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<4.5) which allowed good dispersion On the other hand, commercial alumina samples[19]. In an other work, He et al. also Zeta potential was decreasing 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. 11.107be 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 proper￾ties. 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) alumina particles have less detrimental effect on osteoblasts function and adhesion than other composed with conventional particles (>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 influ￾enced 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 phos￾phatase [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 osteo￾blasts [32–37] but there are less about the in vitro effect of nano￾sized 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 biocom￾patibility of an alumina–zirconia ceramic composite for ortho￾paedic applications, especially for total hip arthroplasty. Here, we present an in vitro and in vivo approach for the biological assess￾ment 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 < 4.5), which allowed good dispersion. On the other hand, Zeta potential was decreasing 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 electro￾static 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