Evaluation of the in vitro cell-material interactions and in vivo osteo-integration of a spinal acrylic bone cement

Abstract

Introduction

Polymethylmethacrylate bone cements have proven performance in arthroplasty and represent a common bone filler, e.g. in vertebroplasty. However, acrylic cements are still subject to controversy concerning their exothermic reaction and osteo-integration potential. Therefore, we submitted a highly filled acrylic cement to a systematic investigation on the cell-material and tissue-implant response in vitro and in vivo.

Materials and methods

Cured Vertecem V+ Cements were characterized by electron microscopy. Human bone marrow-derived mesenchymal stem cell morphology, growth and differentiation on the cured cement were followed for 28 days in vitro. The uncured cement was injected in an ovine cancellous bone defect and analysed 4 and 26 weeks post-implantation.

Results

The rough surface of the cement allowed for good stem cells adhesion in vitro. Up-regulation of alkaline phosphatase was detected after 8 days of incubation. No adverse local effects were observed macroscopically and microscopically following 4 and 26 weeks of implantation of the cement into drill-hole defects in ovine distal femoral epiphysis. Direct bone apposition onto the implant surface was observed resulting in extended signs of osteo-integration over time (35.2 ± 24.2% and 88.8 ± 8.8% at week 4 and 26, respectively).

Conclusion

Contrary to the established opinion concerning bony tissue response to implanted acrylic bone cements, we observed an early cell-implant in vitro interaction leading to cell growth and differentiation and significant signs of osteo-integration for this acrylic cement using standardized methods. Few outlined limitations, such as the use of low cement volumes, have to be considered in the interpretation of the study results.

Introduction

Acrylic bone cements (Polymethylmetacrylates, PMMAs) have been in clinical use, especially for arthroplasty, since 1960 [1]. They are the most commonly used bone filler material for percutaneous vertebroplasty (PVP) and percutaneous kyphoplasty (PKP) since 1987 and 1998, respectively, and more recently for other cancellous bone augmentations [2]. The clinical safety and efficacy of acrylic bone cements has been demonstrated previously. In addition to handling and outstanding mechanical properties, PMMA used in PVP or PKP has been shown to induce immediate and lasting pain relief in 80–90% of the cases [3, 4]. However, the use of acrylic bone cements present disadvantages including high polymerization temperature [5], neurotoxicity of the monomer [6], lack of osteointegration [7] due to their bioinert nature [3] leading to fibrous encapsulation [8, 9]. While mechanical factors were primarily considered responsible for poor response in the past, the biological reaction of the surrounding tissues to the implanted PMMA is currently taken into more consideration [10, 11]. Hermann et al. [12] showed the presence of pseudomembranous fibrous tissue at the bone-implant interface. Mechanical stability, leading to long term stability of the implant, results from bone formation and remodelling at the direct implant-bone interface which leads to implant’s osteo-integration. Osteo-integration is driven by a complex and multi-step process, involving osteogenic cells and their precursors [13]. In vivo, mesenchymal stem cells (MSCs) migrate and attach to the implant, where they will differentiate toward an osteoblastic phenotype able to secrete and mineralize their own extracellular matrix [14]. Osteo-integration is influenced by the implant surface [15] features, as well as by the presence of bio-active components (such as hydroxyapatite or bioglasses) added to the cement [16]. Concerning the osteo-integration of PMMAs, studies showed partial bone attachment to such cements [17]. A recent case report [18] described a large quantity of new bone formation at the interface of the PMMA implant, 3.5 years post-implantation. A similar post-mortem report [19] showed viable bone close to the implanted acrylic cement suggesting bone remodelling.

Due to the controversial reports concerning PMMA cements osteo-integration capacity, we submitted an acrylic spinal bone cement to a systematic investigation of the in vitro cytocompatibility (cell adhesion, cell morphology, cell proliferation) and in vivo cell-material and tissue-implant response.

Materials and methods

Cement preparation and sample preparation

All experiments were conducted using commercial PMMA cement (Vertecem V + Cement Kit, LOT 09CA53010, Synthes GmbH, Oberdorf, Switzerland). It is a radiopaque acrylic bone cement with a medium viscosity for use in percutaneous vertebroplasty. The polymer powder contains 40 wt.% Zirconium dioxide (ZrO₂) as radio-opaque agent and 15 wt.% hydroxyapatite (HA).

PMMA was prepared at room temperature according to the manufacturer’s instructions. The mixed bone cement was then filled into PTFE molds (3 mm deep × 30 mm diameter) and stored under water until complete curing. Samples were then removed from the molds, individually packed in PE/paper bags, and steam sterilized.

For the animal study, the bone cement was used directly after preparation in its pasty state. After filling in 1 ml syringes, the cement was extruded through a 14 Ga needle into the prepared cylindrical bone-defect in a time period of 2–7 min after starting preparation.

Characterization of cement sample surface

The microstructure of the cement surfaces was characterized by scanning electron microscopy (SEM) (Zeiss Evo 60 EP-SEM, Carl Zeiss AG, Switzerland). Cement samples were sputter-coated with gold (BAL-TEC SCD 50 Sputter coater, Oerlikon-Balzers, Liechtenstein) and images were recorded using the secondary electron detector under high vacuum (30 Pa) and an acceleration of 15 kV. Energy dispersive X-ray spectroscopy (EDX) measurements were performed to identify the chemical components in the cement.

Surface roughness (average roughness R_a and maximum roughness R_m) of the cement samples were investigated quantitatively using Contour measurement system (ConturoMatic T1, Q.P.T. GmbH, Germany) according to ISO 4287.

In vitro evaluation

Bone Marrow derived MSC (BMSC) isolation and culture

Human bone marrow (BM) samples were obtained after informed consent of patients (KEK Bern 126/03). After BM homogenization, BMSC were separated on a Ficoll gradient. Interphases containing mononucleated cells were collected and grown in Iscove Modified Dulbecco’s Medium (IMDM, Invitrogen) containing 10% Fetal Bovine Serum (FBS), 1% nonessential amino acids, 100 U/ml PenStrep (all Gibco) and 5 ng/mL basic-FGF [20]. BMSC were used at passage 3, and medium was changed every 3 days.

Cell seeding on the tested surfaces

Prior to cell seeding, the cement samples were rinsed in IMDM for 2 h to remove packaging debris and possible particles release, and dried overnight in a sterile hood. 10,000 cells/cm² in a volume of 250 μL were seeded on each surface and incubated at 37°C and 5% CO₂. After 2 h, 4 ml of osteogenic medium (IMDM, 10% FBS, 10⁻⁸ M dexamethasone, 0.1 mM ascorbic acid and 10 mM sodium β-glycerophosphate) was added to each well.

Cell morphology

For SEM analysis, all cell fixation procedures were carried out at room temperature and in 0.1 M Piperazine-N-N′-bis-(2-ethane sulphonic acid) buffer (PIPES) (Sigma-Aldrich, Switzerland). After washing with PIPES (pH 7.4), samples were fixed in 2.5% glurataldehyde for 5 min. Cell contrasting was obtained through 1 h incubation in 1% osmium tetroxide in PIPES (pH 6.8) (Simec Trade AG, Zofingen, Switzerland). Samples were dehydrated through increased ethanol gradient bath (from 50 to 100%) and critical point dried (CPD) with a Polaron E3100 (Quorum Technologies, East Sussex, UK). Samples were sputter-coated with a 12 nm layer of gold/palladium (80/20) and observed at 5 kV accelerating voltage and 40 μA emission current in a field emission SEM (FESEM S-4700, Hitachi, Japan) with a secondary electron detector and Quartz PCI digital imaging system.

For immunocytochemistry analysis, cryopreserved MSC (Lonza) were seeded at 50,000 cells/sample and incubated overnight at 37°C and 5% CO₂. Samples were washed with Phosphate Buffered Saline (PBS), fixed with 4% paraformaldehyde, and permeabilized with 0.05% Triton X100. Cells were stained using DAPI (Invitrogen) and Alexa Flour Phalloidin 546 (Invitrogen) to highlight the cell nucleus and cytoskeleton, respectively. Samples were imaged using a Zeiss 510 NLO 510 Confocal microscope (Delaware Biotechnology Institute, Newark, DE) equipped with a multiphoton laser at 743 nm for excitation of the DAPI and reflection of the cement surface and a HeNe laser with an excitation wavelength of 543 nm for the phalloidin. For all samples, images were collected with an optical slice ≤2 μm in the z-axis and complied to show a full representation of the cells on the cement surface.

Cell growth

Cell growth was determined via DNA quantification as described by Labarca [21]. After overnight digestion in proteinase K and dilution of the samples in Dulbecco’s phosphate buffered saline (DPBS) containing 0.1% (v/v) Hoechst 33258 (Polysciences Inc, 09460), fluorescence was measured using a PE HTS 7000 Bio Assay Reader at 360 nm excitation and 465 nm emission wavelength.

Quantification of alkaline phosphatase activity

Alkaline phosphatase (ALP) activity was measured according to the Sigma Technical Bulletin Procedure No.104. Cells were lysed using 0.1% Triton-X 10 mM Tris-HCl (pH 7.4) for (2 h, 4°C) and incubated with p-nitrophenyl phosphate substrate (Sigma Kit No.104) for 15 min at 37°C. The p-nitrophenol production was measured at 405 nm.

Alizarin red staining

Cells were fixed in 70% ethanol (12 h, 4°C), washed with purified and deionised water, and incubated in 2 ml of 40 mM Alizarin Red solution (ARS) (1 h, RT). After ARS removal, intensive washes in purified and deionised water and PBS, samples were air dried and imaged with a digital camera.

Gene expression analysis

Total RNA was extracted using TRI-Reagent (MRC Inc. TR-118) according to the manufacturer’s instructions. Reverse transcription was performed on 1 μg of total RNA sample, using TaqMan reverse transcription reagents (Applied Biosystems, Foster City, CA, USA) with random hexamer primers. Real-time Polymerase Chain Reaction (PCR) was performed on a 7500 Real Time PCR System (Applied Biosystems). Genes of interest were detected using specific oligonucleotide primers and TaqMan probes (Microsynth, Switzerland) or Assays on Demand (Applied Biosystems, Foster City, CA, USA) (Table 1), in presence of TaqMan Universal PCR master mix (Applied Biosystems). The PCR was performed using GeneAmp 5700 Sequence Detection System (Applied Biosystems, Foster City, CA, USA). Relative quantification of mRNA targets was performed according to the comparative C_T method with 18S ribosomal RNA as endogenous control (ABI PRISM 7700 Sequence Detector User Bulletin, PE Applied Biosystems 1997).

Table 1.

Real-time PCR assays

18 s	# 4310893E
ALP	Hs00758162_m1
Cbfa1	fw AGC AAG GTT CAA CGA TCT GAG AT rv TTT GTG AAG ACG GTT ATG GTC AA pr TGA AAC TCT TGC CTC GTC CAC TCC G
BSPII	fw TGC CTT GAG CCT GCT TCC rv GCA AAA TTA AAG CAG TCT TCA TTT TG pr CTC CAG GAC TGC CAG AGG AAG CAA TCA
BMP2	fw AAC ACT GTG CGC AGC TTC C rv CTC CGG GTT GTT TTC CCA C Pr CCA TGA AGA ATC TTT GGA AGA ACT ACC AGA AAC TG

In vivo evaluation

This non-clinical investigation was performed by BIOMATECH (Chasse-sur-Rhône, France) following FDA (FDA 21CFR58) and EU (EU 2004/10/EC) regulations under GMP conditions. Further biological evaluations of the bone cement were performed according to ISO10993 standard (Part 2, Part 6).

Animal model

Eight adult Blanche du Massif Central Sheep (BMC) (≥24 month, skeletally mature) weighing between 58 and 71 kg were obtained from GAEC (du Noisetier-Cordesse-15 260 Neuveglise, France). The animal study was approved by the internal ethical committee from BIOMATECH and followed NIH Publication 82-83(1986) and EEC 86/609 directives. The eight sheep were implanted uni-laterally with the test cement for 4 weeks (n = 4) and 26 weeks (n = 4).

After general anaesthesia, the distal femoral epiphysis was surgically approached on the medial side. In each sheep, three cylindrical holes (5 mm in diameter, 10 mm deep) were created in the distal femoral epiphysis under constant irrigation (0.9% NaCl, Versol) and suction. The defects were completely filled with the cement. The incision was then closed by suturing capsule, muscles and subcutaneous layer with absorbable thread (PDSII^®, Vicryl^®, Ethicon). The sheep were observed daily during the whole experimental period.

At 4 and 26 weeks, the designated sheep were sacrified by an injection of pentobarbital (Dolethal^®, Vetoquinol), bones of interest were isolated and surrounding soft tissues were removed for further sample preparation.

Histological preparation

After fixation in 10% neutral buffered formalin, each sample was transversally divided in two equal parts (A and B), dehydrated in graded series of ethanol, and embedded in polymethylmetacrylate (PMMA, Merck, Hohenbrunn, Germany) (part A) or in SPURR resin (Hatfield, PA, USA) (part B). For each part, one longitudinal ground-section of approximately 20 μm thickness was prepared per site (Exakt microcutting) as described by Donath [22]. A modified polychromatic Paragon staining was applied to the sections. The PMMA embedded slides were reserved for histopathology evaluation whereas the SPURR resin slides were used for the histomorphometrical analysis.

Histopathology evaluation

Qualitative and semi-quantitative histological evaluations of the local tissue reaction were performed on 12 parameters for each sample. Signs of inflammation (e.g. presence of macrophages), encapsulation, presence of osteoblastic cells, signs of osteointegration (direct bone–implant contact), osteoconduction (presence of new bone growth onto/into the material), bone neoformation, bone remodeling, neovascularisation and cement degradation were evaluated and graded as follow: 0: not detected, 1: slight evidence, 2: moderate evidence, 3: marked evidence, 4: strong evidence.

Histomorphometric analysis

Histomorphometric analyses were conducted in a blinded manner on the SPURR resin slides using a Zeiss Axioscope microscope equipped with a colour image analyzing system (v.4.27 Samba^®; Samba Technologies, France). Bone contact, relative bone area, fibrous tissue contact, relative fibrous tissue area, relative implant area and the presence of implant within the bone lacunae were quantified within two standardized representative areas in each section (2 mm wide symmetrically located at the bone-cement interface).

Statistical analysis

Statistical analysis was performed on all in vitro data using the pair-wise non-parametric Mann–Whitney test corrected by Bonferroni. Histomorphometrical parameters were analyzed using the Levene test followed by the corrected ANOVA Welch test. SPSS 15.0 (SPSS Inc., USA) was used, and differences were considered significant when p < 0.05 (*).

Results

Cement characteristics

Representative images of the cement surface are shown in Fig. 1. The sample surface appears rough and porous (Fig. 1a). Quantitative analyses of the surface topography revealed values for R_a of 5.2 ± 1.5 μm and for R_m of 37.8 ± 8.9 μm (mean ± SD). At higher magnification (Fig. 1b) it is clearly visible that almost all filler particles (HA and ZrO₂) were embedded/covered by newly polymerized MMA and only a low portion remains apparent at the surface.

Fig. 1.

SEM images of acrylic bone cement. a Overview of the general cement topography. b Higher magnification present all cement ingredients (PMMA, HA and ZrO₂), mainly embedded by polymerized MMA

In vitro evaluation

Cell morphology

Scanning electron microscopy observation revealed a good adhesion of BMSC onto the cement surface. After 2 days of culture, single cells were still visible, presenting a well spread, elongated spindle shaped morphology, with large cell-surface interaction, and the presence of thin pseudopodia (Fig. 2a). The cytoskeleton, as shown by actin immunostaining, is extended over the surface features of the cement (Fig. 2b). As shown in Fig. 2c and d, the cells reached confluence already after 8 days in culture.

Fig. 2.

Cell morphology. After 2 days of culture (a, b). After 21 days in culture (c, d). SEM image (a, c), cytoskeleton immunostaining (b, d)

Cell growth

After 24 h, there was not statistically significant difference in the number of cells (Fig. 3) on plastic compared to cement (p > 0.05). From day 8 of culture however, the number of cells was always found higher on the cell-culture-treated-plastic surface compared to the cement (p < 0.05).

Fig. 3.

Cell growth. DNA quantification at different time points

Cell differentiation

Cell differentiation process was analysed by real-time RT-PCR (Fig. 4a) and ALP enzymatic activity (Fig. 4b). Results shown in Fig. 4a are expressed relative to endogenous 18 s control. Cells on control and cement did not show expression or upregulation of BMP2 gene with time. The amount of ALP detected for at day 1 and day 8 was low, but a significant up-regulation was observed on both surfaces by day 18. The same trend was observed for BSPII; however, due to variability within the sample set, a statistical significance in values was not obtained. The transcription factor cbfa1 was well detected from day 1 of the cell culture on both surfaces and increased with time. None of the tested genes showed any significant differences between the cement and the plastic control surface. For the ALP enzymatic activity (Fig. 4b), more ALP expression was detected with cells on plastic compared to the cement. The ALP level increased over time on both surfaces. All together, these observations indicate that BMSC differentiate on cement similarly to control surfaces.

Fig. 4.

Cell differentiation. a Osteoblastic maker genes expression relative to endogenous 18S control. b ALP activity relative to DNA content

Matrix mineralization

Alizarin Red staining (Fig. 5) confirmed the presence of mineralised nodules on the cement (dark red) at day 18, similar to the plastic (sharp red) control. The number and the size of these nodules increased even more at day 28.

Fig. 5.

Matrix mineralization. Alizarin Red staining as performed at day 18 and day 28 of culture on plastic control or cement surfaces

In vivo evaluation

Throughout this study, sheeps did not indicate any signs of pain. Post euthanasia, histopathological analysis of the lymph nodes showed no signs of adverse effects of the implants (data not shown).

Histopathological evaluation

Polychromatic Paragon staining was performed on the non-decalcified sections at 4 weeks (Fig. 6a, c, e) and 26 weeks (Fig. 6b, d, f) post-implantation. Low magnification images (Fig. 6a, c) showed the presence of active woven bone formation at the direct contact of the implant at 4 weeks post-implantation. After 26 weeks, the bone trabecules showed some thickening and marked signs of bone remodelling at the direct contact of the implant surface (Fig. 6b, f). High magnification images showed the presence of active osteoblasts and areas of osteoconduction at both 4 and 26 weeks (Fig. 6e, f). In Fig. 6c–f, MMA embedding induced the dissolution of the cement; only some remnants were still detectable at the bone-implant interface. No signs of local adverse effects were observed around the implant at two healing times. Results reported in Table 2 summarize the general histological observations that were made on the parameters described.

Fig. 6.

Histology staining as performed at 4 weeks (a, c, e) and at 26 weeks (b, d, f) post implantation. SPURR embedded samples (a, b). MMA embeddings (c–f). I bone cement implant, BT bone tissue, RM remaining bone cement, OsteoC osteoconduction, BR bone remodeling

Table 2.

Semi-quantitative histopathological evaluation (mean score) (scoring scale: 0 not detected, 1 slight evidence, 2 moderate evidence, 3 marked evidence, 4 strong evidence)

Parameters	Time points
	4 weeks (n = 12 sites)	26 weeks (n = 12 sites)
Encapsulation	2	1
Inflammation (macrophages)	1	1
Osteoblastic cells	2	1
Osteointegration	2	3
Osteoconduction	3	3
Bone neoformation	2	3
Remodeling	0	3
Neovascularisation	2	2
Particulate diffusion	0	0

Histomorphometrical analysis

Histomorphometry results presented in Table 3 indicated a statistical significant increase of the bone-implant contact percent between week 4 (35.2 ± 24.2%) and week 26 (88.8 ± 8.8%), while a statistical significant decrease was observed concerning the fibrous tissue related parameter. The osteo-integration of the implant over time reflected a satisfactory level of performance. The relative implant area remained unchanged, indicating a strong stability of the cement (no swelling, plasticity, degradation) over time.

Table 3.

Histomorphometrical analysis

Parameters	Time points (%)
	4 weeks (n = 12 sites)	26 weeks (n = 12 sites)
Bone contact	35.2* ± 24.2	88.8 ± 8.8
Relative bone area	36.0 ± 6.6	43.8 ± 8.7
Fibrous tissue contact	64.8* ± 24.2	11.2 ± 8.8
Relative fibrous tissue area	42.6 ± 8.4	29.8 ± 9.2
Relative implant area	21.4 ± 7.5	26.4 ± 7.7
Implant within bone lacunae	1.3 ± 1.4	4.5 ± 3.3

Results are presented as mean and standard deviation in percentage as obtained from the three samples of each animal

* p < 0.01

Discussion

Osteo-integration of a permanent implant to surrounding bone tissue, like PMMA in arthroplasty or spinal augmentation procedures, is a crucial step for its long-term clinical success. Despite positive clinical outcomes, acrylic cements are the subject of controversy concerning their exothermic reaction [5] and their potential for osteo-integration. In the first part of the present study, we investigated the cell/materials interactions in vitro, and in a second part, we investigated the in vivo osteointegration capacity of a HA containing PMMA bone cement.

To address the in vitro performance of the commercial bone cement, we used BMSC. Since BMSC will colonize the implant and differentiate to osteoblasts in vivo, they represent a relevant cell type for in vitro evaluation of bone implants. The cells settle down on the surface of the material then integrin-dependent adhesion occurs; if appropriate conditions are met, they can further differentiate and mineralize. The initial phases of cell adhesion are, therefore, determinant for subsequent cell behaviour (cell growth, differentiation, matrix synthesis, etc.) and ultimately the implant osteo-integration.

In our study, we compared the BMSC behaviour for three donors on the PMMA cement surfaces in parallel with cell-culture-plastic surface as positive controls. On both surfaces, cells showed similar and reproducible behaviours.

On both surfaces, DNA content increased over time. Many in vitro studies support the hypothesis that the surface roughness and the surface chemistry in general may have a direct influence on how cells will attach, migrate, proliferate and differentiate on biomaterials [23–25]. In the present work, the cement topography showed some R_a and R_m values of about 5 and 30 μm, respectively. Previous studies have shown that surface topography features at the cell scale affect the behaviour of cells [26]. For example, looking at the BMSC behaviour on rough titanium surfaces compared to smoother ones with identical surface chemistry, Deligianni [27] found an increased cell adhesion, a slightly higher proliferation and similar ALP activity on rougher surfaces. Lampin et al. [28] demonstrated that the surface energy of apolar PMMA components significantly increases with roughness. Therefore, the relative high roughness of our test article might partly explain our promising results on cell adhesion, differentiation and in vivo osteo-integration. The latter assumes that the roughness of the implanted cement samples is similar to the in vitro samples cured under water.

However, the surface topography is not the only factor influencing cell behaviour, the material composition is another key aspect [29]. The presence of bio-active elements at the surface of the implant, such as bioglasses, strontium, or hydroxyapatite particle has been shown to increase the cell adhesion, focal contact formation and phenotypic differentiation to PMMA cement [30–32] in vitro but also in vivo. Therefore, HA and ZrO₂ particles, even if most of them were embedded, may have aided in attracting and promoting BMSC adhesion to the surface. Four weeks post implantation; good signs of implant osteo-integration were observed, increasing after 26 weeks. Some fibrous encapsulation was detected at 4 weeks that decreased slightly at week 26.

No cell death was detected during our in vitro-study. Indeed, using cured and resterilised materials any exothermic reaction is eliminated, and the release of superficial monomers—which could induce a local inflammation and cell death—is reduced [33, 34]. The main reason for the use of cured materials for in vitro investigations is the much better reproducibility of experimental conditions like sample surface dimensions compared to uncured materials. In addition, cured materials are relevant for the long term implantation period of non-resorbable materials. Hence, in order to mimic the real clinical application of the materials, in our in vivo study, the cement was injected into the defect in a pasty stage according to well established and standardized methods (ISO 10993 Part 6). Taken together the analyses presented here showed consistency of the overall very positive results between in vitro cytocompatibility and in vivo bio-compatibility, even if some presence of fibrous tissues layer was detected. This could also be attributed to the initial stability of the implant since even micro motion of an implanted material will result in the formation of a fibrous tissue around the implant [35, 36]. The comparability between the motion applied in the femoral condyle (our experimental setting) and in a vertebral body (in case of vetebroplasty) might constitute a limitation of the animal model used in this study [37]. Furthermore, the implanted cement volume and location used in our study was different in comparison to typical augmentation regions (e.g. vertebral body, femural head) and limits the direct transferability to the clinical situation. Especially, the relative low cement volume results in fewer amounts of monomers and heat released during setting. These limitations should be considered in the interpretation of the current results and might be addressed in further studies.

In the light of cancellous bone augmentation procedures, the handling and mechanical properties of PMMA bone cements are superior to other materials such as calcium phosphate cements (CPC) [38, 39]. However, polymeric compounds cannot compete with pure ceramics on bioactivity [10, 11, 17, 40], but using CPCs other drawbacks such as cement failure [38] or particle/ion release could lead to tremendous clinical risks like embolism [41]. The present study demonstrated, that special acrylic cement compositions can combine the good handling properties of common PMMA with a good degree of osteo-integration in vivo and positive effects in cell culture studies, where cell apposition and bone forming potential was investigated. Therefore, implant improvements must consider the whole spectrum of properties of the implant rather than optimizing a single behaviour.

Conclusion

Contrary to the established opinion concerning the lack of bioactivity of a PMMA cement, our study, under the outlined conditions, showed an early cell-implant in vitro interaction leading to cell growth and differentiation followed by clear signs of osteo-integration when implanted in vivo.