ARTICLE IN PRESS Biomaterials xxx(2009)1-12 Contents lists available at science Direct Biomaterial Biomaterials ELSEVIER journalhomepagewww.elsevier.com/locate/biomaterials In vitro and in vivo evaluation of an alumina-zirconia composite for arthroplasty applications Olivier roualdes a,, Marie-Eve Duclos. 1. Dan Gutknecht b, 2, Lucien Frappart Jerome Chevalier b, 2. Daniel]. Hartmann a Universite de Lyon, UPSP 2007.03.135, Reparation Tissulaire, Interactions Biologiques et Biomateriaux, Universite Claude Bemard Lyon 1.8 Rockefeller-69373 b Universite de Lyon, INSA de Lyon, UMR CNRS 5510, 20 avenue Albert Einstein 69621 Villeurbanne Cedex, France Universite de lyon, INSERM U590, Pharmacogenomique et Chimioresistance, Laboratoire d'Anatomopathologie, Hopital Edouard Herriot, Bat. B, 5 place d'Arsonval-69437 LYON cedex 03. franc ARTICLE INFO A BSTRACT Article history. n order to improve the reliability and the mechanical properties of orthor Received 23 October 2009 ceramic composites starting with nanosized powders of alumina and zir ave been recently vailable online xxx developed. The aim of the present study was to investigate the biological tolerance of one of these sintered ceramics and of its alumina and zirconia constitutive nanosized powders with both in vitro and in vivo approaches. At first, osteoblasts and fibroblasts were cultured either upon sintered ceramic discs with polished or rough surfaces or in the presence of the corresponding alumina or zirconia powders at In vitro, the materials showed no deleterious effect on cell proliferation, extra-cellular matrix produ(.ts. various concentrations. Thereafter, we chronically injected these powders in the knee articulation of ra (human type I collagen and fibronectin) or on cell morphology In vivo, the histological examination showed only a very moderate and non-specific granulomatous response of the synovial membrane but o major inflammation as clinically described with metals or polyethylene wear debris. Besides its mproved physical properties, this recently developed alumina-zirconia composite showed satisfactory biocompatibility e 2009 Elsevier Ltd. All rights reserved. 1. Introduction arthroplasty were introduced more than 20 years ago [7-9]. Their use reduces wear rates of bearing components and produces Due to the increase of the life expectancy, the number of total negligible amount of ion release by comparison with metals ol hip arthroplasties is steadily growing. The lifetime of hip implants polymers. The clinical success associated to the use of ceramic has generally ranges from 12 to 15 years. In recent years, we have led to the implantation of more than 3.5 millions alumina observed a significant increase in revision surgery, which are components and more than 600,000 zirconia femoral heads frequently associated with clinical complications. It is therefore worldwide since 1990 [10]. If these materials show many advan- crucial to improve lifetime, reliability and biocompatibility of tages compared to metal, they have also shown some limitations. thopaedic implants such as hip prosthesis There is indeed a higher risk of fracture associated with ceramic Various combinations of different designs and materials have components [11, 12], even if it is strongly reduced today for both been proposed for acetabular cups and femoral heads 1-6. alumina and zirconia [12, 13. Some clinical studies show also Ceramic materials such as zirconia and alumina for total hip a hydrothermal degradation(often referred as aging) of zirconia [14 whose phase transformation is accelerated in aqueous ng author..Te:+33478777518;fax:+33478772819 ronment [15]. E-mail addresses: olivier roualdes@gmailcom(O. Roualdes). me. duclos@vet-lyon. The clinical success of a new ceramic material lies firstly on its ncgevanergr f f D Gutlinec ifrappartechu-lyon fr ability to resist to crack propagation and failure. New generations of che. univ-lyon1 fr(DJ. Hartmann). recently developed alumina-zirconia composites, often starting lel:+33478777518;fax:+33478772819 with very fine nanosized powders, meet this purpose [10, 16, 17]. 33472436125;fax:+33472438528 Such composites have also been demonstrated as safe against aging 3Tel:+33478785673;fax:+33478772819. 18. However, before a human clinical use, biological inocuity must .9612/s- see front matter o 2009 Elsevier Ltd. All rights reserved 0. 1016/ biomaterials. 2009.11. Please cite this article in press as: Roualdes 0, et al, In vitro and in vivo evaluation of an, Biomaterials(2009). doi: 10.1016/ j biomaterials 2009. 11.107
In vitro and in vivo evaluation of an alumina–zirconia composite for arthroplasty applications Olivier Roualdes a,*, Marie-Eve Duclos a,1 , Dan Gutknecht b,2 , Lucien Frappart c , Je´ roˆme Chevalier b,2 , Daniel J. Hartmann a,3 aUniversite´ de Lyon, UPSP 2007.03.135, Re´paration Tissulaire, Interactions Biologiques et Biomate´riaux, Universite´ Claude Bernard Lyon 1, 8 avenue Rockefeller – 69373 Lyon Cedex 08, France bUniversite´ de Lyon, INSA de Lyon, UMR CNRS 5510, 20 avenue Albert Einstein 69621 Villeurbanne Cedex, France cUniversite´ de Lyon, INSERM U590, Pharmacoge´nomique et Chimiore´sistance, Laboratoire d’Anatomopathologie, Hoˆpital Edouard Herriot, Bat. B, 5 place d’Arsonval – 69437 LYON cedex 03, France article info Article history: Received 23 October 2009 Accepted 26 November 2009 Available online xxx Keywords: Ceramic Composite Alumina Zirconia Nanoparticles Biocompatibility abstract In order to improve the reliability and the mechanical properties of orthopaedic hip prosthesis, new ceramic composites starting with nanosized powders of alumina and zirconia have been recently developed. The aim of the present study was to investigate the biological tolerance of one of these sintered ceramics and of its alumina and zirconia constitutive nanosized powders with both in vitro and in vivo approaches. At first, osteoblasts and fibroblasts were cultured either upon sintered ceramic discs with polished or rough surfaces or in the presence of the corresponding alumina or zirconia powders at various concentrations. Thereafter, we chronically injected these powders in the knee articulation of rats. In vitro, the materials showed no deleterious effect on cell proliferation, extra-cellular matrix production (human type I collagen and fibronectin) or on cell morphology. In vivo, the histological examination showed only a very moderate and non-specific granulomatous response of the synovial membrane but no major inflammation as clinically described with metals or polyethylene wear debris. Besides its improved physical properties, this recently developed alumina–zirconia composite showed satisfactory biocompatibility. 2009 Elsevier Ltd. All rights reserved. 1. Introduction Due to the increase of the life expectancy, the number of total hip arthroplasties is steadily growing. The lifetime of hip implants generally ranges from 12 to 15 years. In recent years, we have observed a significant increase in revision surgery, which are frequently associated with clinical complications. It is therefore crucial to improve lifetime, reliability and biocompatibility of orthopaedic implants such as hip prosthesis. Various combinations of different designs and materials have been proposed for acetabular cups and femoral heads [1–6]. Ceramic materials such as zirconia and alumina for total hip arthroplasty were introduced more than 20 years ago [7–9]. Their use reduces wear rates of bearing components and produces negligible amount of ion release by comparison with metals or polymers. The clinical success associated to the use of ceramic has led to the implantation of more than 3.5 millions alumina components and more than 600,000 zirconia femoral heads worldwide since 1990 [10]. If these materials show many advantages compared to metal, they have also shown some limitations. There is indeed a higher risk of fracture associated with ceramic components [11,12], even if it is strongly reduced today for both alumina and zirconia [12,13]. Some clinical studies show also a hydrothermal degradation (often referred as aging) of zirconia [14] whose phase transformation is accelerated in aqueous environment [15]. The clinical success of a new ceramic material lies firstly on its ability to resist to crack propagation and failure. New generations of recently developed alumina–zirconia composites, often starting with very fine nanosized powders, meet this purpose [10,16,17]. Such composites have also been demonstrated as safe against aging [18]. However, before a human clinical use, biological inocuity must * Corresponding author. Tel.: þ33 4 78 77 75 18; fax: þ33 4 78 77 28 19. E-mail addresses: olivier.roualdes@gmail.com (O. Roualdes), me.duclos@vet-lyon. fr (M.-E. Duclos), dan.gutknecht@insa-lyon.fr (D. Gutknecht), l.frappart@chu-lyon.fr (L. Frappart), jerome.chevalier@insa-lyon.fr (J. Chevalier), daniel.hartmann@recherche.univ-lyon1.fr (D.J. Hartmann). 1 Tel.: þ33 4 78 77 75 18; fax: þ33 4 78 77 28 19. 2 Tel.: þ33 4 72 43 61 25; fax: þ33 4 72 43 85 28. 3 Tel.: þ33 4 78 78 56 73; fax: þ33 4 78 77 28 19. Contents lists available at ScienceDirect Biomaterials journal homepage: www.elsevier.com/locate/biomaterials ARTICLE IN PRESS 0142-9612/$ – see front matter 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2009.11.107 Biomaterials xxx (2009) 1–12 Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107
ARTICLE IN PRESS be proved. Indeed, if alumina or zirconia are most often considered 21. Ceramic powders bioinert materials. the new formulation and process of the nposites and their nanosized constitutive particles may lead which are commercially available under the trade ports about in vitro or in vivo biocompatibility on monolithic Tokyo, Japan) respectively. The powders were observed by Scanning Electro umina or zirconia ceramics, there are fewer papers about microscopy, which revealed a mean particle size of 350 nm and 40 nm for alumina alumina-zirconia composites. Affatato et al. demonstrated that the and zirconia respectively, in agreement with data from the producers. The Zeta potential of the powders versus pH has been shown in a previous paper [16] Both primary osteoblast proliferation onto alumina-zirconia composite powders presented a high Zeta potential in the acidic range(ie. Zeta potential samples was found to be not significantly different from that onto higher that 60 mv for pH 6. The isoelectric point for both powders was vitro biocompatibility of such around 9 For physiological pH (pH of the cell culture medium, around 7.4). the Zet a composite, but with a porous structure[20 of soft agglomeration. This is evidenced on Fig. 1(a-b). whicl Throughout the lifetime of a ceramic implant, wear debris are measurements(Malvern, Matersizer, 2000)performed on the powders previous constantly generated from the hip joint. They are recognized as the vortexed in the cell culture media. The particle size distribution is given in number in development of (Fig. la)and in volume %(Fig. 1b). If the majority (in numbers)of particles were and aseptic loosening. One strategy to reduced osteolysis consists dispersed and separated from each other(see Fig 1a: the largest number of partic of choosing bearing surfaces associated with better wear proper consistent with SEM particle size) a large volume portion present. This is especially true for zirconia. This means that both ies. This can be achieved by reducing the particles size and surface single particles(a large portion in number) and agglomerates(a large portion in roughness. They are many reports about these parameters. Some of volume)would be present in the cell culture medium. them demonstrated that materials composed with nanosized (<100 nm)alumina particles have less detrimental effect on 22. Ceramic blocks m production of wear debris, their size may be very small and their promote good dispersion of both powders [16]. Ball milling was then performe ong-term effect relatively unknown. It has been reported that using high purity alumina balls in a plastic jar for 24 h. These dispersed slurries were particles(commercially pure titanium, Ti-6Al-4V alloy)less than spray dried using an ultrasonic spray drier(20 kHz, Sodeva, Ch 3-5 um could be phagocytosed by mature osteoblasts. They had n effect on cell viability but induced a dose dependent decrease in cell a Zirconia Alumina proliferation [27]. More precisely, Gutwein compared the effect of different alumina and titania grain sizes on cell culture. he demonstrated that osteoblast viability and densities were influ- enced solely by particle size and concentration [21 He concluded that osteoblast function and adhesion were more preserved when lumina grain size decreased under 100 nm. Generally wear debris 5 10 pression of type I collagen [28, 29], osteocalcin, alcaline phos- haase[ 30,31 and osteopontin 32 by osteoblasts. To understand ow they might modulate cell activity, we examined in this stud the response of MG-63 osteoblast like cells and fibroblasts to o alumina and zirconia particle 22 To date, there are several in vitro studies about the effect of 5 1 blasts [32-37] but there are less about the in vitro effect of nano- 0.1 sized particles [21, 27,38 and almost none about the in vivo Particles or agglomerate sizes(um) response whatever the ceramic or other orthopaedic materials particles size (39-45]. Most of them consist of the creation of b a contact between particles and bone using a variety of animal nodels. The present in vivo work consisted of an evaluation of the 7 response of a normal articulation chronically exposed to an patibility of an alumina-zirconia ceramic composite for ortho-o ment of this biomaterial it includes the material in its final mposition and its alumina and zirconia particulate constituents hich will enable us to understand and measure the potential toxicity of the associated wear debris 100 1000 Particles or agglomerate sizes(um) ng ceramic powders and the processing of the cor ascribed previously [16, 46]. We present here the main features of the Fig. 1. Particles sizes reparti aber (a) and in volume(b)of alumina an zirconia powder in DMEM afte min by vortex. lease cite this article in press as: Roualdes O, et al, In vitro and in vivo evaluation of an., Biomaterials(2009), doi: 10.1016/ j biomaterials 2009. 11.107
be proved. Indeed, if alumina or zirconia are most often considered as bioinert materials, the new formulation and process of the composites and their nanosized constitutive particles may lead to new unexpected biological responses. Although there are many reports about in vitro or in vivo biocompatibility on monolithic alumina or zirconia ceramics, there are fewer papers about alumina–zirconia composites. Affatato et al. demonstrated that the primary osteoblast proliferation onto alumina–zirconia composite samples was found to be not significantly different from that onto commercial alumina samples [19]. In an other work, He et al. also provide the satisfactory in vitro biocompatibility of such a composite, but with a porous structure [20]. Throughout the lifetime of a ceramic implant, wear debris are constantly generated from the hip joint. They are recognized as the major initiating event in development of periprosthetic osteolysis and aseptic loosening. One strategy to reduced osteolysis consists of choosing bearing surfaces associated with better wear properties. This can be achieved by reducing the particles size and surface roughness. They are many reports about these parameters. Some of them demonstrated that materials composed with nanosized (100 nm) [21–26]. However, if the use of nano-composed ceramic enhances cell properties and reduces the production of wear debris, their size may be very small and their long-term effect relatively unknown. It has been reported that particles (commercially pure titanium, Ti-6Al-4 V alloy) less than 3–5 mm could be phagocytosed by mature osteoblasts. They had no effect on cell viability but induced a dose dependent decrease in cell proliferation [27]. More precisely, Gutwein compared the effect of different alumina and titania grain sizes on cell culture. He demonstrated that osteoblast viability and densities were influenced solely by particle size and concentration [21]. He concluded that osteoblast function and adhesion were more preserved when alumina grain size decreased under 100 nm. Generally, wear debris of various orthopaedic materials have been shown to decrease expression of type I collagen [28,29], osteocalcin, alcaline phosphatase [30,31] and osteopontin [32] by osteoblasts. To understand how they might modulate cell activity, we examined in this study the response of MG-63 osteoblast like cells and fibroblasts to alumina and zirconia particles. To date, there are several in vitro studies about the effect of micron-size ceramic wear particles on cell function such as osteoblasts [32–37] but there are less about the in vitro effect of nanosized particles [21,27,38] and almost none about the in vivo response whatever the ceramic or other orthopaedic materials particles size [39–45]. Most of them consist of the creation of a contact between particles and bone using a variety of animal models. The present in vivo work consisted of an evaluation of the response of a normal articulation chronically exposed to an important amount of nanosized ceramic particles. The aim of the present study was to investigate the biocompatibility of an alumina–zirconia ceramic composite for orthopaedic applications, especially for total hip arthroplasty. Here, we present an in vitro and in vivo approach for the biological assessment of this biomaterial. It includes the material in its final composition and its alumina and zirconia particulate constituents which will enable us to understand and measure the potential toxicity of the associated wear debris. 2. Materials and methods Details on the starting ceramic powders and the processing of the composites have been described previously [16,46]. We present here the main features of the powders and the composite. 2.1. Ceramic powders The zirconia and alumina powders used for the processing of composites were alpha alumina and pure zirconia, which are commercially available under the trade names Ceralox APA 05 (Condea, Hamburg, Germany) and TZ0 (Tosoh Corporation, Tokyo, Japan) respectively. The powders were observed by Scanning Electron Microscopy, which revealed a mean particle size of 350 nm and 40 nm for alumina and zirconia respectively, in agreement with data from the producers. The Zeta potential of the powders versus pH has been shown in a previous paper [16]. Both powders presented a high Zeta potential in the acidic range (i.e. Zeta potential higher that 60 mV for pH 6. The isoelectric point for both powders was around 9. For physiological pH (pH of the cell culture medium, around 7.4), the Zeta potential was not high enough to allow perfect dispersion and promote the presence of soft agglomeration. This is evidenced on Fig. 1 (a–b), which shows granulometry measurements (Malvern, Matersizer, 2000) performed on the powders previously vortexed in the cell culture media. The particle size distribution is given in number (Fig. 1a) and in volume % (Fig. 1b). If the majority (in numbers) of particles were dispersed and separated from each other (see Fig. 1a: the largest number of particle for each material was consistent with SEM particle size), a large volume portion of agglomerates was present. This is especially true for zirconia. This means that both single particles (a large portion in number) and agglomerates (a large portion in volume) would be present in the cell culture medium. 2.2. Ceramic blocks Alumina–zirconia composites were processed by a conventional powder-mixing technique. The powders were mixed in appropriate amounts in water and electrostatic dispersion was used to prepare stable slurries, since pH ¼ 4.5 was shown to promote good dispersion of both powders [16]. Ball milling was then performed using high purity alumina balls in a plastic jar for 24 h. These dispersed slurries were spray dried using an ultrasonic spray drier (20 kHz, Sodeva, Chambery, France). Zirconia Alumina 0.01 0.1 1 10 100 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 a Particles or agglomerate sizes (µm) Particles distribution (% number) Particles distribution (% volume) 0.01 0.1 1 10 100 1000 0 1 2 3 4 5 6 7 8 b Particles or agglomerate sizes (µm) Fig. 1. Particles sizes repartition in number (a) and in volume (b) of alumina and zirconia powder in DMEM after mixing 5 min by vortex. 2 O. Roualdes et al. / Biomaterials xxx (2009) 1–12 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107
ARTICLE IN PRESS O. Roualdes et al. Biomaterials xxx(2009)1-12 Spray dried granules were pressed uniaxially (150 MPa)and isostatically(300 MPa) wells were recovered and stored at -70C for further experiments(see below )and in order to obtain medium and 100 uL of MIT (5 mg/mL in PBS, Gibco) Cells (Hot Isostatic Pressing) at 1520.C for 1 h, with a press p with shaking. In order to eliminate the particles that can They have been fully characterized in interfere with the optical density measurement, each well was centrifuged (5 min, to aging in previous works [10, 16, 17 Finally samples were mirror-polished using 1600 rpm)and the optical density was measured at 570 nm. diamond suspensions(Buhler, Lyon, France), either down to 1 um or only down to 30 um to give a rougher surface state 2.4. Extra-cellular matrix assay uman fibronectin(FN)and human type I collagen(CIH)contents in the culture 3. Cell culture medium were assessed by competition ELISA following standard procedures. Briefly nicro-well plates(96-well, Costar, Coming inc) were previously coated with 200 HL 2.3.1.cels of FN or CiH(respectively diluted at 1 300 and 1 /160 from a 1 mg/mL solution in PBS, man fibroblasts, obtained from healthy new born arm skin explants Biomerieux, Marcy I'Etoile, France) for 24 h at 4C then rinsed in PBS. For FN by Dr M. Cabot, Cent oakville, MD)were cultured in 25 cm" tissues flasks( Costar, Corning Inc )in the Lyon, France). After 1h30 at room temperature wells were rinsed in PBS and 200 uL ence of Dulbecco's modified eagle,'s medium(DMEM, Gibco) supplemented of peroxydase labeled anti-rabbit antibody diluted at 1/1000 were added for 1h30 at vere incubated at 37C, with 5% COz. saturated with humidity until confluent Inc onolayer and then sub-cultured after trypsination MG-63 osteoblast like cells are with 100 uL of CIH antibody diluted at 1/750 in PBS BS ALP activity in ere rinsed in pBs xtra-cellular matrix containing in particular collagen and nd 200 uL of peroxydase labeled anti-rabbit antibody d the expression of classical cell adhesion molecules such integrins 149, 50]. For for 1h30 at room temperature both cell types, experiments were carried out on cells from passage 4 through 30. After the second antibody incubation for both FN and CH assays, micro-well Using von Kossa staining v (named after"osteoblasts")to mineralize(data not shown). 202 were added f ength of 450 nm using a multiscan reader (Multiskan EX Thermo Scientific Electron Corporation). Intra ell culture(dry air, 180C for 4 h). After were lower than 10% for both FN and CIH ELISA. sterilization, ceramic substrates were deposited into a 24-well plate (co 2.5. Cell morphology (7C, 5% COz and wet atmosphere)for 2 h in order to pn hen 1.9 mL of culture medium was added into each well. the medium was ashed 2 times quickly with PBS(Gibco)in controls, cells were seeded directly on the plastic culture surfaces and paraformaldehyde in PBS for 1 h at room e the ceramic substrates massie Blue Brilliant(Sigma Aldrich) and observed using light microscopy. 2二 used for osteoblasts and fibroblasts experiments. 2.6. Animal experiment led onto 24-well culture plates(Costar, Corning Inc )and incubated using standard culture conditi ur male Sprag Dale sighing ading. After 24 (mean+ SEM) were used in this study. The animals were randomly assigned to 3 00 or 1000 ug/mL particles of each powder sterilized prior cell culture(dry groups. The first group received an intra-articular injection of sterile saline at day t 80. for 4 h). This time represented day O. The medium was changed at day 3, 6 The second and the third groups were injected three times with alur ely. Animals during the 10. The cell culture controls followed the same procedure but without particle addition into the medium. riod at the Institut Claude-Bourgelat, Lyon. All experiments were conducted in full ompliance with the National Veterinary School of Lyon(ENVL)Ethical Committee Guideline (p agreement n 0826)according to current European and french Legislations for Animal Protection. At 3, 6 and 10 days of culture, fibroblast and os by MIT assay (3-14,5-dimethylthiazol-2-yll-2, 5-dip um bromide, sigma Analgesia was provided before and after the operation with morphine (3 mg/kg drich) This reagent is transformed by mitochondrial dehydrogenases of active cells The animals were anesthetized with a mixture of isoflurane in oxygen. The right into formazan, providing a measure of cell activation and viability. The medium of al red for injection. Intra-articular injections were nade with 29G needle by a transpatellar approach We injected 50 uL of either saline for control, alumina or zirconia(0. 1% in volume )at 0, 2 and 4 weeks. A arlan re euthanasied two weeks after the last injection(6 weeks after the first injec- tion) by intracardiac injection of Dolethal euthanasia solution after anesthesic amine 90 mg/kg and xylazine 10 mg/kg, intramuscular injection Synovium and articular surface of femur, tibia and patella were removed and examined for gross evidence of abnormalities. 2.7. Histological The whole sy membrane and patella were removed from the joint and fixed saline for 72 h. The hen decalcified with EDTA 10% for 7 days before paraffin-embedding. 5 um sections from the saggital cut and stained with Goldner trichrome toluidine blue or were scored independently by two observers. A semi-quantitative three points scale tero represented the normal or the lowest state, one represented a slight Fig. 2. Scanning electron microscopy image of the composite after polishing and moderate increase and two the highest level of increase. The highest level of the thermal etching showing micron-size alumina grains(in grey)and zirconia sub-micron present score was much lower than that observed in pathologies such as osteoar particles(in white thritis or rheumatoid arthritis. Indeed, we adjusted our scale to the maximum Please cite this article in press as: Roualdes 0, et al, In vitro and in vivo evaluation of an, Biomaterials(2009). doi: 10.1016/ j biomaterials 2009. 11.107
Spray dried granules were pressed uniaxially (150 MPa) and isostatically (300 MPa) in order to obtain small cylinders of 14 mm diameter and 2 mm thickness. Most of the samples were then sintered in air at 1520 C for 2 h, and then post treated by HIP (Hot Isostatic Pressing) at 1520 C for 1 h, with a pressure of 200 MPa. The final microstructure is shown in Fig. 2. All the processed samples reached full density. They have been fully characterized in terms of mechanical properties and re´ sistance to aging in previous works [10,16,17]. Finally samples were mirror-polished using diamond suspensions (Buhler, Lyon, France), either down to 1 mm or only down to 30 mm to give a rougher surface state. 2.3. Cell culture 2.3.1. Cells Human fibroblasts, obtained from healthy new born arm skin explants (provided by Dr MT. Zabot, Centre de Biotechnologie cellulaire, Lyon, France), and osteoblasts (human osteoblast like cell MG-63, American Type Culture Collection, Rockville, MD) were cultured in 25 cm2 tissues flasks (Costar, Corning Inc.) in the presence of Dulbecco’s modified eagle’s medium (DMEM, Gibco) supplemented with 10% foetal calf serum (Gibco) and 1% penicillin/streptomycin (Bayer). Cultures were incubated at 37 C, with 5% CO2, saturated with humidity until confluent monolayer and then sub-cultured after trypsination. MG-63 osteoblast like cells are characterized by an increased ALP activity in response to 1,25-(OH)2D3, the synthesis of an extra-cellular matrix containing in particular collagen and fibronectin [47,48], and the expression of classical cell adhesion molecules such integrins [49,50]. For both cell types, experiments were carried out on cells from passage 4 through 30. Using von Kossa staining we have checked the capacity of these osteoblasts-like cells (named after ‘‘osteoblasts’’) to mineralize (data not shown). 2.3.2. Cell culture on ceramics discs The ceramic discs were sterilized prior cell culture (dry air, 180 C for 4 h). After sterilization, ceramic substrates were deposited into a 24-well plate (Costar, Corning Inc.). 100 mL of culture medium containing 5 104 cells (fibroblasts or osteoblasts) were seeded upon each ceramic substrate. Plates were incubated in standard culture conditions (37 C, 5% CO2 and wet atmosphere) for 2 h in order to promote cell adhesion then 1.9 mL of culture medium was added into each well. The medium was changed at 3, 6 and 10 days. For the controls, cells were seeded directly on the plastic culture surfaces and treated like the ceramic substrates. 2.3.3. Cell culture in the presence of particles The same procedure was used for osteoblasts and fibroblasts experiments. 2.5 104 cells were seeded onto 24-well culture plates (Costar, Corning Inc.) and incubated using standard culture conditions in order to promote cell adhesion and spreading. After 24 h, the medium was replaced by 2 mL of fresh medium containing 100 or 1000 mg/mL particles of each powder sterilized prior cell culture (dry air, 180 C for 4 h). This time represented day 0. The medium was changed at day 3, 6 and 10. The cell culture controls followed the same procedure but without particles addition into the medium. 2.3.4. Cell proliferation At 3, 6 and 10 days of culture, fibroblast and osteoblast proliferation was assessed by MTT assay (3-[4,5-dimethylthiazol-2-yl]-2, 5-diphenyl tetrazolium bromide, Sigma Aldrich). This reagent is transformed by mitochondrial dehydrogenases of active cells into formazan, providing a measure of cell activation and viability. The medium of all wells were recovered and stored at 70 C for further experiments (see below) and replaced by 900 mL of fresh medium and 100 mL of MTT (5 mg/mL in PBS, Gibco). Cells culture plates were incubated in 5% (v/v) CO2 in air at 37 C for 3 h. The medium was removed and replaced by 500 mL of ethanol/DMSO solution (vol/vol) and incubated for 5 min at room temperature with shaking. In order to eliminate the particles that can interfere with the optical density measurement, each well was centrifuged (5 min, 1600 rpm) and the optical density was measured at 570 nm. 2.4. Extra-cellular matrix assay Human fibronectin (FN) and human type I collagen (CIH) contents in the culture medium were assessed by competition ELISA following standard procedures. Briefly micro-well plates (96-well, Costar, Corning inc) were previously coated with 200 mL of FN or CIH (respectively diluted at 1/300 and 1/160 from a 1 mg/mL solution in PBS, Biomerieux, Marcy l’Etoile, France) for 24 h at 4 C then rinsed in PBS. For FN assessment, 100 mL of standard point or medium sample were incubated with 100 mL of human FN antibody diluted at 1/1600 in PBS BSA 3% (24 911 antibody, Novotec, Lyon, France). After 1h30 at room temperature wells were rinsed in PBS and 200 mL of peroxydase labeled anti-rabbit antibody diluted at 1/1000 were added for 1h30 at room temperature. For CIH assessment, 100 mL of standard point or medium sample were incubated with 100 mL of CIH antibody diluted at 1/750 in PBS BSA 3% (Human type I collagen antibody 20111, Novotec). After 3 h at room temperature wells were rinsed in PBS and 200 mL of peroxydase labeled anti-rabbit antibody diluted at 1/800 were added for 1h30 at room temperature. After the second antibody incubation for both FN and CIH assays, micro-well plates were washed in PBS and 200 mL of o-Phenylenediamine dihydrochloride in H2O2 were added for 30 min on the dark at room temperature. Optical densities were obtained at the wavelength of 450 nm using a multiscan reader (Multiskan EX, Thermo Scientific Electron Corporation). Intra and inter-assay reproductibilities were lower than 10% for both FN and CIH ELISA. 2.5. Cell morphology Cells were cultured in the presence of particles at 100 mg/mL in DMEM as described previously (paragraph 2.3.3). At 2, 6, 24 h and 3, 6 10 and 15 days of culture, cells were washed 2 times quickly with PBS (Gibco) in order to remove particles then fixedwith 4% paraformaldehyde in PBS for 1 h at room temperature. Cells were stained with Coomassie Blue Brilliant (Sigma Aldrich) and observed using light microscopy. 2.6. Animal experiment Thirty-four male Sprage Dawley rats 4 months old and weighing 479 38 g (mean SEM) were used in this study. The animals were randomly assigned to 3 groups. The first group received an intra-articular injection of sterile saline at day 0. The second and the third groups were injected three times with alumina and zirconia particles respectively. Animals were housed in groups during the same period at the Institut Claude-Bourgelat, Lyon. All experiments were conducted in full compliance with the National Veterinary School of Lyon (ENVL) Ethical Committee Guideline (protocol agreement n 0826) according to current European and French Legislations for Animal Protection. Analgesia was provided before and after the operation with morphine (3 mg/kg). The animals were anesthetized with a mixture of isoflurane in oxygen. The right stifle joint was clipped and prepared for injection. Intra-articular injections were made with 29G needle by a transpatellar approach. We injected 50 mL of either saline for control, alumina or zirconia (0.1% in volume) at 0, 2 and 4 weeks. A veterinarian closely monitored the animals for infections and other complications. The animals were euthanasied two weeks after the last injection (6 weeks after the first injection) by intracardiac injection of Dolethal euthanasia solution after anesthesic protocol (ketamine 90 mg/kg and xylazine 10 mg/kg, intramuscular injection). Synovium and articular surface of femur, tibia and patella were removed and examined for gross evidence of abnormalities. 2.7. Histological preparation and examination The whole synovial membrane and patella were removed from the joint and fixed in neutral-buffered formal saline for 72 h. They were then decalcified with EDTA 10% for 7 days before paraffin-embedding. 5 mm sections from the saggital plane of the synovium was cut and stained with Goldner trichrome, toluidine blue or hematoxilin eosine safranin. Morphological studies were performed using a light microscope (Zeiss Axioskop 50 Germany, Oberkochen, software Explora Nova Morpho Expert, La Rochelle, France). All areas of each biopsy section were examined and histological features were scored independently by two observers. A semi-quantitative three points scale was adapted from previously used scoring systems for synovial tissue [44,45,51–55]. Zero represented the normal or the lowest state, one represented a slight or moderate increase and two the highest level of increase. The highest level of the present score was much lower than that observed in pathologies such as osteoarthritis or rheumatoid arthritis. Indeed, we adjusted our scale to the maximum Fig. 2. Scanning electron microscopy image of the composite after polishing and thermal etching showing micron-size alumina grains (in grey) and zirconia sub-micron particles (in white). O. Roualdes et al. / Biomaterials xxx (2009) 1–12 3 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107
ARTICLE IN PRESS teration observed in our conditions. Moreover, the present limited three points particles at 100 ug/mL in DMEM. Results are presented in the Fig. 4. very reproducible. The histological assessment was made in two times. l From 3 to 15 days of culture, we observed a regular increase of the ncluding the synovial lining su cell growth for both types of cell. The cell proliferation was not field. The synod embrane thickness and the villous hyperplasia were scored. affected by the presence of alumina or zirconia particles. Indeed, no hereafter, using 40x power field we examined separately the synovial lining cell significant differences were observed in the cell proliferation synovial tissue. Results presented were the between cultures with or without particles. The study of the cell bserved in five continuous selected helds from the proximal region of the synovial viability by blue trypan exclusion confirmed this result(data not ssue near the patella. In the synovial lining cell layer, we scored the cell layer shown). In order to assess the cell ability to produce ECM, all DMEM perplasia(number of cell layers)and the cell hypertrophy(o= normal flat cells, cuboidal cells, 2= cylindrical cytoplasmic extensions) In the sub-synovial supernatants were collected for CIH and FN assays Results are gion, the cell infiltration was evaluated by scoring the amount of presented in the Fig. 4c-f. Both CIH or FN productions and cell istiocytes cells(O=no or very low. 1=moderate and 2= high amount) and growth increased in comparable way. Compared to the control aggregates). Finally, vascularisation was scored (0=f culture (without particles). cells cultured in the presence of 1= moderate increase and 2 alumina or zirconia at concentration of 100 ug/mL did not affect these two proteins synthesis Results were similar for fibroblasts and osteoblasts from 3 to 15 days of culture. 3.2.2. Particles at 1000 ug/mL In order to evaluate the influence of the particles concentration on (r)and Spearman (r) correlation tests using GraphPad Prism version 4.02 software cell functions, we reproduced accurately all this experiment with ten for Macintosh(GraphPad Software, San Diego, CA, USA). Statistical significance was fold more particles(1000 ug/mL) Due to this particles amount and considered at p<0.05. All in vitro experiments were run in triplicate and repeated at compared to the cell number and the cell culture surface, we quickly observed an important particle layer that recovered all cells. This extreme condition for an in vitro study. 3. Results However, fibroblasts and osteoblasts growth were not really affected by this particles amount. Results are presented in the Fig. 5. Compared 3. 1. In vitro cytocompatibility of ceramic discs with control culture, we observed just a light increase of the fibroblast growth when cultured in the presence of alumina aPA(p 0.05) and Compared to cells cultured on plastic (reference substrate rep- zirconia Tz-0(p <0.05)after 15 days. This result was associated with sented by an empty well), fibroblasts and osteoblasts proliferation an increase of the fibronectin synthesis in the pi e of both decreased on both rough and polished ceramics discs tested in this alumina or zirconia particles at the same culture period (respectively study at all time periods even if the differences were not always p<0.01 and p<0.001 compared with control culture). However, no tatistically significantFig3 a-b). The proliferation of fibroblasts significant variation of collagen synthesis was detected In the pres- was significantly decreased on rough discs after 3 and 6 day of ence of the same particle amount, osteoblasts growth was also not ulture compared to the control(Fig. 3a, P<0.01). For the osteo- affected by the presence of alumina or zirconia after 3, 6 and 10 days of blasts, the proliferation was significantly lower on rough ceramics culture. It was just slightly reduced at day 15(p<0.01)for both discs after 3, 6 and 10 days of culture(respectively p<0.01, 0.05 and alumina and zirconia. At this time, this variation was associated with p<0.05 and p <0.01)than control( Fig. 3b). The cell functions were Regarding the results of cell growth and ECM synthesis in our study assessed by the assays of two important proteins of the extra-cellular conditions, alumina and zirconia particles showed no cytotoxiceffect. matrix produced by fibroblasts and osteoblasts: human type I collagen and fibronectin(Fig. 3c-f). Compared to respective control 3. 3. Cell morphology in the presence of ceramic particles cultures, the levels of the two proteins were generally weaker but not significantly decreased for both fibroblasts and osteoblasts. Two The present study aimed at investigating the influence of exceptions were observed, only for cells cultured on rough ceramic. particles on osteoblasts and fibroblasts morphology and spreading. The first one concerned the production of CIH by fibroblasts after 10 Just after addition of alumina and zirconia in the cell culture media, days( Fig 3c, p<0.01)and the second concerned the production of we observed a slight particle aggregation in the cell culture fibronectin by osteoblasts after 6 days of culture( Fig. 3f, p<0.05). medium according to the granulometry studies discussed in para- Moreover, although not statistically significant, the production of the graph 2.1. As with the cells cultured without particles, cells extra-cellular matrix(ECM) was always greater in the presence of morphology and spreading appeared normal when cultured in the the polished ceramic than with the rough one For respective cells presence of alumina or zirconia for both fibroblasts and osteoblasts and substrates, these variations of the production of the ECM fol- up to 15 days of contact. After incubation with both alumina lowed the evolution of the proliferation zirconia powders, fibroblasts and osteoblasts quickly endocytosed an important proportion of particles. This internalization seems to 3. 2. In vitro cytocompatibility of ceramic particles start after a 2 h d appears clearly after 24 h(data not shown). Particles could be seen within and around the cells(the In order to assess the effect of the ceramic constituents, we also particles appears in black and are indicated by arrows in Fig. 6). The performed in vitro studies with particles. Fibroblasts and osteo- intracellular localization was especially peri-nuclear. lasts were cultured in the presence of alumina APA or zirconia tzO articles at two different concentrations(100 and 1000 ug/mL) 3.4. In vivo biocompatibility of alumina and zirconia particles luring 15 days. At 3, 6, 10 and 15 days, cell growth, CIH and FN synthesis were measured The rats well tolerated the particles or physiological fluid injection and ambulated without any pain or locomotion difficulties. Animals 3.2.1. Particles at 100 ug/mL showed no inflammation nor infection signs in all groups and all The first step of this study was the assessment of fibroblasts and along the 6 weeks of the study. These observations were confirmed by osteoblasts proliferation in the presence of alumina or zirconia autopsies. During dissection, the macroscopic examination revealed lease cite this article in press as: Roualdes O, et al, In vitro and in vivo evaluation of an., Biomaterials(2009), doi: 10.1016/ j biomaterials 2009. 11.107
alteration observed in our conditions. Moreover, the present limited three points scale was very reproducible. The histological assessment was made in two times. At first, we performed a general histological evaluation of the whole synovial tissue including the synovial lining surface and the underlining stroma using 5 power field. The synovial membrane thickness and the villous hyperplasia were scored. Thereafter, using 40 power field we examined separately the synovial lining cell layer and the sub-synovial tissue. Results presented were the mean of the scores observed in five continuous selected fields from the proximal region of the synovial tissue near the patella. In the synovial lining cell layer, we scored the cell layer hyperplasia (number of cell layers) and the cell hypertrophy (0 ¼ normal flat cells, 1 ¼ cuboı¨dal cells, 2 ¼ cylindrical cytoplasmic extensions). In the sub-synovial region, the cell infiltration was evaluated by scoring the amount of macrophages/ histiocytes cells (0 ¼ no or very low, 1 ¼ moderate and 2 ¼ high amount) and lymphocytes (0 ¼ no or very low, 1 ¼ moderate/diffuse infiltration and 2 ¼ high infiltration or aggregates). Finally, vascularisation was scored (0 ¼ normal, 1 ¼ moderate increase and 2 ¼ high increase). 2.8. Statistical analysis Cell proliferation, human fibronectin and type I collagen levels and in vivo data were analyzed statistically using a standard unparametric analysis of variance techniques (ANOVA Kruskall walis) following Dun’s multiple range test or Pearson (r0 ) and Spearman (r) correlation tests using GraphPad Prism version 4.02 software for Macintosh (GraphPad Software, San Diego, CA, USA). Statistical significance was considered at p < 0.05. All in vitro experiments were run in triplicate and repeated at least 3 times. Data presented on graphs or in the text are means SEM. 3. Results 3.1. In vitro cytocompatibility of ceramic discs Compared to cells cultured on plastic (reference substrate represented by an empty well), fibroblasts and osteoblasts proliferation decreased on both rough and polished ceramics discs tested in this study at all time periods even if the differences were not always statistically significant (Fig. 3 a–b). The proliferation of fibroblasts was significantly decreased on rough discs after 3 and 6 day of culture compared to the control (Fig. 3a, p < 0.01). For the osteoblasts, the proliferation was significantly lower on rough ceramics discs after 3, 6 and 10 days of culture (respectively p < 0.01, 0.05 and p < 0.001) and on polished discs after 3 and 10 days (respectively p < 0.05 and p < 0.01) than control (Fig. 3b). The cell functions were assessed by the assays of two important proteins of the extra-cellular matrix produced by fibroblasts and osteoblasts: human type I collagen and fibronectin (Fig. 3c–f). Compared to respective control cultures, the levels of the two proteins were generally weaker but not significantly decreased for both fibroblasts and osteoblasts. Two exceptions were observed, only for cells cultured on rough ceramic. The first one concerned the production of CIH by fibroblasts after 10 days (Fig. 3c, p < 0.01) and the second concerned the production of fibronectin by osteoblasts after 6 days of culture (Fig. 3f, p < 0.05). Moreover, although not statistically significant, the production of the extra-cellular matrix (ECM) was always greater in the presence of the polished ceramic than with the rough one. For respective cells and substrates, these variations of the production of the ECM followed the evolution of the proliferation. 3.2. In vitro cytocompatibility of ceramic particles In order to assess the effect of the ceramic constituents, we also performed in vitro studies with particles. Fibroblasts and osteoblasts were cultured in the presence of alumina APA or zirconia TZ0 particles at two different concentrations (100 and 1000 mg/mL) during 15 days. At 3, 6, 10 and 15 days, cell growth, CIH and FN synthesis were measured. 3.2.1. Particles at 100 mg/mL The first step of this study was the assessment of fibroblasts and osteoblasts proliferation in the presence of alumina or zirconia particles at 100 mg/mL in DMEM. Results are presented in the Fig. 4. From 3 to 15 days of culture, we observed a regular increase of the cell growth for both types of cell. The cell proliferation was not affected by the presence of alumina or zirconia particles. Indeed, no significant differences were observed in the cell proliferation between cultures with or without particles. The study of the cell viability by blue trypan exclusion confirmed this result (data not shown). In order to assess the cell ability to produce ECM, all DMEM supernatants were collected for CIH and FN assays. Results are presented in the Fig. 4c–f. Both CIH or FN productions and cell growth increased in comparable way. Compared to the control culture (without particles), cells cultured in the presence of alumina or zirconia at concentration of 100 mg/mL did not affect these two proteins synthesis. Results were similar for fibroblasts and osteoblasts from 3 to 15 days of culture. 3.2.2. Particles at 1000 mg/mL In order to evaluate the influence of the particles concentration on cell functions, we reproduced accurately all this experiment with ten fold more particles (1000 mg/mL). Due to this particles amount and compared to the cell number and the cell culture surface, we quickly observed an important particle layer that recovered all cells. This concentration was an extreme condition for an in vitro study. However, fibroblasts and osteoblasts growth were not really affected by this particles amount. Results are presented in the Fig. 5. Compared with control culture, we observed just a light increase of the fibroblast growth when cultured in the presence of alumina APA (p < 0.05) and zirconia TZ-0 (p < 0.05) after 15 days. This result was associated with an increase of the fibronectin synthesis in the presence of both alumina or zirconia particles at the same culture period (respectively p < 0.01 and p < 0.001 compared with control culture). However, no significant variation of collagen synthesis was detected. In the presence of the same particle amount, osteoblasts growth was also not affected by the presence of alumina or zirconia after 3, 6 and 10 days of culture. It was just slightly reduced at day 15 (p < 0.01) for both alumina and zirconia. At this time, this variation was associated with a light but not significant decrease of CIH synthesis and increase of FN. Regarding the results of cell growth and ECM synthesis in our study conditions, alumina and zirconia particles showed no cytotoxic effect. 3.3. Cell morphology in the presence of ceramic particles The present study aimed at investigating the influence of particles on osteoblasts and fibroblasts morphology and spreading. Just after addition of alumina and zirconia in the cell culture media, we observed a slight particle aggregation in the cell culture medium according to the granulometry studies discussed in paragraph 2.1. As with the cells cultured without particles, cells morphology and spreading appeared normal when cultured in the presence of alumina or zirconia for both fibroblasts and osteoblasts, up to 15 days of contact. After incubation with both alumina and zirconia powders, fibroblasts and osteoblasts quickly endocytosed an important proportion of particles. This internalization seems to start after a 2 h contact, and appears clearly after 24 h (data not shown). Particles could be seen within and around the cells (the particles appears in black and are indicated by arrows in Fig. 6). The intracellular localization was especially peri-nuclear. 3.4. In vivo biocompatibility of alumina and zirconia particles The rats well tolerated the particles or physiological fluid injection and ambulated without any pain or locomotion difficulties. Animals showed no inflammation nor infection signs in all groups and all along the 6 weeks of the study. These observations were confirmed by autopsies. During dissection, the macroscopic examination revealed 4 O. Roualdes et al. / Biomaterials xxx (2009) 1–12 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107
ARTICLE IN PRESS O. Roualdes et al. Biomaterials xxx(2009)1-12 a10 3.5 3.0 oE 左02 0.5 0.0 Days c6.5 d45 5.5 盟5.0 4.5 3.0 4.0 332 2.0 510 1.0 函 Days f7.5 10 55 7 5544 86 635 3.0 2.5 2.0 505 自爵自 Fig 3. Fibroblasts(a c, e)and osteoblasts(b, d, f) were cultured for 3, 6 and 10 days on rough C or polished (2 ceramic and on plastic(control E under standard cell culture conditions in 24-well culture plates At each time of culture, cell proliferation was assessed using the mit assay (a, b)and culture media were harvested for human type I collagen(c. d)and human fibronectin(e f) synthesis quantification Values are expressed as mean+SEM, n=9: +p<0.05, +tp<0.01, +++p<0.001(compared to control cultures a normal knee articulation associated with a normal articular capsul The histological appearances of the synovial membrane were However, a fine and granular deposit was observed around the patella normal at low magnification (5x). For control and particles for some animals that received particles. It certainly injected groups, no alteration of the synovial membrane sponded of the accumulation of the powder in the tissues thickness or hyperplasia could be observed Please cite this article in press as: Roualdes 0, et al, In vitro and in vivo evaluation of an, Biomaterials(2009). doi: 10.1016/ j biomaterials 2009. 11.107
a normal knee articulation associated with a normal articular capsule. However, a fine and granular deposit was observed around the patella for some animals that received ceramic particles. It certainly corresponded of the accumulation of the powder in the tissues. The histological appearances of the synovial membrane were normal at low magnification (5). For control and particles injected groups, no alteration of the synovial membrane thickness or hyperplasia could be observed. 3 6 10 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 ++ ++ a Days 3 6 10 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 ++ +++ + + ++ b Days Cell proliferation (optical density) Cell proliferation (optical density) Human type I collagen (µg/mL) Human type I collagen (µg/mL) Human fibronectin (µg/mL) Human fibronectin (µg/mL) 3 6 10 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 ++ c Days 3 6 10 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Days d 3 6 10 0 1 2 3 4 5 6 7 8 9 10 11 e Days 3 6 10 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 + f Days Fig. 3. Fibroblasts (a, c, e) and osteoblasts (b, d, f) were cultured for 3, 6 and 10 days on rough ( ) or polished ( ) ceramic and on plastic (control, ) under standard cell culture conditions in 24-well culture plates. At each time of culture, cell proliferation was assessed using the MTT assay (a, b) and culture media were harvested for human type I collagen (c, d) and human fibronectin (e, f) synthesis quantification. Values are expressed as mean SEM, n ¼ 9; þp < 0.05, þþp < 0.01, þþþp < 0.001 (compared to control cultures). O. Roualdes et al. / Biomaterials xxx (2009) 1–12 5 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107
ARTICLE IN PRESS 1.0 0.9 2.0 15 80.1 0.0 6 6 Days d 5.0 4.0 2.5 .0 6 15 10 321 9是5E 09876543 510 目 broblasts (a, c e)an d centrifuged in order to eliminate residual particles Supernatant were used for human type I expressed as mean+SEM, n=9 for culture in the presence of particles and n= 12 for controls: *p< 0.05. *p<0.01(compared to control cultures). At higher magnification(high power field: 40x ) we clearly order to avoid confusion between the visualization of particles and observed the presence of particles in the synovial membrane for all blood deposition, the presence of particles was confirmed with two animals, which were injected with alumina or zirconia(Fig. 7). In different colorations (Goldner trichrome and toluidin blue). As lease cite this article in press as: Roualdes O, et al, In vitro and in vivo evaluation of an., Biomaterials(2009), doi: 10.1016/ j biomaterials 2009. 11.107
At higher magnification (high power field: 40), we clearly observed the presence of particles in the synovial membrane for all animals, which were injected with alumina or zirconia (Fig. 7). In order to avoid confusion between the visualization of particles and blood deposition, the presence of particles was confirmed with two different colorations (Goldner trichrome and toluidin blue). As 3 6 10 15 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 a Days 3 6 10 15 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 b Days 3 6 10 15 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 + c Days 3 6 10 15 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 d 5.0 Days Days 3 6 10 15 0 5 10 15 20 25 30 35 ++ e Days 3 6 10 15 0 1 2 3 4 5 6 7 8 9 10 11 12 13 f Cell proliferation (optical density) Cell proliferation (optical density) Human type I collagen (µg/mL) Human type I collagen (µg/mL) Human fibronectin (µg/mL) Human fibronectin (µg/mL) Fig. 4. Fibroblasts (a, c, e) and osteoblasts (b, d, f) were cultured for 3, 6, 10 and 15 days in the presence of particles (100 mg/mL) of alumina ( ), zirconia ( ) or without particles ( ) under standard cell culture conditions in 24-well culture plates. At each time of culture, cell proliferation was assessed using the MTT assay (a, b) and culture media were harvested and centrifuged in order to eliminate residual particles. Supernatant were used for human type I collagen (c, d) and human fibronectin (e, f) synthesis quantification. Values are expressed as mean SEM, n ¼ 9 for culture in the presence of particles and n ¼ 12 for controls; þp < 0.05, þþp < 0.01 (compared to control cultures). 6 O. Roualdes et al. / Biomaterials xxx (2009) 1–12 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107
ARTICLE IN PRESS 09) 111 b5 54433 050 075 0 左 0 左 0 ys 111 654 766 1.3 0 2 5.5 110 1 5.0 品8- 3. 5 0 000000 50 0 0.0 10 15 Days 443 E5 01血图象自 10 15 6 Fibroblasts (a, c e)and osteoblasts(b, d, f) were cultured for 3, 6, 10 and 15 days in the presence of particles (1000 ug/mL)of conditions in 24-well culture ssay(a, b) sted and centrifuged in order to eliminate residual particles. are expressed as mean* SEM, n=9 for culture in the presence of particles and n= 12 for controls: *p<0.05, ++p<0.01, +++p<0.001(compared to control cultures). described previously in the cell culture medium, alumina and anatomical region. Most of them were found near the patella zirconia particles looked agglomerated (paragraph 3.3). The particular in the folds of the synovial membrane. Our histological amount of particles observed was different depending on the evaluation score exhibited that both alumina and zirconia particles Please cite this article in press as: Roualdes 0, et al, In vitro and in vivo evaluation of an, Biomaterials(2009). doi: 10.1016/ j biomaterials 2009. 11.107
described previously in the cell culture medium, alumina and zirconia particles looked agglomerated (paragraph 3.3). The amount of particles observed was different depending on the anatomical region. Most of them were found near the patella, in particular in the folds of the synovial membrane. Our histological evaluation score exhibited that both alumina and zirconia particles 3 6 10 15 0.00 0.25 0.50 0.75 1.00 1.25 1.50 + +++ + + Days 3 6 10 15 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 ++++ a b Days 3 6 10 15 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 c Days 3 6 10 15 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 + + d Days 3 6 10 15 0 5 10 15 20 25 30 35 40 45 50 55 ++ +++++ e Days 3 6 10 15 0 1 2 3 4 5 6 7 8 9 10 11 f Days Cell proliferation (optical density) Human type I collagen (µg/mL) Human fibronectin (µg/mL) Cell proliferation (optical density) Human type I collagen (µg/mL) Human fibronectin (µg/mL) Fig. 5. Fibroblasts (a, c, e) and osteoblasts (b, d, f) were cultured for 3, 6, 10 and 15 days in the presence of particles (1000 mg/mL) of alumina ( ), zirconia ( ) or without particles ( ) under standard cell culture conditions in 24-well culture plates. At each time of culture, cell proliferation was assessed using the MTT assay (a, b) and culture media were harvested and centrifuged in order to eliminate residual particles. Supernatant were used for human type I collagen (c, d) and human fibronectin (e, f) synthesis quantification. Values are expressed as mean SEM, n ¼ 9 for culture in the presence of particles and n ¼ 12 for controls; þp < 0.05, þþp < 0.01, þþþp < 0.001 (compared to control cultures). O. Roualdes et al. / Biomaterials xxx (2009) 1–12 7 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107
ARTICLE IN PRESS O. Roualdes et al/ Biomaterials xox(2009)1-12 产 b d olemented with 10% FBS, 1% penicillin/streptomycin) for 24 h At that time, medium was replaced hie由ga media containing either alumina (a, b), zirconia(c, d)or no particles(control: e, f) and cultured for 72 h under standard conditions. After this period, cells were with Coomassie blue brilliant in order to observe cell morphology and spreading (original magnification= 40x, arrows indicate particles localization and bars indicate 10 um). gere more present in the intracellular compartment (respectively correlation r=0.5151, p< 0.05). All synovial lining cell reactions 42+0.07 and 3. 23+0. 42)than extra-cellularly (respectively observed were similar for both alumina and zirconia powders. 2.49+0.15 and 1.84+1.13). Thereafter, we compared the scores of In a second time, we evaluated the sub-synovial tissue reat total particles amount in the synovial lining cell layers and in the tion Rats who received particles showed a significantly increase ub-synovial tissue. Results presented in Fig. 8 showed a signifi- of the macrophages, histiocytes and foreign body giant cells cantly higher repartition of both powders in the synovial lining cell populations(Fig. 10). Most of these cells contained an important layers than in sub-synovial tissue (p< 0.001). This amount intra-cytoplasmic particle amount. The scores of number of decreased quickly in the depth of the sub-synovial tissu macrophages and of the total particles amounts were positively The injection of particles led to some cellular and tissue modi- correlated in the sub-synovial tissue, all injected animal being fications of the synovial membrane. At first, we analyzed the taken together (Spearman correlation r=0.5105, p< 0.05 and synovial lining cell reaction. It was characterized by a slight Pearson correlation r= 0.5414, p<0.01). However, like in the and 9). Indeed, animals injected with particles showed a synovial lymphocytes infiltration and no fibrosis Vascularisation appeared ning cell composed with more cell layers(p< 0.01)and the cells to be slightly but not significantly increased( Fig. 10)but there was ere bigger or more cuboidal (p < 0.001 and p<0.05 for alumina a positive correlation between the vascularisation score and the nd zirconia respectively)compared to the control group(Fig 9). macrophage amount(Spearman correlation r=0.5413, p<0.001 The addition of the scores of cell layer hyperplasia and cell hyper- and Pearson correlation r=0.4845, p<0.001). More impo trophia constituted an adequate score for the evaluation of the there was a stronger correlation between the score of the synovial lining cell reaction. In this study, it was increased for lining cell reaction and the score of macrophage animals that received alumina or zirconia powders (p<0.001 and (Spearman correlation r=0.7435, p<0.001 and Pearson 0.01 respectively) and positively correlated with the score tion r=0.7522, p<0.001). All this results were in accordance corresponding to the total particle amount found in the synovial with a moderate non-specific granulomatous response of the lining cell(Spearman correlation r=0.5200, p< 0.01 and Pearson synovial tissue. lease cite this article in press as: Roualdes O, et al, In vitro and in vivo evaluation of an., Biomaterials(2009), doi: 10.1016/ j biomaterials 2009. 11.107
were more present in the intracellular compartment (respectively 3.42 0.07 and 3.23 0.42) than extra-cellularly (respectively 2.49 0.15 and 1.84 1.13). Thereafter, we compared the scores of total particles amount in the synovial lining cell layers and in the sub-synovial tissue. Results presented in Fig. 8 showed a signifi- cantly higher repartition of both powders in the synovial lining cell layers than in sub-synovial tissue (p < 0.001). This amount decreased quickly in the depth of the sub-synovial tissue. The injection of particles led to some cellular and tissue modi- fications of the synovial membrane. At first, we analyzed the synovial lining cell reaction. It was characterized by a slight increase of the cell layer hyperplasia and cell hypertrophia (Figs. 7 and 9). Indeed, animals injected with particles showed a synovial lining cell composed with more cell layers (p < 0.01) and the cells were bigger or more cuboidal (p < 0.001 and p < 0.05 for alumina and zirconia respectively) compared to the control group (Fig. 9). The addition of the scores of cell layer hyperplasia and cell hypertrophia constituted an adequate score for the evaluation of the synovial lining cell reaction. In this study, it was increased for animals that received alumina or zirconia powders (p < 0.001 and p < 0.01 respectively) and positively correlated with the score corresponding to the total particle amount found in the synovial lining cell (Spearman correlation r ¼ 0.5200, p < 0.01 and Pearson correlation r0 ¼ 0.5151, p < 0.05). All synovial lining cell reactions observed were similar for both alumina and zirconia powders. In a second time, we evaluated the sub-synovial tissue reaction. Rats who received particles showed a significantly increase of the macrophages, histiocytes and foreign body giant cells populations (Fig. 10). Most of these cells contained an important intra-cytoplasmic particle amount. The scores of number of macrophages and of the total particles amounts were positively correlated in the sub-synovial tissue, all injected animal being taken together (Spearman correlation r ¼ 0.5105, p < 0.05 and Pearson correlation r0 ¼ 0.5414, p < 0.01). However, like in the control group, animals injected with ceramic particles showed no lymphocytes infiltration and no fibrosis. Vascularisation appeared to be slightly but not significantly increased (Fig. 10) but there was a positive correlation between the vascularisation score and the macrophage amount (Spearman correlation r ¼ 0.5413, p < 0.001 and Pearson correlation r0 ¼ 0.4845, p < 0.001). More importantly, there was a stronger correlation between the score of the synovial lining cell reaction and the score of macrophage amount (Spearman correlation r ¼ 0.7435, p < 0.001 and Pearson correlation r0 ¼ 0.7522, p < 0.001). All this results were in accordance with a moderate non-specific granulomatous response of the synovial tissue. Fig. 6. Fibroblasts (a, c, e) and osteoblasts (b, d, f) were cultured in DMEM (supplemented with 10% FBS, 1% penicillin/streptomycin) for 24 h. At that time, medium was replaced with fresh culture media containing either alumina (a, b), zirconia (c, d) or no particles (control: e, f) and cultured for 72 h under standard conditions. After this period, cells were fixed and stained with Coomassie blue brilliant in order to observe cell morphology and spreading (original magnification ¼ 40, arrows indicate particles localization and bars indicate 10 mm). 8 O. Roualdes et al. / Biomaterials xxx (2009) 1–12 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107
ARTICLE IN PRESS O. Roualdes et al. Biomaterials xxx(2009)1-12 3.5 3.0 50 2.5 2.0 e1.0 0.5 Synovial lining cells Subsynovial tissue Fig 8. Comparison of the ng cell and sub-synovial tissue for alumina( group B, n= 12) and zirconia ( group C. n=12) Values are ed as mean+ SEM, *p<0.001 compared to the sar material in the sub-synovial compartment bALA to the sintering process itself. Thus, even if the material used in this study cannot be characterized as a nano-phase ceramic, it is a good picture of the new generation of alumina-zirconia composites currently developed and used in arthroplasty [10, 16, 18 Although their superiority in terms of mechanical strength and aging resis- tance has been clearly demonstrated [16, 46], there is lack of careful biological evaluation. This work tries to respond to this need in two parts. At first, we focalized on the in vitro cytocompatibility of the alumina-zirconia sintered ceramic assessment Then we evaluated the in vitro and in vivo biological influence of it alumina and zirconia constitutive powders as a model of wear debris. When osteoblasts were cultured upon a nano-structured sin- tered alumina ceramic, Webster et al. showed that cell adhesion and proliferation increased compared with conventional ceramics. However, it was the opposite for fibroblasts [22-24. If osteoblasts are bone forming cells, fibroblasts can be considered as cells that contribute to fibrous encapsulation and callus formation, events that may lead to loosening failure. These results and this theory may explain the good osteointegrative potential of the material. In Fig. 7. Histological es(stained with Goldner the present study, we only used micron-structured sintered synovial lining cell alumina-zirconia ceramic. Then we were not comparing the nano (c)represent section of animals that were injected three time respectively with and conventional structure of ceramics on cell comportment but lumina or zirconia particles and showed a slight reaction of the synovial lining cell rather to observing the difference between the response of osteo- layer and macrophage infiltration. blasts and fibroblasts, depending on the surface of ceramics. After cell seeding on the material, the adhesion of cells were indirectly Discussion nvestigated by comparing the cell proliferation upon and beside the discs by mrt assay which is representative of the proportion of The present study consisted of the evaluation of the biological cells that have not been fixed ompatibility of an alumina-zirconia composite. If the cyto- studies, we did not observe any modification or difference of the compatibility of conventional ceramics (micro-structured and adhesion for both fibroblasts and osteoblasts(data not shown). The monolithic composed) has been clearly demonstrated there is less micro-structured ceramic discs, with polished or rough surface. information about their nano-structured and or composite version. appear as surfaces adapted for cell promoting Here, we investigated on an alumina-zirconia ceramic processed In fact, this composite only slightly reduced cell proliferation of om nano-powders. Although starting from nano-phases powders both osteoblasts and fibroblasts, even if it was statistically significan (40 and 350 nm for zirconia and alumina respectively). sintering for osteoblasts but not for fibroblasts. The cell functions, represented mechanisms leaded to a sub-micron microstructure(600 nm to by CIH and fn productions appeared also accurate for both cell types I um for zirconia and alumina respectively ) which is obvious due In the conditions of our study, this material showed a satisfactory Please cite this article in press as: Roualdes 0, et al, In vitro and in vivo evaluation of an, Biomaterials(2009). doi: 10.1016/ j biomaterials 2009. 11.107
4. Discussion The present study consisted of the evaluation of the biological compatibility of an alumina–zirconia composite. If the cytocompatibility of conventional ceramics (micro-structured and monolithic composed) has been clearly demonstrated, there is less information about their nano-structured and/or composite version. Here, we investigated on an alumina–zirconia ceramic processed from nano-powders. Although starting from nano-phases powders (40 and 350 nm for zirconia and alumina respectively), sintering mechanisms leaded to a sub-micron microstructure (600 nm to 1 mm for zirconia and alumina respectively), which is obvious due to the sintering process itself. Thus, even if the material used in this study cannot be characterized as a nano-phase ceramic, it is a good picture of the new generation of alumina–zirconia composites currently developed and used in arthroplasty [10,16,18]. Although their superiority in terms of mechanical strength and aging resistance has been clearly demonstrated [16,46], there is lack of careful biological evaluation. This work tries to respond to this need in two parts. At first, we focalized on the in vitro cytocompatibility of the alumina–zirconia sintered ceramic assessment. Then, we evaluated the in vitro and in vivo biological influence of it alumina and zirconia constitutive powders as a model of wear debris. When osteoblasts were cultured upon a nano-structured sintered alumina ceramic, Webster et al. showed that cell adhesion and proliferation increased compared with conventional ceramics. However, it was the opposite for fibroblasts [22–24]. If osteoblasts are bone forming cells, fibroblasts can be considered as cells that contribute to fibrous encapsulation and callus formation, events that may lead to loosening failure. These results and this theory may explain the good osteointegrative potential of the material. In the present study, we only used micron-structured sintered alumina–zirconia ceramic. Then we were not comparing the nano and conventional structure of ceramics on cell comportment but rather to observing the difference between the response of osteoblasts and fibroblasts, depending on the surface of ceramics. After cell seeding on the material, the adhesion of cells were indirectly investigated by comparing the cell proliferation upon and beside the discs by MTT assay which is representative of the proportion of cells that have not been fixed on the material. Contrary to Webster studies, we did not observe any modification or difference of the adhesion for both fibroblasts and osteoblasts (data not shown). The micro-structured ceramic discs, with polished or rough surface, appear as surfaces adapted for cell promoting. In fact, this composite only slightly reduced cell proliferation of both osteoblasts and fibroblasts, even if it was statistically significant for osteoblasts but not for fibroblasts. The cell functions, represented by CIH and FN productions appeared also accurate for both cell types. In the conditions of our study, this material showed a satisfactory Fig. 7. Histological appearances of rat synovial membranes (stained with Goldner trichrome, magnification 40). (a) The control section showed a synovial lining cell layer composed by one layer of flat synoviocytes and no sign of inflammation. (b) and (c) represent section of animals that were injected three time respectively with alumina or zirconia particles and showed a slight reaction of the synovial lining cell layer and macrophage infiltration. Synovial lining cells Score of total particles amount Sub-synovial tissue 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 +++ +++ Fig. 8. Comparison of the scores of total particle amount between synovial lining cell and sub-synovial tissue for alumina ( , group B, n ¼ 12) and zirconia ( , group C, n ¼ 12). Values are expressed as mean SEM, þþþp < 0.001 compared to the same material in the sub-synovial compartment. O. Roualdes et al. / Biomaterials xxx (2009) 1–12 9 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107
ARTICLE IN PRESS 1.75 1.50 0.50 0.00 Cell layer hyperplasia Cell hypertrophy Synovial linging cell reaction Fig 9. The evaluation of the synovial lining cell reaction was based on the addition of two individual parameters: cell hyperplasia and hypertrophia of the synovial lining cell. Results are presented for alumina( group B, n= 12). zirconia (2 group C, n=12)and for the control group ( E n=10). Values are expressed as mean+ SEM, p<0.05, cytocompatibility. Apart the control, the polished surface more device. According to several studies, bone cell functions may be preserved cell functions and proliferation than the roughness one, influenced by particles size, chemistry and shape 27,33 ven if the difference between these two materials was not statis- 35, 58,61,62] The clinical success of a new alumina-zirconia ceramic tically significant. This may confirm the real influence of the surface lies partly to the low production of wear particles. Because the topography on cell behavior described in many works (48, 56, 57 amount of particles we obtained after tribologic test was too low for If the cytocompatibility of the composite is the key element to be an in vitro evaluation, we decided to replace them by alumina and checked in order to use it in an implant, it is also known that wear zirconia constitutive powders. Germain et al. reported different debris induce bone loss and is an important cause of aseptic loos- toxicities for alumina wear debris and commercially or constituen ening of the hip prosthesis [27, 58-61. Thus, biological evaluation of particles if their size and volume are different [63]. Here, we have not these wear debris is needed to predict the long-term reliability of the verified these settings but we think that it is a reliable model especially in our condition of study. Indeed we tested in vitro twe powders of alumina and zirconia at two different concentrations with two cell types on many parameters like proliferation, viability 1.6 data not shown and ECm production. The concentrations of particles(100 and 1000 ug/mL) were chosen in accordance with 1.4 those used by gutwein in previous works [21, 38. Moreover, the results of tribological tests made us think they are both very 12 important compared to the amount that should be observed in clinical use. In parallel, the in vivo study came to reinforce the quality 1.0蕨 of this approach. If there were any significant toxicity, we would have underscored it. The present study has shown that alumina and zirconia constitutive powders at 100 ug/mL have no influence either 0.8 on the cell proliferation and viability(data not shown), or on the roduction of extra-cellular matrix. this was verified for both oste- oblasts and fibroblasts. similar results were observed with a concentration of 1000 ug/mL up to 10 days of culture. After this time, proliferation of osteoblasts decreased while that of fibroblasts 0.4 increased. These alterations were very light and associated only with an increased of the production of fibronectin for both cell types. This 0.2 is in agreement with literature reports of the very small detriment observed [21, 27, 38, 64,65. Furthermore, contrary to other works Macrophages Vascularisation about metal and polymer particles, we have not observed any gnificant decrease Fig- 10. Score of the evaluation of sub-synovial tissue reaction Macrophage and vas- 28, 29, 32, 66. Moreover, we have not observed any differences in acaeasetionswstre e anucted ed wings a degrees ed os maea n sEnt of ive eate cellular response in the presence of either alumina or zirconia valuated, n=12 for animals injected with alumina () or zirconia (50) and n=10 for powders. Their effect seems to depend on their concentration but the control group g and +++p<0.001 compared to the control for the same not on their chemical nature. The present study thus demonstrates the excellent in vitro cytocompatibility of the constitutive particles lease cite this article in press as: Roualdes O, et al, In vitro and in vivo evaluation of an., Biomaterials(2009), doi: 10.1016/ j biomaterials 2009. 11.107
cytocompatibility. Apart the control, the polished surface more preserved cell functions and proliferation than the roughness one, even if the difference between these two materials was not statistically significant. This may confirm the real influence of the surface topography on cell behavior described in many works [48,56,57]. If the cytocompatibility of the composite is the key element to be checked in order to use it in an implant, it is also known that wear debris induce bone loss and is an important cause of aseptic loosening of the hip prosthesis [27,58–61]. Thus, biological evaluation of these wear debris is needed to predict the long-term reliability of the device. According to several studies, bone cell functions may be influenced by particles size, chemistry and shape [27,33– 35,58,61,62]. The clinical success of a new alumina–zirconia ceramic lies partly to the low production of wear particles. Because the amount of particles we obtained after tribologic test was too low for an in vitro evaluation, we decided to replace them by alumina and zirconia constitutive powders. Germain et al. reported different toxicities for alumina wear debris and commercially or constituent particles if their size and volume are different [63]. Here, we have not verified these settings but we think that it is a reliable model, especially in our condition of study. Indeed, we tested in vitro two powders of alumina and zirconia at two different concentrations with two cell types on many parameters like proliferation, viability (data not shown) and ECM production. The concentrations of particles (100 and 1000 mg/mL) were chosen in accordance with those used by Gutwein in previous works [21,38]. Moreover, the results of tribological tests made us think they are both very important compared to the amount that should be observed in clinical use. In parallel, the in vivo study came to reinforce the quality of this approach. If there were any significant toxicity, we would have underscored it. The present study has shown that alumina and zirconia constitutive powders at 100 mg/mL have no influence either on the cell proliferation and viability (data not shown), or on the production of extra-cellular matrix. This was verified for both osteoblasts and fibroblasts. Similar results were observed with a concentration of 1000 mg/mL up to 10 days of culture. After this time, proliferation of osteoblasts decreased while that of fibroblasts increased. These alterations were very light and associated only with an increased of the production of fibronectin for both cell types. This is in agreement with literature reports of the very small detrimental effect of nanosized particles despite the important endocytosis observed [21,27,38,64,65]. Furthermore, contrary to other works about metal and polymer particles, we have not observed any significant decrease on the production of type I collagen [28,29,32,66]. Moreover, we have not observed any differences in cellular response in the presence of either alumina or zirconia powders. Their effect seems to depend on their concentration but not on their chemical nature. The present study thus demonstrates the excellent in vitro cytocompatibility of the constitutive particles. Cell layer hyperplasia Score Cell hypertrophy Synovial linging cell reaction 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 + +++ ++ ++ +++ +++ Fig. 9. The evaluation of the synovial lining cell reaction was based on the addition of two individual parameters: cell hyperplasia and hypertrophia of the synovial lining cell. Results are presented for alumina ( , group B, n ¼ 12), zirconia ( , group C, n ¼ 12) and for the control group ( , n ¼ 10). Values are expressed as mean SEM, þp < 0,05, þþp < 0,01, þþþp < 0.001 compared to control group for the same parameter. Macrophages Score Vascularisation 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 +++ +++ Fig. 10. Score of the evaluation of sub-synovial tissue reaction. Macrophage and vascularisation were evaluated following 3 degrees: 0 ¼ normal, 1 ¼ slight or moderate increase, 2 ¼ strong increased. Values are expressed as mean SEM of five fields evaluated, n ¼ 12 for animals injected with alumina ( ) or zirconia ( ) and n ¼ 10 for the control group ( ) and þþþp < 0.001 compared to the control for the same parameter. 10 O. Roualdes et al. / Biomaterials xxx (2009) 1–12 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107