permeable net
`(Nylon) and immersed for 24 h at 251C in 50 ml of
`distilled water. They were then placed between two
`pieces of dry filter paper to wipe off the excess of water.
`The bags containing the swollen films were weighed in
`order to calculate Pw.
`
`2.4. Experimental in vivo study
`
`The study was conducted in compliance with the
`European and Italian Laws on animal experimentation,
`the principles stated in the Guide for the Care and Use
`of Laboratory Animals and the Animal Welfare
`Assurance No. A5424–01 by the National Institute of
`Health (NIH-Rockville, Maryland, USA).
`Sixteen New Zealand adult-male rabbits (2.970.5 kg
`body weight), aged eight months, were used. They were
`housed individually in stainless-steel cages and main-
`tained under the same environmental conditions (tem-
`perature 2070.51C, relative humidity 5575%, 12 h
`photoperiod). The animals were fed with a standard
`rabbit diet and filtered water ad libitum.
`
`All surgical procedures were performed in aseptic
`conditions, under general anaesthesia,
`induced by
`intramuscular
`injection of 35 mg/kg b.w. ketamine
`(Ketavet, Farmaceutici Gellini, Aprilia, LT, Italy) and
`5 mg/kg b.w. xylazine (Rompun, Bayer Italia, SpA,
`Milan, Italy), and maintained by means of a mixture of
`2% halothane (Fluothane, Zeneca Ltd., Macclesfield,
`UK) and oxygen/nitrous oxide (1/0.4 l/min), delivered
`by an automatic ventilator using a special-designed
`mask. The right femoral condyle of all rabbits was
`exposed by lateral–longitudinal skin incision. With the
`use of an operative microscope, the weight-bearing
`surface of each lateral femoral condyle was lightly
`scarified with a sharp scalpel blade to obtain a chondral
`defect of 5  4 mm, without damaging the subchondral
`bone. Afterwards, the surgical wounds were sutured.
`Immediately after surgery, eight animals (Hyal 50%
`group) received Hyal 50% solution (20 mg freeze-dried
`Hyal 50%/0.6 ml NaCl 0,9%) by percutaneous injection
`in the synovial capsule. The other eight rabbits were
`treated under the same conditions using NaCl 0.9% as
`control (Control Group). The rabbits were allowed to
`move freely in their cages and for 2 days all of them
`received antibiotics (Flumequine-Flumexil, ATI srl,
`Ozzano, BO, Italy) and analgesics (Orudis-Ketoprofene,
`Rhone Poulenc-Rorer, Milan, Italy). The Hyal 50%
`treatments and controls were repeated every 10 days, for
`three (30 days) or five times (50 days).
`At the end of the experimental times (30 and 50 days)
`four animals of each group were euthanised (Tanax,
`Hoechst, Frankfurt a M., Germany), after induction of
`general anaesthesia for the explant of condyles.
`
`2.4.1. Histomorphology
`condyles were dissected and
`All
`right
`femoral
`articular cartilage was examined macroscopically. For
`the histological evaluation the condyles were fixed in 4%
`buffered formalin, decalcified, dehydrated and em-
`bedded in paraffin. Sections of 7 mm in thickness were
`prepared from the centre of the repair area and stained
`with hematoxylin and eosine, alcian blue and toluidine
`blue. Samples were analysed by a blinded investigator,
`using a Zeiss Axioscope microscope, (C. Zeiss, Oberko-
`chen, Germany) connected to an image analyser
`(Kontron KS300, v.2, Kontron Elektronik, Munchen,
`Germany) to evaluate surface regularity and cellular
`morphology in the area of cartilage lesion.
`
`3. Results
`
`3.1. Rheological analysis
`
`3.1.1. Viscoelastic properties
`The rheological analysis provided the quantitative
`evaluation of the viscous and elastic responses of Hyal
`
`

`

`R. Barbucci et al. / Biomaterials 23 (2002) 4503–4513
`
`4507
`
`100
`
`10
`
`G', G'' [Pa]
`
`G' before injection
` G'' before injection
`
` G' after injection
`
` G'' after injection
`
`10
`
`1
`0.01
`
`0.1
`
`1
`
`Frequency [Hz]
`
`Fig. 1. Mechanical spectra of Hyal 50% before and after injection
`(C ¼ 10 mg/ml, T¼ 251C) through a 22-gauge needle.
`
`before injection
`after injection
`
`1
`
`Shear rate [1/s]
`
`10
`
`100
`
`10
`
`1
`
`0.1
`
`Viscosity [Pa s]
`
`0.01
`0.1
`
`Fig. 2. Flow curves of Hyal 50% before and after
`(C ¼ 10 mg/ml, T¼ 251C) through a 22-gauge needle.
`
`injection
`
`samples led to a reduction of the viscosity, demonstrated
`also by mechanical spectra of Hyal 50%, before and
`after injection (Fig. 1).
`
`3.2. Sterilisation
`
`is heat-sensitive (glass
`It is well known that Hyal
`transition temperature is about 601C) and sensible to g-
`rays, presenting chemical degradation and depolymer-
`isation with a decrease of molecular weight average [27].
`In any case, cross-linking may increase Hyal resistance
`to the sterilisation process, necessary to obtain a sterile
`hydrogel suitable for clinical application.
`Steam, g-rays and EtO sterilisation processes were
`conducted on liophilised Hyal 50%. The results of
`swelling degree measurements showed in Fig. 3 are
`
`50% and the estimation of the effect of the hydrogel
`injection through a 22-gauge needle on the visco-elastic
`properties.
`The storage moduli (G0) were greater than the loss
`(G00) within the frequency range analysed
`moduli
`(Fig. 1). The storage moduli, before and after injection,
`weakly depended upon frequency, while both the loss
`moduli showed a more marked frequency dependency.
`This polymer showed the features of ‘‘gel-like’’ material;
`it exhibited the typical behaviour of visco-elastic solids
`under small deformation conditions. The strain sweep
`tests showed that Hyal 50% was strain independent only
`for small deformation regions; this behaviour is typical
`of gels and provides an immediate distinction between
`gels and entanglement networks [20,21]. The rheological
`behaviour of Hyal 50% was characterised by a marked
`elasticity. This feature is not so remarkable in other
`commercial products such as Hyalgan, which shows
`essentially a viscous behaviour [22]. The introduction of
`chemical cross-linking between the chains led to a
`permanent stable network, and the intrinsic mobility of
`the molecules decreased dramatically. The network was
`also stabilised by chemical interactions, such as hydro-
`gen bonding [21].
`Both moduli of the non-injected sample were higher
`than that of the injected one. The hydrogel
`injected
`through the needle still behaved like a gel, but it showed
`a reduction of
`the dynamic moduli. The storage
`modulus,
`in fact, decreased from about 60 Pa to 25
`and the loss modulus from about 25 to 8, at 2.5 Hz
`(Fig. 1). The storage modulus of Hyal 50% before
`injection was similar to the modulus of non-injected
`Hylan GF 20 (Biomatrix, USA) [23], a cross-linked
`Hyaluronan of high molecular weight already used in
`OA treatment. Futhermore, the loss modulus of Hylan
`is one order of magnitude lower than that of Hyal 50%
`as well. The rheological behaviour of Hylan in this
`frequency range, in fact, is typical of entangled solution
`beyond the cross over frequency rather than a gel. For
`this reason the residence time of Hyal 50% should be
`higher than that of Hylan [24,25]. The injected samples
`still maintained rheological properties of great interest
`for biomedical applications because of the storage and
`loss modulus that were higher in comparison with those
`of normal adult synovial fluids at 2.5 Hz, which are 23
`and 7, respectively [24].
`Steady shear experiments were performed in order to
`evaluate the effect of the injection through a needle on
`the hydrogel. In Fig. 2, the flow curves of the injected
`and non-injected samples are shown. Both samples
`exhibited a non-Newtonian behaviour: shear thinning.
`The viscosity sharply drops with the shear rate (thin-
`ning). The dependency of the viscosity upon the shear
`rate could be satisfactory, represented by a power law
`with an exponent o1. This feature has been reported
`to be a characteristic of the gel [26]. The injection of the
`
`

`

`4508
`
`R. Barbucci et al. / Biomaterials 23 (2002) 4503–4513
`
`27000
`
`22000
`
`17000
`
`12000
`
`H A 50 %
`
`H A 50 % ga m m a rays
`
`H A 50 % EtO
`
`H A 50 % stea m (dry)
`
`Fig. 3. Hyal 50% swelling degree before and after g-ray, EtO and
`steam sterilization. Results are expressed as mean7SEM of analysis
`performed on four samples.
`
`relative to the hydrogel, before and after sterilisation.
`The data show that g-rays and EtO treatments did not
`affect the swelling behaviour of Hyal 50%, demonstrat-
`ing that the two sterilisation methods did not modify the
`physico-chemical characteristics of the hydrogel as well
`as their morphology, as revealed by SEM analysis also
`(Fig. 4a and b).
`On the contrary, steam sterilisation induced a marked
`decreasing of swelling degree (Fig. 3), probably due to
`breakages of hyaluronic acid chains in lower and
`different molecular weight fragments. These results are
`in accordance with the ones obtained for the other types
`of polysaccharidic hydrogels (alginic acid, carboxi-
`methylcellulose), obtained with the same cross-linking
`reaction, where sterilisation treatment of the liophilised
`samples provoked the reduction of the swelling values of
`about 50% [28].
`Since,
`toxicity of ethylene oxide residues after
`sterilisation is the major problem with this standard
`technique, there is a need to carefully estimate the best
`aerating time for porous materials, such as hydrogels,
`and evaluate the residual EtO levels and the non-toxicity
`of sterilised samples.
`As determined by chromatographic dosage of eluates,
`a very low level of EtO residues after seven days of
`areation was found in the samples (2.7  102 mg/ml).
`The EtO sterilised hydrogels were therefore tested for
`cytotoxicity of residual gas by direct contact method
`with fibroblast and endothelial cells. Two different cell
`lines were used because their behaviour towards EtO
`residual may be different (3T3 are tumoral and more
`resistant cells in comparison to primary endothelial
`cells).
`As shown in Fig. 5, after EtO sterilisation, the Hyal
`50% hydrogel did not reveal cytotoxic effect towards
`mouse fibroblast 3T3, after 24 h of contact. The cells
`showed a flat morphology, even if their number was
`lower than that in contact with the negative control
`
`(A)
`
`(B)
`
`Fig. 4. SEM analysis of: (A) native Hyal 50%; (B) EtO sterilized Hyal
`50% hydrogel. The morphology of the hydrogel is not modified as a
`consequence of the sterilisation process. Hyal 50% before and after
`sterilization appears to be composed by wide, stretched out and
`superimposed sheets. Among the sheets, hollows with varying
`dimensions are present.
`
`trend may be due to the
`polystyrene (PS). This
`experimental procedure. In fact, cells in contact with
`hydrogels adhered to PS substrate but were covered by
`Hyal 50% samples, reducing gas exchange and conse-
`quently cell-growth. On the contrary, in the negative
`control (i.e. the bottom of polystyrene wells), the cells
`were not covered by any other materials and gas
`exchanges were not limited. Few cells in contact with
`the positive control (PVC) were alive but showed a
`round morphology indicating cell suffering.
`Cells viability of microvascular endothelial cells was
`evaluated with contrast phase microscope after 4 and
`24 h. In endothelial cells in suspension, no significant
`toxic effect was found after 4 or 24 h incubation in the
`presence of the tested materials. The percentage of alive
`cells was in fact the same in contact with PS and Hyal
`50% and the trend of CVEC growth, the same of 3T3
`fibroblasts. When cells treated with the biomaterials for
`
`

`

`R. Barbucci et al. / Biomaterials 23 (2002) 4503–4513
`
`4509
`
`demonstrated to be a reliable model of OA [31,32],
`permitting to obtain an uniform and standard lesion
`with respect to other indirect (joint immobilisation or
`compression, mechanical
`loading)
`[37–39] or direct
`(joint fracture, ligament transection, papain injection)
`[40] methods.
`The realised model requires a meticulous technique
`and a correct methodology because of
`the small
`cartilage thickness (about 0.3 mm thick), and defect
`must not affect the subchondral bone, a variable that
`could influence tissue healing [33].
`The wounds of all animals healed without complica-
`tions and all
`the animals were in good general
`conditions.
`
`3.3.1. Macroscopic observations
`Gross examination of condyles (Fig. 6a) showed that
`at 30 days, the surface of the lesion in the control group
`appeared uneven, without signs of healing. In Hyal 50%
`group, at 30 days (Fig. 6b), the defect was filled with
`slightly irregular tissue that looked smoother, with less-
`defined margins. At 50 days, NaCl 0.9% treated defects
`in the control group (Fig. 6c) were incompletely filled
`with white and irregular fibrous tissue, in which the
`margin was clearly visible. At 50 days, in Hyal 50%
`
`24 h were let to adhere to tissue culture plastic, all the
`cells adhered and proliferated. Morphology of adherent
`cells was not modified by the treatment.
`
`3.3. In vivo experiment
`
`the most-used animals for
`is one of
`The rabbit
`studying experimental cartilage repair [8,29–36]. Among
`the different techniques to develop an experimental OA,
`we chose to follow a surgical method, taking some
`cautiousness as described in a previous paper [33]. This
`
`PS
`
`PVC
`
`Hyal 50%
`
`Materials
`
`40
`
`30
`
`20
`
`10
`
`0
`
`fields
`
`Number of cells for five
`
`Fig. 5. 3T3 cell viability after 24 h of contact with polystyrene (PS),
`organ-tin stabilized poly(vinylchloride) (PVC) and Hyal 50%. Data
`are means 7 SEM of counts from five fields on four specimens.
`
`(a)
`
`(b)
`
`
`
`(c)
`
`Fig. 6. Macroscopic photographs of the chondral lesion after surgery at 30 days: (a) Control group; (b) Hyal 50% group; and at 50 days: (c) Control
`group on the left and Hyal 50% group on the right.
`
`

`

`4510
`
`R. Barbucci et al. / Biomaterials 23 (2002) 4503–4513
`
`treated defects (Fig. 6d), the area of the lesion was more
`difficult to clearly define and showed less signs of
`degenerative changes. No signs of tissue reaction or
`inflammation were observed.
`
`3.3.2. Histological observations
`At 30 days, the chondral defect in the control group
`was evident, and there was no sign of healing (Fig. 7a).
`Hyal 50% group showed a fibro-cartilage layer, along
`the whole surface of the lesion, slightly thinner than
`adjacent intact articular cartilage (Fig. 7b).
`At 50 days, cartilage lesion of the control group
`(Fig. 8a) appeared covered by a thin and slightly
`
`irregular layer of fibrous tissue. This layer consisted of
`fibrous cells without any presence of chondrocytes. No
`samples showed evidence of proteoglycans production
`in the reparative tissue. On the contrary, Hyal 50%
`group (Fig. 8b) showed a thick mixed hyaline and fibro-
`cartilage layer. Chondrocytes are present as cluster and
`columnar formations (Fig. 9), dipped into the hyaline-
`like matrix.
`these evaluations are summarised in
`Results of
`the histological grading scale (which is a modification
`of that described by Driscoll et al.
`[41]) shown in
`Table 1, where a score was assigned to cell morphology,
`matrix staining, defect filling and surface regularity, with
`a total score for ‘‘perfect’’ repair or normal articular
`cartilage, equal
`to 12, deriving from the following
`characteristics of newly formed cartilage and defect
`
`(a)
`
`(b)
`
`Fig. 7. Histological sections from the chondral defect at 30 days in
`Control group (a) and Hyal 50% group (b). Hematoxylin and eosin
`staining, 2.5  .
`
`Fig. 8. Histological sections from the chondral defect at 50 days in
`Control group (a) and Hyal 50% group (b). Toluidine blue staining,
`1.25  .
`
`

`

`R. Barbucci et al. / Biomaterials 23 (2002) 4503–4513
`
`4511
`
`4; normal matrix
`healing: normal morphology,
`staining, 4; 100% filling of defect, 2; regular and smooth
`surface, 2.
`
`Fig. 9. Chondrocytes grown in the repaired tissue of Hyal 50% group.
`Toluidine blue staining, 10  .
`
`4. Discussion
`
`The clear applicability of Hyal 50% in the OA
`treatment depends on two factors: (i) the injection of the
`hydrogel through a needle which does not substantially
`alter its rheological characteristics; (ii) easy sterilisation.
`For bio-medical applications a shear thinning beha-
`viour, as well as that showed by Hyal 50%, is required.
`Moreover, in orthopaedic surgery low viscosity at high
`shear rate (when the material
`is forced through the
`needle) facilitates both the injection and removal of the
`gel, while the high viscosity at low shear rate permits
`higher residence time in the joint.
`The usual sterilisation techniques (EtO and g-ray) do
`not modify the characteristics of the hydrogel, in terms
`of swellability. In all incoming, both the sterilisation
`processes do not destroy the network structure of the
`Hyal 50% hydrogel. On the contrary, g-rays sterilisation
`induces breakage of the linear polysaccharide Hyal.
`The in vivo effect