Morphological characterization and in vitro biocompatibility of a porous nickel–titanium alloy

ARTICLE IN PRESS

Biomaterials 26 (2005) 5801–5807 www.elsevier.com/locate/biomaterials

Morphological characterization and in vitro biocompatibility of a porous nickel–titanium alloy Oleg Prymaka, Denise Bogdanskib, Manfred Ko¨llerb, Stefan A. Esenweinb, Gert Muhrb, Felix Beckmannc, Tilmann Donathc, Michel Assadd, Matthias Epplea, a

Institute for Inorganic Chemistry, University of Duisburg-Essen, Universitaetsstr. 5-7, D-45117 Essen, Germany b Department of Surgery, BG Kliniken Bergmannsheil, D-44789 Bochum, Germany c GKSS-Research Center Geesthacht, Institute for Materials Research, Max-Planck-Strasse 1, D-21502 Geesthacht, Germany d Centre for Bone and Periodontal Research, McGill University, 740 Dr. Penfield Avenue, Rm. 2200, Montreal, QC H3A 1A4, Canada Received 17 February 2005; accepted 21 February 2005 Available online 26 April 2005

Abstract Disks consisting of macroporous nickel–titanium alloy (NiTi, Nitinols, Actipores) are used as implants in clinical surgery, e.g. for fixation of spinal dysfunctions. The morphological properties were studied by scanning electron microscopy (SEM) and by synchrotron radiation-based microtomography (SRmCT). The composition was studied by X-ray diffractometry (XRD), differential scanning calorimetry (DSC), and energy-dispersive X-ray spectroscopy (EDX). The mechanical properties were studied with temperature-dependent dynamical mechanical analysis (DMA). Studies on the biocompatibility were performed by co-incubation of porous NiTi samples with isolated peripheral blood leukocyte fractions (polymorphonuclear neutrophil granulocytes, PMN; peripheral blood mononuclear leukocytes, PBMC) in comparison with control cultures without NiTi samples. The cell adherence to the NiTi surface was analyzed by fluorescence microscopy and scanning electron microscopy. The activation of adherent leukocytes was analyzed by measurement of the released cytokines using enzyme-linked immunosorbent assay (ELISA). The cytokine response of PMN (analyzed by the release of IL-1ra and IL-8) was not significantly different between cell cultures with or without NiTi. There was a significant increase in the release of IL-1ra (po0:001), IL-6 (po0:05), and IL-8 (po0:05) from PBMC in the presence of NiTi samples. In contrast, the release of TNF-a by PBMC was not significantly elevated in the presence of NiTi. IL-2 was released from PBMC only in the range of the lower detection limit in all cell cultures. The material, clearly macroporous with an interconnecting porosity, consists of NiTi (martensite; monoclinic, and austenite; cubic) with small impurities of NiTi2 and possibly NiCx. The material is not superelastic upon manual compression and shows a good biocompatibility. r 2005 Elsevier Ltd. All rights reserved. Keywords: Shape memory alloys; Biocompatibility; Cytokines; Leukocytes; Porosity; Nickel–titanium alloys

1. Introduction Nickel–titanium alloys (NiTi; Nitinols) have gained access to current medical technology due to their extraordinary mechanical properties, namely, the superelasticity and the shape memory effect [1]. In particular, orthodontic wires and intravascular stents are often Corresponding author.

E-mail address: [email protected] (M. Epple). 0142-9612/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2005.02.029

made of superelastic NiTi. In trauma surgery, NiTi has found only few applications so far despite the very useful shape memory effect when it comes to fixation of, e.g., bone fractures. Staples for foot surgery are on the market that make use of this effect [1–3]. Intramedullary nails have been proposed and tested [4]. For spinal surgery, porous implants are used as interbody fusion device without the need for an additional bone graft material (Actipores) [1,5]. These materials are prepared by a self-heated combustion synthesis (SHS) process,

ARTICLE IN PRESS 5802

O. Prymak et al. / Biomaterials 26 (2005) 5801–5807

where nickel and titanium powders are intimately mixed and the exothermic reaction occurs by heating the mixture with a tungsten wire [6,7]. This leads to an interconnecting macroporosity that is advantageous for the ingrowth of bone cells at the implantation site. It promotes long-term fixation and intervertebral fusion without the need for bone grafting. The biocompatibility of such porous NiTi implants was thoroughly studied both in vitro and in vivo [5]. Porous NiTi was an excellent osteoconductive material for lumbar interbody fusion in sheep and triggered an osteoid front penetration from the surrounding cancellous bone towards the center of the implant. The behavior of the porous NiTi device was similar to that of a titanium cage. No significant difference in distraction capacities between both materials regardless of time was found [8]. Concerns about an adverse reaction due to a release of nickel have not been confirmed [9–11]. NiTi demonstrated good resistance to in vivo surface corrosion and also to nickel ion release. In vivo, it compared very well to a titanium implant in a 12-month sheep study [12]. We have studied the material in terms of its morphology, composition, mechanical properties, and cell biological responses in order to gain a full view of this material for an extended clinical use.

2. Materials and methods Disks of macroporous NiTi (Actipores) from selfpropagating high-temperature synthesis (SHS) [7] were obtained from Biorthex, Montreal, QC, Canada, and used as obtained. Their diameter was either 7 or 10 mm, and the height was 2 mm. The average porosity was 6575% and average pore diameter was 2307130 mm, both according to the manufacturer. Scanning electron microscopy (SEM) and energydispersive X-ray spectrometry were carried out with a LEO 1530 Gemini instrument. For metallic specimens, no sputtering was required. Samples from cell culture experiments were sputtered with gold. X-ray diffractometry (XRD) was carried out with a Siemens D5000 system with CuKa radiation (1.54178 A˚; Bragg–Brentano geometry; Al sample holders) on mechanically flattened samples which were, however, correctly positioned in the diffractometer. Differential scanning calorimetry was carried out with a Netzsch DSC 204 instrument with the following temperature program: 150 to
100 1C with
10 K min
1, followed by
100 to 150 1C with +10 K min
1. The typical sample size was 65 mg (sealed aluminium crucibles). Dynamical mechanical analysis (DMA) was carried out with a Netzsch DMA 242 C4 instrument on a porous disc of 2.02 mm height and 6.97 mm diameter

under air in a dynamical compression experiment. The diameter of the flat circular indenter was 3 mm. The applied force was oscillating sinusoidally between 2.32 and 11.64 N, the frequency was 5 Hz, the heating rate was 1 K min
1, and the temperature range was
50 to +150 1C. Synchrotron radiation-based microtomography (SRmCT) was carried out at beamline W2 at HARWI at HASYLAB/DESY at 54 keV in the range of 1–1801 with steps of 0.251 [13,14]. The sample size was 2  2  2 mm3, taken from an intact sample by cutting. The resolution is of the order of a few micrometers in each voxel dimension. Three-dimensional reconstruction work was done at HASYLAB. Images were created with the program VG Studio MAX 1.2. The cell biological responses were studied in vitro using equally sized test disks (10 mm diameter, 2 mm thickness) of porous NiTi. Polymorphonuclear neutrophil leukocytes (PMN) and peripheral blood mononuclear cells (PBMC) were isolated by a single-step procedure based on a discontinuous double Ficoll-gradient as described earlier [15,16]. This method led to more than 95% pure and viable PMN or PBMC. Isolated cells were adjusted to 1  106 cells ml
1 in RPMI1640 cell culture medium (Invitrogen, Eggenstein, Germany) supplemented with L-glutamine (0.3 g l
1) and sodium bicarbonate (2.0 g l
1), 10% fetal calf serum (FCS, Invitrogen), and 20 mM N-(2-hydroxyethyl)-piperazineN0 -(2-ethanesulfonic acid) (Sigma-Aldrich). Cell counting was performed using Tuerk staining solution (Sigma-Aldrich). Cell viability was measured by the trypan blue (Sigma-Aldrich) exclusion test. Isolated PMN or PBMC (1  106 cells ml
1 in supplemented RPMI 1640, see above) were added to each well of a 24-well cell culture plate with or without porous NiTi samples and incubated for 24 h using cell culture conditions (37 1C, 5% CO2, humidified atmosphere). Subsequently, the cell culture supernatants were harvested and centrifuged at 2000g for 2 min at room temperature (Biofuge-Pico, Eppendorf, Hamburg, Germany) to eliminate cells and particulate material. Cell adherence and cell viability were qualitatively analyzed scanning electron microscopy and fluorescence microscopy as described earlier [15,16]. For fluorescence analysis the samples were carefully replaced from the culture plates and washed three times with RPMI1640. Subsequently, adherent cells were stained with 14 mg ml
1 Calcein-AM (Calbiochem-Novabiochem, Bad Soden, Germany) and 50 mg ml
1 propidium iodide (Molecular Probes Inc., Eugene, OR, USA). After a twofold washing with RPMI1640, the stained cells were microscopically photographed using a fluorescence microscope (Photomicroscope 3, Zeiss, Oberkochem, Germany) and a digital camera (Camedia C3030,

ARTICLE IN PRESS O. Prymak et al. / Biomaterials 26 (2005) 5801–5807

Olympus, Hamburg, Germany). The images were processed using AnalySIS 3.2 software (Soft Imaging System, Mu¨nster, Germany) and Photoshop 5.0 software (Adobe, UnterschleiXheim, Germany). Cytokine concentrations (IL-1ra, IL-2, IL-6, IL-8, TNF-a) in cell culture supernatants were quantified by enzyme-linked immunosorbant assay (ELISA) as described previously [17] (IL ¼ interleukine; ra ¼ receptor antagonist). Antibodies as well as recombinant human cytokines (standards) were supplied by R&D Systems (Wiesbaden, Germany). The sensitivities of the immunoassays were 25 pg ml
1 (IL-1ra), 10 pg ml
1 (IL-2), 0.6 pg ml
1 (IL-6), 10 pg ml
1 (IL-8), and 5 pg ml
1 (TNF-a). Differences in cytokine releases in the presence or absence of NiTi samples were statistically calculated using Student’s t-test.

5803

3. Results and discussion The destruction-free method of SRmCT clearly shows that the resulting biomaterials possess pore size diameters of 2307130 mm that fall well in the conventional 50–500 mm range for tissue integration (Fig. 1) [14]. The surface structure can be studied with scanning electron microscopy in higher resolution (Fig. 2). The porous structure is also clearly indicated. In higher magnification, small crystallites of the order of a few micrometers are seen which may result from cutting (sample preparation for microscopic analysis). The surface of the cut parts is clearly different from that of the initial surface which solidified from the melt. Energy-dispersive X-ray spectrometry shows only the elements nickel and titanium (Fig. 3).

Fig. 1. Volume renderings in different magnification of a 2  2  2 mm3 section of a porous NiTi implant investigated by SRmCT. The interconnected pore system is shown. The scale bar refers to the picture at the left, and the pictures to the right represent magnifications.

Fig. 2. Representative SEM micrographs of the surface of NiTi implants. The porous structure is obvious (top pictures). There are, however, clear differences between cut surfaces (bottom left) and undisturbed surfaces from melt solidification.

ARTICLE IN PRESS O. Prymak et al. / Biomaterials 26 (2005) 5801–5807

5804

800 Ti intensity / a.u.

600 Ni

400

Ni

200 Ti

Ti

Ni

0 1

2

3

4

5 6 energy / keV

7

8

9

10

Fig. 3. Energy-dispersive X-ray spectrometry of the sample surface, showing only the elements Ni and Ti.

of porous NiTi implants [5,8,12,18], they are probably negligible. The thermal parameters of the sample were studied by differential scanning calorimetry (DSC; Fig. 5). The cooling curve shows that the material is in the transition region between austenite and martensite. However, the material is not superelastic upon manual deformation, i.e. upon compression or bending with hands (note that it is necessary to perform either compression or tension tests to clearly demonstrate the superelastic behavior). This is shown by DMA (Fig. 6) [19–21]. The elastic modulus (approximated as storage modulus) around room temperature was about 340 MPa. Kim et al. reported 120–200 MPa for similar samples [22]. This is of the order of the elastic modulus of cancellous bone (human vertebral trabecular bone: 56.8–77.9 MPa [23]).

0.20 30

0.10 DSC / mW mg-1

25

Intensity

20 15

NiTi2

NiCx (?)

NiCx (?)

62.5°C

0.15

NiTi (cubic)

NiCx (?)

10

M

A

0.05 0.00 -0.05 -0.10 -0.15 M

5

-0.20 -100

NiTi2 30

33

36

39 42 45 48 51 Diffraction angle / ˚2Θ

54

57

60

Fig. 4. X-ray powder diffractogram of compacted NiTi, showing the peaks of NiTi in its martensitic form. The peaks are slightly shifted due to the non-ideal diffraction geometry. The monoclinic NiTi (B19 ( phase) with the unit cell parameters a ¼ 2:885; b ¼ 4:622; c ¼ 4:120 A; b ¼ 96:81 is shown as vertical intensity bars on the bottom, all other ( phases are labelled directly (cubic NiTi; B2; a ¼ 2:998 A).

0 50 Temperature / °C

100

150

Fig. 5. Differential scanning calorimetry (DSC) shows the martensite–austenite transition upon heating (dotted line) at 62.5 1C (Ap) and the austenite–martensite transition upon cooling (solid line) at 22.8 1C (Mp). Note the directions of the phase transitions between M(artensite) and A(ustenite).

6.0

420

storage modulus ------- amplitude

Storage modulus / MPa

400

The crystallographic composition of the sample is accessible by XRD (Fig. 4). For strongly absorbing materials (like NiTi), this method can be performed only with flat samples, therefore we used a part of the porous material and pressed it between two steel plates to a flat, compact disk. This disk was then measured in reflection geometry on the diffractometer. The diffractogram shows mainly the peaks of monoclinic NiTi and cubic NiTi, in accordance with the work by Li et al. [7], and traces of other intermetallic phases. The minor impurities probably result from the special materials processing route, i.e., from the SHS process. Their influence on the microscopic corrosion and degradation properties is not known, but, given the good biological performance

-50

5.5

380

5.0

360 4.5 340 4.0

320 300

Amplitude / µm

0

A

22.8°C

3.5

280 -50

0

50 Temperature / °C

100

3.0 150

Fig. 6. Dynamical mechanical analysis (DMA) shows the martensite–austenite transition in the region of 60–80 1C (heating from
50 to +150 1C; oscillating force between 2.32 and 11.62 N, 5 Hz).

ARTICLE IN PRESS O. Prymak et al. / Biomaterials 26 (2005) 5801–5807

Fig. 7. Scanning electron microscopy of adherent peripheral blood leukocytes (PMN) at the surface of a porous NiTi sample. Isolated PMN were incubated with the NiTi sample.

5805

For cortical bone, values of elastic modulus of E ¼ 10220 GPa [24,25], 11.5–27 GPa (compact human bone [26]), and 20.175.4 GPa (diaphyseal cortical bone [27]) were found. For morsellized particles of bone, 85–135 MPa were found [28]. Solid (non-porous) NiTi has an elasticity modulus of E ¼ 30248 GPa [29–32] (for superelastic orthodontic wires), which is low compared to other standard metallic biomaterials: Ti-6Al-4 V alloy (E ¼ 115 GPa) [33], Ti (E ¼ 116 GPa), and stainless steel E ¼ 200 GPa: In most cases, medical implants are initially exposed to a blood-containing microenvironment. As it is shown in Figs. 7 and 8, cells immediately adhere to metallic implant surfaces which was also reported previously [34]. Furthermore, the vast majority of adherent cells was viable after 24 h cell culture (indicated by green fluorescence, Fig. 8), indicating a good biocompatibility of the porous material. Although the early interaction of leukocytes with biomaterial surfaces represents only one single aspect which governs the fate of an implant, early released mediators are intimately involved in wound healing, tissue remodelling, and local host defence [35,36]. Since IL-2 is mainly generated from activated lymphocytes [37] and IL-6 or IL-1ra are typically produced by myeloid leukocytes [37], the analysis of a selected pattern of released cytokines allows a conclusion which leukocyte subsets are preferentially activated. As shown in Table 1, there was a profound activation of monocytes (myeloids within the PBMC fraction) in the

Fig. 8. Fluorescence microscopy of adherent peripheral blood leukocytes (PMN) at the surface of a porous NiTi sample. Isolated PMN were incubated with the NiTi sample and stained with calcein-AM (green, living cells) and propidium iodide (red, dead cells). The inset represents a magnified image section; the white arrow indicates a single dead cell.

ARTICLE IN PRESS O. Prymak et al. / Biomaterials 26 (2005) 5801–5807

5806

Table 1 Cytokines (pg ml
1) released from PBMC after contact to porous NiTi in comparison to spontaneous cytokine release (control)

Control Porous NiTi

IL-1 ra

IL-6

IL-8

TNF-a

1687116 7587248**

1637228 187972107*

225371244 369371189*

1377 25716ns

Data represent average values 7standard deviation obtained from different experiments (n ¼ 5). Statistically significant increase compared to control * (po0.05); ** (po0.01); ns not significant. Table 2 Cytokines (pg ml
1) released from PMN after contact to porous NiTi in comparison to spontaneous cytokine release (control)

Control Porous NiTi

IL-1 ra

IL-8

99764 99748ns

134786 86733ns

Data represent average values 7standard deviation obtained from different experiments (n ¼ 3). ns Statistically not significantly different compared to control.

presence of porous NiTi samples, indicated by an increase in cytokine release compared to the controls without added test sample. This increased cytokine release in the presence of porous NiTi samples is obviously not related to the material itself because a comparable increase in cytokine production was observed in the presence of a porous tantalum biomaterial with similar porosity and pore size [38]. Lymphoid cells were not activated at the porous NiTi surface, and also a significant activation of PMN could not be demonstrated (Table 2). The reason for the different behavior of the two cell types is currently investigated. Whether the metallic nature of the porous material as a physical transition phase between austenite and martensite may influence these biological results is not known and may be the subject of future studies on NiTi with a clearly defined (and pure) solid phase.

4. Conclusions Macroporous NiTi from solid-combustion synthesis was tested with structural and biological tools. The material is macroporous, consists mainly of monoclinic NiTi and is not superelastic but shows a low modulus of elasticity, resembling that of cancellous bone which reduces the risk of stress shielding. The porous nature of this biomaterial should permit tissue/bone cell penetration and integration. In early implantation phases the porous material will be filled with blood and the large surface promote the adhesion of blood cells. Among them, myeloids will be predominately activated and will release inflammatory/regulatory mediators. For subse-

quent tissue integration, a fine balance between inflammatory signals which are necessary for the induction of tissue remodelling and a subsequent clearance of the inflammatory reactions is important (for reviews see Refs. [35–37]). At least, activation of myeloid leucocytes at porous metal implant surfaces will also provide a microenvironment of enhanced host defence as we have recently shown [38]. This is a natural part of the body’s reaction towards implants.

Acknowledgements This work was supported by a grant of the Deutsche Forschungsgemeinschaft (SFB 459: Shape Memory Technique).

References [1] Yahia L. Shape memory implants. Berlin: Springer; 2000. [2] Barouk LS. Osteotomies of the great toe. J Foot Surg 1992;31: 388–99. [3] Barouk LS. Scarf osteotomies in the hallux valgus. Personal experience. Therapeutic proposition. Foot Dis 1994;1:79–89. [4] Kujala S, Ryha¨nen J, Ja¨msa¨ T, Danilov A, Saaranen J, Pramila A, Tuukkanen J. Bone modeling controlled by a nickel–titanium shape memory alloy intramedullary nail. Biomaterials 2002;23: 2535–43. [5] Assad M, Likibi F, Jarzem P, Leroux MA, Coillard C, Rivard CH. Porous nitinol vs. titanium intervertebral fusion implants: computer tomography, radiological and histological study of osseointegration capacity. Mat-wiss u Werkstofftech 2004;35: 219–23. [6] Gunter V. Delay law and new class of materials and implants in medicine. STT/ Scientific & Technical Translations; 2000. [7] Li BY, Rong LJ, Li YY, Gjunter VE. Fabrication of cellular NiTi intermetallic compounds. J Mater Res 2000;15:10–3. [8] Assad M, Jarzem P, Leroux MA, Coillard C, Chernyshov AV, Charette S, Rivard CH. Porous titanium-nickel for intervertebral fusion in a sheep model: part 1. Histomorphometric and radiological analysis. J Biomed Mater Res Appl Biomater 2003;64B:107–20. [9] Kapanen A, Ryha¨nen J, Danilov A, Tuukkanen J. Effect of nickel–titanium shape memory metal alloy on bone formation. Biomaterials 2001;22:2475–80. [10] Kapanen A, Kinnunen A, Ryha¨nen J, Tuukkanen J. TGF-b1 secretion of ROS-17/2.8 cultures on NiTi implant material. Biomaterials 2002;23:3341–6. [11] Kapanen A, Danilov A, Lehenkari P, Ryha¨nen J, Ja¨msa¨ T, Tuukkanen J. Effect of metal alloy surface stresses on the viability of ROS-17/2.8 osteoblastic cells. Biomaterials 2002;23:3733–40. [12] Assad M, Chernyshov AV, Jarzem P, Leroux MA, Coillard C, Charette S, Rivard CH. Porous titanium–nickel for intervertebral fusion in a sheep model: part 2. Surface analysis and nickel release assessment. J Biomed Mater Res Appl Biomater 2003;64B:121–9. [13] Beckmann F, Bonse U, Biermann T. New developments in attenuation and phase-contrast microtomography using synchrotron radiation with low and high photon energies. Proc SPIE 1999;3772:179–87. [14] Tadic D, Beckmann F, Donath T, Epple M. Comparison of different methods for the preparation of porous bone substitution

ARTICLE IN PRESS O. Prymak et al. / Biomaterials 26 (2005) 5801–5807

[15]

[16]

[17]

[18]

[19] [20]

[21]

[22]

[23]

[24] [25]

materials and structural investigations by synchrotron m-computer tomography. Mat-wiss u Werkstofftech 2004;35:240–4. Choi J, Bogdanski D, Ko¨ller M, Esenwein SA, Mu¨ller D, Muhr G, Epple M. Calcium phosphate coating on nickel–titanium shape memory alloys. Coating procedure and adherence of leukocytes and platelets. Biomaterials 2003;24:3689–96. Bogdanski D, Esenwein SA, Prymak O, Epple M, Muhr G, Ko¨ller M. Inhibition of PMN apoptosis after adherence to dip-coated calcium phosphate surfaces on a NiTi shape memory alloy. Biomaterials 2004;25:4627–32. Bogdanski D, Epple M, Esenwein SA, Muhr G, Petzoldt V, Prymak O, Weinert K, Ko¨ller M. Biocompatibility of calcium phosphate-coated and of geometrically structured nickel–titanium (NiTi) by in-vitro testing methods. Mater Sci Eng A 2003;378:527–31. Assad M, Yahia LH, Rivard CH, Lemieux N. In vitro biocompatibility assessment of a nickel–titanium alloy using electron microscopy in situ end-labeling (EM-ISEL). J Biomed Mater Res 1998;41:154–61. van Humbeeck J. Cycling effects, fatigue and degradation of shape memory alloys. J Phys IV France 1991;1:189–97. Liu Y, van Humbeeck J, Stalmans R, Delaey L. Some aspects of the properties of NiTi shape memory alloys. J Alloys Compounds 1997;247:115. Liu Y, Xie ZL, van Humbeeck J, Delaey L. Asymmetry of stress–strain curves under tension and compression for NiTi shape memory alloys. Acta Mater 1998;46:4325. Kim JS, Kan JH, Kang SB, Yoon KS, Kwon YS. Porous TiNi biomaterial by self-propagating high-temperature synthesis. Adv Eng Mater 2004;6:403–6. Kopperdahl DL, Morgan EF, Keaveny TM. Quantitative computed tomography estimates of the mechanical properties of human vertebral trabecular bone. J Orthop Res 2002;20:801–5. Burnstein AH, Reilley DT, Martens M. Ageing of bone tissue: mechanical properties. J Bone Jt Surg 1976;58A:82–6. Swadener JG, Rho JY, Pharr GM. Effects of anisotropy on elastic moduli measured by nanoindentation in human tibial cortical bone. J Biomed Mater Res 2001;57:108–12.

5807

[26] Suchanek W, Yoshimura M. Processing and properties of hydroxyapatite-based biomaterials for use as hard tissue replacement implants. J Mater Res 1998;13:94–117. [27] Zysset PK, Guo XE, Hoffler CE, Moore KE, Goldstein SA. Elastic modulus and hardness of cortical and trabecular bone lamellae measured by nanoindentation in the human femur. J Biomechanics 1999;32:1005–12. [28] Verdonschot N, van Hal CTH, Schreurs BW, Buma P, Huiskes R, Slooff TJJH. Time-dependent mechanical properties of HA/TCP particles in relation to morsellized bone grafts for use in impaction grafting. J Biomed Mater Res Appl Biomater 2001;58:599–604. [29] Drake SR, Wayne DM, Powers JM. K.A. Mechanical properties of orthodontic wires in tension, bending, and torsion. Am J Orthod 1982;82:206–10. [30] Asgharnia MK, Brantley WA. Comparison of bending and tension tests for orthodontic wires. Am J Orthod 1986;89:228–36. [31] Kusy RP, Stush AM. Geometric and material properties of a nickel–titanium and a betatitanium orthodontic archwire alloy. Dent Mater 1987;3:207–17. [32] Prymak O, Klocke A, Kahl-Nieke B, Epple M. Fatigue of orthodontic nickel–titanium (NiTi) wires in different fluids under constant mechanical stress. Mater Sci Eng A 2003;378:110–4. [33] Davidson JA, Mishra AK, Kovacs P, Poggie RA. New surfacehardened, low-modulus, corrosion-resistant Ti–13Nb–13Zr alloy for total hip arthroplasty. Biomed Mater Eng 1994;4:231–43. [34] Eriksson C, Lausmaa J, Nygren H. Interactions between human whole blood and modified TiO2-surfaces: influence of surface topography and oxide thickness on leukocyte adhesion and activation. Biomaterials 2001;22:1987–96. [35] Horowitz MC. The role of cytokines in bone remodeling. J Clin Dent 1998;1:187–98. [36] Jimenez PA, Jimenez SE. Tissue and cellular approaches to wound repair. Am J Surg 2004;187:56S–64S. [37] Thompson A. The cytokine handbook. 3rd ed. San Diego: Academic Press; 1998. [38] Ko¨ller M, Schildhauer TA, Bogdanski D, Muhr G. Activation of human leukocytes after contact to highly porous tantalum biomaterial. Eur J Trauma 2004;30(Suppl.).