A new austenitic stainless steel with negligible nickel content: an in vitro and in vivo comparative investigation

Biomaterials 24 (2003) 4929–4939

ARTICLE IN PRESS

*Correspondin

6366580.

E-mail addres

0142-9612/$ - see

doi:10.1016/S014

A new austenitic stainless steel with negligible nickel content:an in vitro and in vivo comparative investigation

M Finia,*, N. Nicoli Aldinia, P. Torricellia, G. Giavaresia, V. Borsaria, H. Lengerb,J. Bernauerb, R. Giardinoa,c, R. Chiesad, A. Cigadad

aExperimental Surgery Department, Research Institute Codivilla-Putti, Rizzoli Orthopaedic Institute, Via di Barbiano 1/10, Bologna 40136, ItalybResearch Department B .ohler Edelsthal Gmbh & Co Kg-Kapfenberg, Austria

cUniversity of Bologna, Bologna, ItalydDepartment of Chemistry, Materials and Chemical Engineering ‘‘G. Natta’’, Polytechnic of Milan, Milan, Italy

Received 28 December 2002; accepted 30 May 2003

Abstract

New nickel (Ni)-reduced stainless-steel metals have recently been developed to avoid sensitivity to Ni. In the present study, an

austenitic Ni-reduced SSt named P558 (P558, B .ohler, Milan, Italy) was studied in vitro on primary osteoblasts and in vivo after

bone implantation in the sheep tibia, and was compared to ISO 5832-9 SSt (SSt) and Ti6Al4V. Cells were cultured directly on P558

and Ti6Al4V. Cells cultured on polystyrene were used as controls. Osteoblast proliferation, viability and synthetic activity were

evaluated at 72 h by assaying WST1, alkaline phosphatase activity (ALP), nitric oxide, pro-collagen I (PICP), osteocalcin (OC),

transforming growth factor-b1 (TGFb-1) and interleukin-6 (IL-6) after 1.25(OH)2D3 stimulation. Under general anaesthesia, foursheep were submitted for bilateral tibial implantation of P558, SSt and Ti6Al4V rods. In vitro results demonstrated that the effect of

P558 on osteoblast viability, PICP, TGF b-1, tumor necrosis factor-a production did not significantly differ from that exerted byTi6Al4V and controls. Furthermore, P558 enhanced osteoblast differentiation, as confirmed by ALP and OC levels, and reduced IL-

6 production. At 26 weeks, the bone-to-implant contact was higher in P558 than in SSt (28%, po0:005) and Ti6Al4V (4%, po0:05),and was higher in Ti6Al4V than in SSt (22%, po0:005). The tested materials did not affect bone microhardness in pre-existing hostbone as evidenced by the measurements taken at 1000 mm from the bone–biomaterial interface (F ¼ 1:89; ns). At the bone–biomaterial interface the lowest HV value was found for SSt, whereas no differences in HV were observed between materials

(F ¼ 1:55; ns). The current findings demonstrate P558 biocompatibility both in vitro and in vivo, and osteointegration processes areshown to be significantly improved by P558 as compared to the other materials tested.

r 2003 Elsevier Ltd. All rights reserved.

Keywords: Stainless steel; Nickel; Titanium; Osteoblasts; Bone

1. Introduction

Since the mean life expectation in humans isconstantly increasing, the working life of orthopaedicimplants should be likewise prolonged. The mechanicalproperties of biomaterials, as well as corrosion resis-tance and biocompatibility, therefore take on a growingimportance in the choice of the materials to be used forimplant manufacturing.Fracture fixation, joint replacement and traumatic

and iatrogenic segmental bone losses usually require the

g author. Tel.: +39-51-6366557; fax: +39-51-

s: [email protected] (M. Fini).

front matter r 2003 Elsevier Ltd. All rights reserved.

2-9612(03)00416-2

implantation of metallic devices [1]. Consequently, localand systemic metallosis, as well as the products of metalcorrosion and ion release, has been widely investigatedover the last decades [1–10]. In severe cases, notable andpermanent tissue changes can occur around metallicimplants resulting in the clinical entity of inflammation,pain and, at worst, implant failure [11].Standard metallic orthopaedic materials include

stainless steels (SSt), cobalt (Co)-based alloys, commer-cially pure (cp) titanium (Ti) and Ti-based alloys,with an increasing number of devices being made ofcpTi and Ti alloys. Regarding dental and non-cementedorthopaedic implants, cpTi and Ti alloys are generallypreferred to SSt and Co-alloys because of their lowermodulus, superior biocompatibility and corrosion

ARTICLE IN PRESSM. Fini et al. / Biomaterials 24 (2003) 4929–49394930

resistance [12]. However, SSt is still the most used metalfor internal fixation devices thanks to a favourablecombination of mechanical properties, acceptable bio-compatibility and cost effectiveness when compared toother metallic implant materials [13,14].A new class of austenitic SSt with interesting

mechanical and electrochemical properties has beenISO-standardized (ISO 5832-9). The high nitrogen con-tent of ISO 5832-9 SSt explains its superior corrosionresistance compared with traditional ISO 5832-1/D andISO 5832-1/E SSt and the higher mechanical properties(Rmin 770Mpa, Rs min 465Mpa, A min 35%) observedeven in the annealed state.However, a disadvantage seen for SSt is its tendency

towards corrosion under physiological conditions caus-ing a release of metal ions such as those of nickel (Ni)and chromium (Cr) [14]. This effect takes on a growingimportance in the case of ISO 5832-1/D and ISO 5832-1/E SSt. Both types of steel, in fact, are not immune tocrevice corrosion in the human body, which can increaseion release in the surrounding tissues by several ordersof magnitude.Ni has been reported to be the most common metal

sensitizer in humans [15] and some concern has beenexpressed regarding toxicity, susceptibility to bacterialinfection and cancerogenous effects, even though noevidence of any direct relationship between implants andcancer development exists [1,11,16,17]. In vitro studieshave shown that 10–50 mg/ml of Ni ions cause fibroblast,endothelial cell and monocyte total suppression ofmitochondrial function [16,18–20]. In vivo investiga-tions have demonstrated that more than 25 mg/g of Niimplanted in soft tissue may elicit severe inflammationand necrosis [16] and are therefore consistent within vitro citotoxicity studies. Furthermore, a concentra-tion of less than 10 mg/g can cause a severe inflammatoryresponse depending on the length of tissue exposure [16].This specific problem has led researchers to investigatevalid alternatives to these materials, in accordance withthe EU Directives on the usage of Ni-alloyed materialsfor manufacturing objects to come into contact with thehuman body (EU Dir. 94/27/CEE, para.2).To avoid the above-mentioned severe side effects of

SSt, Ti and its alloys have been considered the metal ofchoice for bone implants because of their high osteoin-tegration properties, suitable modulus of elasticity,lower density, improved biocompatibility and MRIcompatibility [21–23]. However, one might argue thatTi and Ti alloys are known for their relatively poor wearproperties; moreover, the tissue reactions around Ti-based implants, as well as the tendency to leave theorthopaedic devices in the body, have resulted in anumber of studies on biocompatibility [14,24,25]. Somerecent in vitro data have shown the negative effect of Tiparticles on osteoblast gene expression and on therelease of proinflammatory cytokines [26–30]. Ti and Ti

alloys have been associated with inflammation andcorrosion [14] and the toxicity of vanadium andaluminium has increased the interest in developing newTi-based alloys [12]. Furthermore, a shorter time offailure in total hip arthroplasties with Ti alloy versuscobalt chrome alloy has been reported, together with ahigher loosening rate and peri-prosthetic osteolysis [31].Ni-reduced SSt metals with high nitrogen content

have recently been developed to address the issue ofsensitivity to Ni and appear to have superior mechanicalproperties and better corrosion resistance [13].A new austenitic SSt named P558 (Bohler, Milan,

Italy), which has been recently patented, may provide aninteresting alternative to conventional SSt, Co-basedalloys, and Ti and Ti-alloys. P558 has a high Mn and Ncontent and a negligible Ni (o0.20%) content (inaccordance with ASTM E 112: 4–5). Such a low Nicontent does not induce Ni ion release and, conse-quently, prevents allergic reactions to Ni, as confirmedby a previous in vivo maximization-sensitization testperformed in guinea pigs (in accordance with ISO10993-10, 1995; Biological evaluation of medical de-vices, Part.10: Test for irritation and sensitization)(unpublished data).The aim of the present study was to evaluate in vitro

and in vivo biocompatibility in terms of osteoblastproliferation, differentiation and synthetic activity, aswell as in vivo osteointegration, through comparison ofa Ni-reduced SSt (P558) with the Ti alloy Ti6Al4V andISO 5832-9 SSt (SSt).

2. Materials and methods

2.1. Materials

Disks made of P558 and Ti6Al4V with a diameter of10mm and thickness of 1mm were used for the in vitrostudy, and cylindrical rods made of P558, SSt andTi6Al4V with a diameter of 4mm and length of 12mmfor the in vivo implantation in sheep. Specimens forthe in vitro and in vivo testing were used in the‘‘as-received’’ conditions and not submitted to surfacefinishing process.The chemical compositions of P558, SSt and Ti6Al4V

are reported in Table 1.The surface roughness of P558, SSt and Ti6Al4V

implants for in vitro and in vivo tests was measured bymeans of a 3D laser profilometer (UBM-MicrofocusCompact, NanoFocus AG, Germany). Roughnessmeasurements were calculated on 10 profiles acquiredon the surface of the 10mm diameter disks and on five5.6mm long profiles, 500 points/mm, acquired on thelateral surface of the cylindrical rods in a directionparallel to the cylinder axis. The following parameterswere calculated for both in vitro and in vivo specimens:

ARTICLE IN PRESS

Table 1

Chemical composition (%) of ISO 5832-9 stainless steel (SSt), Ni-

reduced stainless steel (P558) and Ti6Al4V used for the present study

Chemical element Material

SSt P558 Ti6Al4V

C 0.04 0.16 0.009

Si 0.25 0.43 —

Mn 3.58 9.47 —

P 0.009 o0.005 —

S 0.001 0.0002 —

Cr 21.43 16.62 —

Mo 2.39 3.25 —

Ni 9.38 o0.02 —

V 0.05 o0.02 3.84

W 0.05 — —

Cu 0.06 0.02 —

Co o0.05 — —

Ti o0.005 — Balance

Al 0.01 0.006 5.9

H — — 0.0088

Y — — 0.001

O — — 0.11

Nb 0.340 0.026 —

N 0.43 0.49 0.009

Fe Balance 0.1

1The in vivo study was performed strictly following Italian and

European Law on animal experimentation Italian D.L. January 27,

1992; ISO 10993-2; N.I.H. No. A5424-01.

M. Fini et al. / Biomaterials 24 (2003) 4929–4939 4931

Ra (the arithmetic mean of the departures of theroughness profile from the mean line) and Rmax ISO4287 (the maximum profile valley depth).Materials were sterilized at gamma rays (25 kGy)

before their in vitro and in vivo use.

2.2. In vitro study

Osteoblasts were isolated sterilely from small speci-mens of trabecular bone derived from the femoralcondyles of adult male rats (b.w.: about 300 g).Trabecular bone fragments were put in DMEM:F12serum-free culture medium and immediately processedto obtain primary cultured osteoblasts. Bone fragmentswere washed with DMEM:F12 serum-free medium anddigested in F12 medium with 1mg/ml collagenase for90min at 37�C. The enzymatic reaction was stopped byadding an equal volume of medium with 10% of FCS,and the supernatant containing the released cells wascollected. Washing and collecting were repeated threetimes. The cells obtained were pelletted by centrifuga-tion, resuspended, seeded in 35mm dishes, cultured inDMEM medium supplemented with ascorbic acid(50 mg/ml) and b-glycerophosphate (10�8m), and incu-bated at 37�C in a humidified 95% air/5% CO2atmosphere. Cells were released at confluence with0.05% (w/v) trypsin and 0.02% (w/v) EDTA, countedand used (first passage) for the experiment in 24-wellculture plates. Then, they were also seeded (1� 104 cells/ml) in four-well chamber slides for characterization, by

adding 10�9m 1,25(OH2)D3 to allow mineralization andexpression of the osteoblast phenotype.Five Ni-reduced stainless-steel (P558) samples and

five equally sized Ti6Al4V samples were placed on eachof the two 24-well plates and a cell suspension(0.5� 104 cells in 100 ml) was directly seeded over eachsample. The same amount of osteoblast-like cells wasplated into the same number of empty polystyrene wellsas Control Group. After a 2-h incubation to allow celladhesion to the substrate, 900 ml of DMEM (supple-mented with 50 mg/ml ascorbic acid and b-glyceropho-sphate 10�8m) were added to each well. Cultures weremaintained in the same conditions as described abovefor 72 h, and the medium was changed after 24 h.1.25(OH2)D3 (10

�9m) (Sigma, UK) was added to the

plates in the last 48 h. No bacterial or fungal contam-ination was found during the experiments.After 72 h the supernatant was collected from all the

wells and centrifuged to remove particulates, if any.Aliquots were dispensed in Eppendorf tubes for storageat �70�C and assayed for type I collagen (PICP,Prolagen-C enzyme Immunoassay kit, Metra Biosystem,CA, USA), interleukin-6 (IL-6, Human IL-6 Immu-noassay kit, Biosource International, CA, USA), trans-forming growth factor-b1 (TGF-b1, Quantikine humanTGF-b1, Immunoassay, R&D Systems, MN, USA) andtumor necrosis factor-a (TNF-a, Quantikine humanTNF-a, Immunoassay, R&D Systems, MN, USA).Alkaline phosphatase activity (ALP, Sigma Kinetic

method kit, St. Louis, MO, USA), nitric oxide (NO,Sigma colorimetric assay, St. Louis, MO, USA), andosteocalcin (OC, Novocalcin enzyme Immunoassay kit,Metra Biosystem, CA, USA), were tested on super-natants immediately after collection. The measuredconcentrations were normalized by cell number. Finally,the Cell Proliferation Reagent WST-1 test was done toassess cell proliferation and viability: 100 ml of WST1solution and 900 ml of medium (final dilution: 1:10) wereadded to the cell monolayers, and the multi-well plateswere incubated at 37�C for a further 4 h. Supernatantswere quantified spectrophotometrically at 450 nm with areference wavelength of 640 nm. Results are reported asoptical density (OD).

2.3. In vivo implantation1

Four adult mongrel sheep—body weight 6575.0 kg(Allevamento Pancaldi Massimo, Budrio, Bologna,Italy)—were used. After a quarantine period of 7 days,48 h of starvation and 24 h without drinking water,the animals received general anaesthesia following astandardized protocol: pre-medication with 10mg/kg


ketamine i.m. (Ketavet 100, Farmaceutici Gellini SpA,Aprilia, Italy), 0.3mg/kg b.w. xylazine i.m. (RompunBayer AG, Leverkusen, Germany) and 0.0125mg/kgb.w. atropine sulphate s.c.; induction with 6mg/kgsodium thiopentone (2.5% solution, Pentothal, HoechstAG, Germany) i.v.; maintenance with O2, N2O and2–3% fluothane (Halothan Hoechst AG Germany)under assisted ventilation (Servo Ventilator 900 D,Siemens, Germany).After skin incision and fasciae dissection, a segment of

about 10 cm of the tibial diaphyses was exposed underaseptic conditions: it extended from about 5 cm from theknee joint to about 3 cm from the distal extremity of thetibia. Then, 2.5, 3 and 3.9mm-diameter drills were usedto pre-drill six holes under external irrigation withsaline solution to minimize the risk of bone overheating.Six experimental cylindrical rods (P558) were thenimplanted in the left tibia of all the animals using abeater. Following the same surgical technique, sixcylindrical rods made of SSt and Ti6Al4V wereimplanted in the right tibial diaphysis. Overall, a totalof 24 implants of P558, 12 implants of SSt and 12implants of Ti6Al4V were inserted. The wounds weresutured in two layers and medicated until stitch re-moval. Antibiotics (cefalosporin 1 g/day for 5 days) andanalgesics (ketoprofen 500mg/day for 3 days) wereadministered postoperatively.Sheep were kept in single boxes at a T of 2271�C and

relative humidity of 55%, and allowed standard food(Piccioni Settimo Milanese, Milan, Italy, 0.85% Ca and0.51% P) and tap water ad libitum. About 2 weeks aftersurgery, the animals were housed externally untilsacrifice (26 weeks after surgery). A fluorescent dye(30mg/kg b.w. of oxytetracycline) was injected i.m 10, 9,2 days and 1 day before killing the sheep in order tolabel the newly formed bone next to the implants.The animals were euthanized with intravenous admin-

istration of Tanax (Hoechst, Frankfurt am Main,Germany) under general anaesthesia.The tibiae were explanted: after macroscopic and

radiographic observation, bone specimens of about2.5 cm containing the implants were obtained and usedfor histological, histomorphometrical and microhard-ness investigations.

2.4. Histology and histomorphometry

The bone specimens containing the implants werefixed in 10% formalin solution buffered at 7.2 pH,dehydrated in graded series of alcohols, and finallyincluded in polymethylmethacrylate. The whole processwas carried out at a T of 2271�C and humidity rate of48%. Blocks were sectioned along a plane parallel to thelong axis of the implants by using a Leica 1600 diamondsaw microtome (Ernst Leitz, Wetzlar, Germany) andyielding undecalcified sections of 100 mm in thickness.

Afterwards, 20–30 mm thick slices were obtained usingan Exact cutting system (boronitride—saw blade;Kulzer System, Norderstedt, Germany), as well as discsof sandpaper with progressively smaller granules; carewas taken to avoid damaging the bone–biomaterialinterface.The sections were stained with Goldner’s trichrome

(basic fuchsin and light green) in order to identifyosteoid or connective tissue (red-orange) and miner-alized bone (green). After a light microscopy evaluationat various different magnifications, histomorphometri-cal analyses were performed using an optic microscope(Zeiss Axioskop, Carl Zeiss, Jena, Germany) connectedto an image analyser system (Kontron KS300 v.2,Kontron Elektronik, Munchen, Germany).The bone-to-implant contact (the fractional, linear

extent of bone opposed to the implant surface dividedby the total surface perimeter of the implant� 100) wasmeasured to evaluate the bone–biomaterial interfaceand osteointegration. Each measurement was takensemi-automatically at an original magnification of� 40 by an experienced blinded investigator.Unstained 100-mm-thick sections were observed under

fluorescent light to evaluate new bone deposition next tothe implants.

2.5. Microhardness

The same resin-embedded blocks containing theremaining part of the implanted screws were used tomeasure bone hardness by means of an indentation test(Microhardness VMHT 30, Leica, Wien, Austria).Briefly, a rotatory wheel set at 150 rpm was used withSiC paper and water lubrification. The resin-embeddedblocks containing the specimen were then polished usinga diamond paste with a progressively finer grain size.The smooth surface obtained was observed under themicroscope and the bone–material interface and otherareas to be examined were clearly visible.The microhardness measurements were taken tangen-

tially to the interface with a Vickers indenter (four-sidedpyramid with square base and an apex angle betweenopposite sides of 136�7150) applied to the bone at aload of 0.05 kgf and dwell time of 5 s. The Vickershardness degree (HV) was calculated by dividing theindentation force by the surface of the imprint (fourpyramid surfaces) observed at the microscope. Theaverage value for each sample was calculated on a meanof 10 for each examined area at two sites: in boneregrown within 200 mm from the interface (HV200mm)and in pre-existing host bone, at 1000 mm from theinterface (HV1000mm). Finally, the bone maturationindex (BMI) was calculated by dividing the microhard-ness (HV) of the bone regrown at the interface by themicrohardness (HV) of the pre-existing bone multipliedby 100.

ARTICLE IN PRESS

able 3

oughness measurements for in vivo specimens of nickel-reduced

ainless steel (P558), ISO 5832-9 stainless steel (SSt) and Ti6Al4V

arameter Unit P558, n ¼ 5 SSt, n ¼ 5 Ti6Al4V, n ¼ 5

a mm 0.72670.043a 0.91570.045b 0.36570.027

maxISO 4287 mm 6.7470.77c 8.8070.82 3.1570.35d

oughness was calculated on the lateral surface of a cylindrical rod in

direction parallel to the cylinder axis.

onferroni multiple comparison test:aP558 versus SSt and Ti6Al4V ðpo0:0005Þ:bSSt versus Ti6Al4V ðpo0:0005Þ:cP558 versus SSt (po0:005).dTi6Al4V versus P558 and SSt (po0:0005).

Table 4

In vitro results in terms of osteoblast proliferation, differentiation and

synthetic activity after culturing samples on Ni-reduced stainless steel

(P558) and Ti6Al4V plates for 72 h

Unit Control P558 Ti6Al4V

WST1 OD at 450nm 1.01670.088 1.02870.028 0.96670.064


2.6. Statistics

Statistical analysis was performed using the SPSSv.10.1 software (SPSS Inc., Chicago, IL, USA). Data arereported as mean7SD at a significant level of po0:05:After testing data for normal distribution and homo-geneity of the variance, a multiple way ANOVA was usedto assess significant interactions between selected factors(sheep id., implant sites, materials) and histomorphome-trical and microhardness data. When these interactionswere found, a univariate ANOVA was done to investigatethe effects of the factors on the data by means ofhypotheses expressed as linear matrix according to theSPSS syntax. When no interaction was found, in the caseof in vitro results and between roughness parameters, aone-way ANOVA was done, followed by the Bonferronimultiple comparison test, in order to check for thepresence of significant differences between the resultsobtained from the different materials. The paired Student’st-test was performed to compare data between microhard-ness measurement sites within the same material.

ALP IU/l 14.4470.65 17.1871.22a,b 15.4270.82OC ng/ml 9.6571.03c 19.5870.68 19.4370.95PICP ng/ml 5.2070.10 4.7770.35 4.6770.65TGF-b1 pg/ml 822.67728.51 951.00718.00 838.33724.50IL-6 pg/ml 47.8875.81 46.6674.34b 57.6273.53TNF-a pg/ml 25.3276.59 21.2273.63 23.6672.10

The control group cells were cultured in polystyrene wells in the

absence of materials (Mean7SD; n ¼ 5).Bonferroni multiple comparison test:aP558 versus control (po0:005).bP558 versus Ti6Al4V (po0:05).cControl versus P558 and Ti6Al4V ðpo0:0005Þ:

3. Results

3.1. Roughness measurement

The roughness measurements taken on in vitro andin vivo samples are shown in Tables 2 and 3. The Ra(63%) and Rmax (33%) calculated on the Ti6Al4Vspecimens for the in vitro test showed the highest valuesbut, on average, little difference was found betweenthem and the corresponding values for P558 (Table 2).Although no particular surface treatment was admi-

nistered to the various different materials, somedifferences in Ra (F ¼ 251:23; po0:0005) and Rmax(F ¼ 87:84; po0:0005) were found between the speci-mens of the materials used for the in vivo testing.Ti6Al4V showed the lowest roughness value in terms ofRa and Rmax (Table 3).

3.2. In vitro study

Data on cell characterization confirmed the osteoblastphenotype. The level of ALP activity and OC produc-

Table 2

Roughness measurements for in vitro specimens of nickel-reduced

stainless steel (P558) and Ti6Al4V

Parameter Unit P558, n ¼ 10 Ti6Al4V, n ¼ 10

Ra mm 0.14570.066� 0.23670.054�

RmaxISO 4287 mm 1.67170.478�� 2.22470.616��

Unpaired Student’s t-test between roughness measurements within

each surface treatment: �po0.005; ��po0.05. Roughness wascalculated on the surface of one disk for each material in random

directions.

T

R

st

P

R

R

R

a

B

tion increased significantly in the presence of 10�9m1,25(OH2)D3. The Von Kossa staining further con-firmed these results (data not shown).No significant differences in cell viability (WST1),

PICP, TGF-b1 and TNF-a production were observedbetween materials and control (polystyrene) (Table 4).The ALP production was significantly higher (F ¼11:18; po0:005) in P558 cultured cells than in thecontrol (19%, po0:005) and Ti6Al4V (11%, po0:05)groups. OC levels were significantly higher in osteoblastscultured in the presence of biomaterials than in thecontrol group (F ¼ 159:41; po0:0005). Finally, IL-6production was significantly higher in Ti6Al4V (23%)than in P558 cultures (F ¼ 10:52; po0:05).

3.3. In vivo implantation

All animals tolerated surgery well and surviveduntil the final experimental time. No postoperativecomplications were observed. Macroscopically, no signsof infection were observed around the implants. Apostoperative X-ray confirmed the correct positioning of


the rods in the cortical tibial bone and the absence ofosteolytic areas around the implants. A periosteal/endosteal callus covered the external lateral andintramedullary surfaces of all the rods.Figs. 1–4 show the histological sections of P558, SSt

and Ti6Al4V. All of the tested materials osteointegratedwith bone at 26 weeks, even though some differenceswere seen between the various different compositions.Histologically, some areas of fibrous tissue were foundonly for SSt (Fig. 5), whereas direct bone apposition wasalways observed with P558 and Ti6Al4V. Fig. 6(A–C)shows the histological unstained sections under fluor-escent light: fluorescence is not very marked at the

Fig. 1. Histological section of Ni-reduced SSt (P558) 26 weeks after

surgical implantation in sheep tibial diaphysis. Osteointegration with

cortical bone and abundant periosteal callus covering the external

lateral surface of the implant (basic fuchsin and light green, � 2).

Fig. 3. Histological section of ISO 5832-9 SSt (SSt) 26 weeks after

surgical implantation in sheep tibial diaphysis. A worst bone–

biomaterial contact is visible as compared to P558 (basic fuchsin and

light green, � 2).

Fig. 2. Histological section of Ni-reduced SSt (P558) 26 weeks after

surgical implantation in sheep tibial diaphysis. The higher magnifica-

tion clearly reveals direct bone apposition to the material surface. The

bone tissue near the implant surface is of high quality, and resembles

the compact bone (basic fuchsin and light green, � 4).

Fig. 4. Histological section of Ti6Al4V 26 weeks after surgical

implantation in sheep tibial diaphysis. Bone is regrown in direct

contact with the material surface (basic fuchsin and light green, � 4).

bone–biomaterial interface of all the tested implants andis not visible in the cortical bone far from the implants.However, a greater fluorescence can be seen adjacent tothe Ti6Al4V implants.The one-way ANOVA revealed significant differences

in bone-to-implant contact (F ¼ 5:95; po0:005) be-tween tested materials (Table 5): the bone-to-implantcontact of P558 was higher than that of SSt (28%,po0:005) and Ti6Al4V (4%, po0:05) and, similarly,that of Ti6Al4V was higher than that of SSt (22%,po0:005).Finally, Table 6 shows data on the microhardness

investigations performed at the bone–biomaterial inter-face and in normal bone far from the implants. The

ARTICLE IN PRESS

Fig. 5. Histological section of ISO 5832-9 SSt (SSt) 26 weeks after

implantation in sheep tibial diaphysis. An area of red-stained fibrous

tissue is clearly visible at the bone–biomaterial interface (basic fuchsin

and light green, � 4).

(a)

(b)


tested materials did not affect bone microhardness inpre-existing host bone (HV1000 mm) (F ¼ 1:89; ns). At thebone–biomaterial interface, the lowest HV200 mm valuewas found for SSt, whereas no differences in HV200 mmwere observed between materials at 26 weeks (F ¼ 1:55;ns). The BMI showed that the regrown bone matureddifferently, in the following order: P558 (84%)>SSt(79%)>Ti6Al4V (77%).

(c)

Fig. 6. A representative fluorescent light micrograph of a cross-section

area of P558 (A), ISO 5832-9 SSt (B), and Ti6Al4V (C) implants and

adjacent bone. No differences in terms of areas stained by the

fluorescent dye can be seen for P558 and SSt. A higher fluorescence

uptake is visible adjacent to the Ti6Al4V implant.

4. Discussion

In the present study, attention was focused on an SStwith negligible amount of Ni (named P558) to obtainuseful information on in vitro and in vivo behaviour interms of osteoblast activity and osteointegration rateafter implantation in the sheep tibial cortical bone.According to the present in vitro results, at 72 h no

significant differences were found in cell viability,proliferation and PICP production between controlosteoblasts and osteoblasts cultured on P558 andTi6Al4V. The ALP production was higher for P558than for Ti6Al4V and controls, and the difference wassignificant when the P558 group was compared to thecontrol group. It should be pointed out that a strongcorrelation between ALP and bone-like nodulus forma-tion was observed in osteoblast cultures, thus suggestingthat ALP may be a good indicator of in vitro osteogen-esis [32]. No significant differences were seen betweenP558 and Ti6Al4V cultures.

OC levels were significantly higher in osteoblastscultured with the materials than in control cultures(102%). On the one hand, ALP is an early differentia-tion marker of osteoblasts, on the other hand thenoncollagenous protein OC is considered a highly

ARTICLE IN PRESS

Table 5

Histomorphometrical result of nickel-reduced stainless steel (P558), ISO 5832-9 stainless steel (SSt) and Ti6Al4V at 26 weeks. (Mean7SD)

Parameter Unit P558, n ¼ 24 SSt, n ¼ 12 Ti6AlV, n ¼ 12

Bone-to-implant contact % 69.96713.64a,b 55.06727.50c 67.33712.84

Bonferroni multiple comparison test:aP558 versus SSt (po0:005).bP558 versus Ti6Al4V ðpo0:05).cSSt versus Ti6Al4V (po0:005).

Table 6

Microhardness results of nickel-reduced stainless steel (P558), ISO

5832-9 stainless steel (SSt) and Ti6Al4V at 26 weeks. (Mean7SD)

Parameter Unit P558 n ¼ 24 SSt n ¼ 12 Ti6AlV n ¼ 12

HV200 mm Vickers 63.575.5� 60.677.8� 61.476.8�

HV1000mm Vickers 73.377.3 73.776.9 76.476.8

HV200mm: in the regrown bone within 200mm from the interface;HV1000mm: in the pre-existing host bone at 1000mm from the interface.Paired Student’s t-test between hardness measurement sites within each

surface treatment: �po0:0005:

M. Fini et al. / Biomaterials 24 (2003) 4929–49394936

specific osteoblast marker and represents a later stage ofosteoblast differentiation [14].As far as cytokines and growth factors are concerned,

P558 cultured osteoblasts showed the highest TGF b-1and the lowest TNF-a values, even if statisticalsignificance was not reached versus other groups, asalso observed by other authors investigating SSt andcpTi [33]. The highest IL-6 levels were found forTi6Al4V cultured cells with a significant differencebetween the latter and the P558 group (23%). IL-6 is aknown inducer of osteoclastogenesis and a stimulator ofbone resorption [27].The effect of P558 on osteoblast proliferation, ALP,

PICP, TGF b-1 and TNF-a production did notsignificantly differ from that exerted by Ti6Al4V. Inaddition, it significantly improved osteoblast differentia-tion, as demonstrated by ALP and OC levels, andsignificantly reduced the pro-inflammatory cytokine (IL-6) responsible for in vivo bone implant failure.Primary osteoblasts were used for the in vitro study

because the biocompatibility of an orthopaedic implantstrictly depends on the effect of the implant on bone-forming cells, and changes in osteoblast proliferation,maturation and differentiation are important events inossification [33]. On the other hand, osteoblasts alsoplay a direct role in implant loosening [26]. Primary cellswere preferred in order to avoid mutated proliferationand marker expression of immortalized cell lines [14].The data obtained from the in vitro step of the

research strongly supported subsequent animal implan-tation.The present in vivo results demonstrated tissue

compatibility of P558: further evidence was providedby the osteointegration rate of the Ni-reduced SSt(P558), which was better than that of ISO 5832-9 SSt(SSt) and Ti6Al4V.At the beginning of the study, the present authors

hypothesized that P558 would behave better than SStbecause of the absence of any Ni-related negative effectson bone tissue. This hypothesis was confirmed by thehistological observation and histomorphometrical mea-surement of bone-to-implant contact. In accordancewith expectations, comparison between Ti6Al4V andSSt showed a significant enhancement of bone apposi-tion in Ti6Al4V. Although some authors have not foundany difference in osteointegration between titanium and

stainless-steel implants inserted in rabbit tibial corticalbone and dog calvaria, in particular at long-experi-mental times [34,35], there is a general agreement aboutthe superiority of titanium-based materials over stain-less-steel implants when inserted both in trabecular andin cortical bone [23,36,37]. Moreover, the comparisonbetween P558 and Ti6Al4V revealed a significantlybetter osteointegration rate of P558. This unexpectedresult led researchers to have bone-to-implant contactmeasurements repeated by another blinded and experi-enced investigator, in order to check data reliability.However, results obtained through the objective estima-tion of bone-to-implant contact in P558 and Ti6Al4Vimplants did not differ by more than 0.4% and 1.6%,respectively, and the final determination reported in thisstudy is the average of the two measurements.To the present authors’ knowledge, no in vivo data

are available on bone implantation to make a compar-ison between a Ni-reduced SSt, such as P558 andTi6Al4V. When comparing an SSt with negligibleamount of Ni with conventional SSt and Ti6Al4V interms of implantation in soft tissues, results have showna notably better performance of the SSt with negligibleamount of Ni versus SSt, and Ti-based materials havebeen demonstrated to be the gold standard amongmetallic implant materials, as far as inflammation isconcerned [12]. The current results were obtained afterbone implantation and also showed the superiority ofP558 over Ti6Al4V. At this stage of the research, thesefindings cannot be explained. Further investigations arerequired, firstly because of the limited number ofanimals used and secondly on account of the singleexperimental time selected. The very fact that only asingle and, what is more, long-term experimental time

ARTICLE IN PRESSM. Fini et al. / Biomaterials 24 (2003) 4929–4939 4937

was chosen could have masked the beneficial effects ofthe Ti6Al4V surface during the early phases of theosteointegration processes, leading to the impossibilityof evaluating a possible acceleration of bone fixation bythe osteointegrative properties of the Ti6Al4V implants.However, the fluorescent dye labelling near the implantswas more evident for Ti6Al4V than for SSt and P558.The former event may reveal ongoing bone-tissueremodelling at the interface, whereas the latter shows asteady-state bone condition at 26 weeks.The experimental time of 26 weeks was selected

because no previous in vivo data were available andverification was necessary for the presence of osteolysisor side effects brought about by the implantation ofthe experimental material and Ni-release from ISO5832-9 SSt.Bone microhardness data revealed that no significant

differences existed in terms of bone–biomaterial inter-face between the three materials tested. BMI progres-sively decreased at the interface in the following order:P558>SSt>Ti6Al4V. Microhardness was measuredwithin 200 and at 1000 mm from the bone–biomaterialinterface not only to assess the microarchitecturalquality of the bone regrown around the tested bioma-terials but also to compare the bone regrown at theinterface with the pre-existing bone far from theimplant. In previous bone microhardness investigationsafter biomaterial implantation conducted by the presentauthors, the effect of the surgical implant was evident upto a 500-mm distance from the bone–biomaterial inter-face, whereas bone at a greater distance was demon-strated to be unaffected by implantation surgery [38].The comparison between bone hardness at the interfaceand bone hardness of normal bone far from the interfaceshowed that bone maturation at the bone–implantinterface was still in progress in all of the testedmaterials, as demonstrated by the significant differencestill existing between all materials at the bone interfaceand in normal pre-existing bone.Microhardness is considered to be the expression of

microstructural bone parameters, such as calcificationdegree, arrangement and number of collagen fibers,ratio between collagen fibers and ground substances,and mineral quantity per volume unit [38–44]. Thehigher the value of bone hardness, the higher the level ofbone mineralization and maturation. Mechanical ana-lyses, such as push-out or pull-out tests, can furtherimprove the current findings and provide better in-formation about the osteointegration rate throughdetermination of the bone–implant fixation stability.However, at this stage of the study, and given the limitednumber of animals used, it was necessary to adopt amechanical method, such as microhardness evaluation,that would not destroy the bone–biomaterial interfaceand could be used on the same specimens employed forthe histomorphometrical investigation. The evaluation

of the orthopaedic implant–bone interface should ensuregood reliability when using both a morphologic and amicrostructural method [41]. Morphological studiesalone, in fact, cannot supply precise information onbone maturity and mechanical resistance. As reportedby Huja et al. [44], the microhardness technique allowsthe study of bone hardness variations over smalldistances, and such information is considered usefulwhen investigating the adaptation of the bone adjacentto an implant. Moreover, a good correlation coefficientbetween microhardness, elastic modulus, yield stress,volume fraction of mineral and calcium content wasregistered, together with a close relationship betweenmineralization and hardness [44].According to the present results, P558 provides a very

compatible material with a promising future as animplant material. The biological behaviour of P558 isprobably due to the absence of Ni-related negativeeffects on cells and tissues. Further investigations arenecessary to verify the hypothesized indirectly improvedmicrovascularization of peri-implant bone tissue, whichhas been observed in soft tissue implants made of SStwith a negligible amount of Ni when compared toimplants made of conventional SSt (15). Vascularizationis considered to be very important for fracture healingand osteointegration, since bone growth does not occurwithout an adequate blood supply [45,46]. In addition,the Cr content is lower in the experimental materialP558 than in SSt (Table 1). Cr is accepted as a sensitizerand reported to impair osteoblast proliferation, differ-entiation and cytokine release [1,29]. Consequently,the lower amount of Cr, associated with the absenceof Ni, could also have played a role in the enhancementof osteoblast function and tissue healing rate observedin the present investigations when comparing P558with SSt.Methods and materials used for implant manufactur-

ing should meet high-standard requirements to ensurethe long-term functional life of the prosthesis. Thechoice of a material group suitable for a particularpurpose is guided not only by its chemical compositionbut also by its manufacturing process, which positivelyand lastingly influences the specific features of thematerial. Corrosion resistance and non-noxiousness of aproduct are crucial factors for material biocompatibil-ity. Good resistance to corrosion, a highly polishedsurface and good material characteristics all depend onthe structural homogeneity and highly advanced gradeof purity of the material itself. The properties of theaustenitic SSt P558 used for the present study, such asthe absence of Ni-ion release and allergic reactions, itslow cost versus Ti-based metals, the best machinabilityversus cpTi and Ti biphasic alloys, its high mechani-cal strength and toughness, its excellent general andlocalized corrosion resistance, make it suitable fororthopaedic reconstructive surgery.


Since the main advantage offered by the clinical use ofbioactive cpTi and Ti alloys in orthopaedics is reportedto be its application in the treatment of osteoporotic andelderly patients with decreased osteogenesis [23], a directcomparison between conventional materials and P558using standardized in vitro and in vivo models ofosteoporosis could be, in the present authors’ opinion,of extreme interest to obtain additional useful informa-tion for clinical purposes.

Acknowledgements

The research project was partially supported bygrants from the Rizzoli Orthopedic Institute. Theauthors are deeply indebted to B .ohler Edelstahl Gmbh& Co. KG (Austria) for their support and to the B .ohlerDivision of B .ohler Uddeholm Italia Spa (Milan, Italy)for providing the materials. Moreover, they wouldlike to thank Claudio Dalfiume, Nicola Corrado,Patrizio Di Denia, Franca Rambaldi and Patrizia Nini(Experimental Surgery Department) for their technicalassistance. No benefits in any form have been receivedor will be received from a commercial party directly orindirectly related to the subject of this article.

References

[1] Swiontkowski MF, Agel J, Schwappach J, McNair P, Welch M.

Cutaneous metal sensitivity in patients with orthopaedic injuries.

J Orthop Trauma 2001;2:86–9.

[2] Jacobs JJ, Gilbert JL, Urban RM. Corrosion of metal ortho-

paedic implants. J Bone Jt Surg 1998;80A(2):269–82.

[3] Bianco PD, Ducheyne P, Cuckler MJ. Titanium serum and urine

levels in rabbits with a titanium implant in the absence of wear.

Biomaterials 1996;17:1937–42.

[4] Savarino L, Stea S, Granchi D, Visentin M, Ciapetti G, Donati

ME, Rollo G, Zinghi G, Pizzoferrato A, Montanaro L, Toni A.

Sister chromatid exchanges and ion release in patients wearing

fracture fixation devices. J Biomed Mater Res 2000;50(1):21–6.

[5] Pierini G, Fini M, Giavaresi G, Dallari S, Brayda Bruno M,

Rocca M, Nicoli Aldini N, Giardino R. Atomic absorption

spectrophotometry (AAS) for the evaluation of metallosis in

prostheses and artficial organs use: a new approach. Int J Artif

Organs 1999;7(22):522–7.

[6] Brayda Bruno M, Fini M, Pierini G, Giavaresi G, Rocca M,

Giardino R. Evaluation of systemic metal diffusion after spinal

pedicular fixation with titanium alloy and stainless steel systems:

a 36 month experimental study in sheep. Int J Artif Organs

2001;24(1):41–9.

[7] Case CP, Lankamer VG, James C, Palmer MR, Kemp AJ, Heap

PF, Solomn L. Widespread dissemination of metal debris from

implants. J Bone Jt Surg Br 1994;76-B(5):701–12.

[8] Lugowski SJ, Smith DC, McHugh AD, Van Loon C. Release of

metal ions from dental implant materials in vivo: determination of

Al, Co, Cr, Mo, Ni, V, and Ti in organ tissue. J Biomed Mater

Res 1991;25:1443–58.

[9] Lalor PA, Revell PA, Gray AB, Wright S, Railton GT, Freeman

MAR. Sensitivity to titanium. A cause of implant failure. J Bone

Jt Surg Br 1991;73(1):25–8.

[10] Michel R, Nolte M, Reich M, Loer F. Systemic effects of

implanted prostheses made of cobalt–chromium alloys. Arch

Orthop Trauma Surg 1991;110:61–74.

[11] Bischoff UW, Freeman MAR, Smith D, Tuke MA, Gregson PJ.

Wear induced by motion between bone and titanium or cobalt-

chrome alloys. J Bone Jt Surg 1994;76-B:713–6.

[12] Woodman JL, Jacobs JJ, Galante JO, Urban RM. Metal ion

release from titanium-based prosthetic segmental replacements of

long bones in baboons: a long-term study. J Orthop Res

1984;1(4):421–30.

[13] Disegi JA, Eschbach L. Stainless steel in bone surgery. Injury

2000;31(S4):4–6.

[14] Schmidt C, Ignatius A, Claes LE. Proliferation and differentiation

parameters of human osteoblasts on titanium and steel surfaces.

J Biomed Mater Res 2001;54:209–15.

[15] Kraft CN, Burian B, Perlick L, Wimmer MA, Wallny T, Schmitt

O, Diedrich O. Impact of a nickel-reduced stainless steel implant

on striated muscle microcirculation: a comparative in vivo study.


[16] Long M, Rack HJ. Titanium alloys in total joint replacement—a

material science perspectives. Biomaterials 1998;19(18):1621–39.

[17] Hallab N, Jacobs JJ, Black J. Hypersensitivity to metallic

biomaterials: a review of leukocyte migration inhibition assays.

Biomaterials 2000;21(13):1301–14.

[18] Wataha JC, O’Dell NL, Singh BB, Ghazi M, Whitford GM,

Lockwood PE. Relating nickel-induced tissue inflammation to

nickel release in vivo. J Biomed Mater Res (Appl Biomater)

2002;58:537–44.

[19] McDonald DJ, Enneking WF, Sundaram M. Metal-associated

angiosarcoma of bone: report of two cases and review of the

literature. Clin Orthop 2002;396:206–14.

[20] Wataha JC, Hanks CT, Craig RG. The in vitro effects of metal

cations on eukaryotic cell metabolism. J Biomed Mater Res

1991;25:1133–49.

[21] Wataha JC, Hanks CT, Sun ZL. In vitro reaction of macrophages

to metal ions from dental biomaterials. Dent Mater 1995;11:

239–45.

[22] Wataha JC, Sun ZL, Hanks CT, Fang DN. Effect of Ni ions on

expression of intercellular adhesion molecule 1 by endothelia cells.


[23] Christensen FB, Dalstra M, Sejling F, Overgaard S, Bunger C.

Titanium alloy enhances bone-pedicle screw fixation: mechanical

and histomorphometrical results of titanium alloy versus stainless

steel. Eur Spine J 2000;9:97–103.

[24] Perren SM. The concept of biological plating using the limited

contact-dynamic compression plate (LC-DCP). Injury 1991;

22(Suppl.):1–41.

[25] Pohler OEM. Unalloyed titanium for implants in bone surgery.

Injury, Int J Care Injured 2000;31:SD7–13.

[26] Shida J, Trindade MC, Goodman SB, Schurman DJ, Smith RL.

Induction of interleukin-6 release inhuman osteoblast-like cells

exposed to titanium particles in vitro. Calcif Tissue Int

2000;67(2):151–5.

[27] Takei H, Pioletti DP, Kwon SY, Sung KL. Combined effect of

titanium particles and TNF-alpha on the production of IL-6 by

osteoblast-like cells. J Biomed Mater Res 2000;52(2):382–7.

[28] Pioletti DP, Takei H, Kwon SY, Wood D, Sung KL. The

cytotoxic effect of titanium particles phagocytosed by osteoblasts.

J Biomed Mater Res 1999;46(39):399–407.

[29] Wang JY, Wicklund BH, Gustilo RB, Tsukayama DT. Prosthetic

metals interfere with the functions of human osteoblast cells

in vitro. Clin Orthop 1997;339:216–26.

[30] Fritz EA, Glant TT, Vermes C, Jacobs JJ, Roebuck KA.

Titanium particles induce the immediate early stress responsive

chemokines IL-8 and MCP-1 in osteoblasts. J Orthop Res 2002;

20(3):490–8.

ARTICLE IN PRESSM. Fini et al. / Biomaterials 24 (2003) 4929–4939 4939

[31] Sott AH, Rosson JW. The influence of biomaterial on patterns of

failure after cemented total hip replacement. Int Orthop

2002;26(5):287–90.

[32] Perizzolo D, Lacefield WR, Brunette DM. Interaction between

topography and coating in the formation of bone nodules in

culture for hydroxyapatite- and titanium-coated micromachined

surfaces. J Biomed Mater Res 2001;56:494–503.

[33] Kapanen A, Kinnunen A, Ryhanen J, Tuukkanen J. TGF-beta1

secretion of ROS-17/2.8 cultures on NiTi implant material.

Biomaterials 2002;23(16):3341–6.

[34] Linder L, Lundskog J. Incorporation of stainless steel, titanium

and vitallium in bone. Injury 1975;6:277–85.

[35] Millar BG, Frame JW, Browne RM. A histological study of

stainless steel and titanium screws in bone. Br J Oral Maxillofac

Surg 1990;28:92–5.

[36] Rocca M, Fini M, Giavaresi G, Nicoli Aldini N, Giardino R.

Tibial implants: biomechanical and histomorphometric studies of

hyroxyapatite-coated and uncoated stainless steel and titanium

screws in long term ovariectomized sheep. Int J Artif Organs

2001;24(9):649–54.

[37] Sun C, Huang G, Christensen FD, Dalstra M, Vergaard S,

Bunger . Mechanical and histological analysis of bone-pedicle

screw interface in vivo: titanium versus stainless steel. Chin Med

J (Engl) 1999;112(5):456–60.

[38] Fini M, Cadossi R, Can"e V, Cavani F, Giavaresi G, Krajewski A,

Martini L, Nicoli Aldini N, Ravaglioli A, Rimondini L, Torricelli

P, Giardino R. The effect of pulsed electromagnetic fields on the

osteointegration of hydroxyapatite implants in cancellous bone: a

morphologic and microstructural in vivo study. J Orthop Res

2002;20:756–63.

[39] Fini M, Giavaresi G, Rimondini L, Giardino R. Titanium alloy

osseointegration in cancellous and cortical bone of ovariecto-

mized animals: histomorphometric and bone hardness measure-

ments. J Oral Maxillofac Surg 2002;17(1):28–37.

[40] Huja SS, Katona TR, Roberts WE. Microhardness testing of

bone. In: An YH, Draughn RA, editors. Mechanical testing of

bone and the bone-implant interface. Boca Raton: CRC Press;

2000. p. 247–56.

[41] Stea S, Visentin M, Savarino L, Ciapetti G, Donati ME, Moroni

A, Caja V. Pizzoferrato A: microhardness of bone at the interface

with ceramic-coated metal implants. J Biomed Mater Res

1995;29:695–9.

[42] Moroni A, Caja VL, Maltarello MC, Savarino L, Marinelli F,

Stea S, Visentin M, Giannini S. Biochemical, scanning electron

microscopy, and microhardness analyses of the bone pin interface

in hydroxyapatite coated vs. uncoated pins. J Orthop Trauma

1997;11:154–61.

[43] Savarino S, Stea S, Ciapetti G, Paganetto G, Donati ME,

Alvergna P, Pizzoferrato A. Microstructural investigation

of bone cement interface. J Biomed Mater Res 1995;29:

701–5.

[44] Huja SS, Katona TR, Moore BK, Roberts WE. Micro-

hardness and anisotropy of the vital osseous interface and

endosseous implants supporting bone. J Orthop Res 1998;16:

54–60.

[45] Albrektsson T, Johansson C. Osteoinduction, osteoconductionm

and osseointegration. Eur Spine J 2001;10:S96–101.

[46] Albrektsson T. The healing of autologous bone grafts after

varying degrees of surgical trauma. J Bone Jt Surg 1980;32:

403–10.

Documents

A new austenitic stainless steel with negligible nickel content: an in vitro and in vivo comparative investigation