Growth inhibition of cultured smooth muscle cells bycorrosion products of 316 L stainless steel wire

Chun-Che Shih,1,2 Chun-Ming Shih,3,4 Yuh-Lien Chen,5 Yea-Yang Su,6 Jeng-Shong Shih,7 Ching-Fai Kwok,8

Shing-Jong Lin1,9,10

1Institute of Clinical Medicine, National Yang-Ming University School of Medicine, Taipei 112, Taiwan2Division of Cardiovascular Surgery, Taipei Veterans General Hospital, Taipei 112, Taiwan3Graduate Institute of Medical Science, Taipei Medical University School of Medicine, Taipei 110, Taiwan4Division of Cardiology, Department of Internal Medicine Taipei Medical University Hospital, Taipei 110, Taiwan5Institute of Anatomy and Cell Biology, National Yang-Ming University, Taipei 112, Taiwan6Amorphous Technologies Inc., Marietta, Georgia7Department of Chemistry, National Taiwan Normal University, Taipei 112, Taiwan8Division of Endocrinology and Metabolism, Taipei Veterans General Hospital, Taipei 112, Taiwan9Division of Cardiology, Taipei Veterans General Hospital, No. 201, Sec. 2, Shih-Pai Road, Taipei 112, Taiwan10Cardiovascular Research Center, National Yang-Ming University, Taipei 112, Taiwan

Received 21 August 2000; revised 7 March 2001; accepted 30 March 2001

Abstract: The potential cytotoxicity on vascular smoothmuscle cells of corrosion products from 316 L stainless steel,one of most popular biomaterials of intravascular stents, hasnot been highlighted. In this investigation, 316 L stainlesssteel wires were corroded in Dulbecco’s modified eagle’smedium with applied constant electrochemical breakdownvoltage, and the supernatant and precipitates of corrosionproducts were prepared as culture media. The effects of dif-ferent concentrations of corrosion products on the growth ofrat aortic smooth muscle cells were conducted with the [3H]-thymidine uptake test and cell cycle sorter. Both the super-natant and precipitates of corrosion products were toxic tothe primary culture of smooth muscle cells. The growth in-hibition was correlated well with the increased nickel ions in

the corrosion products when nickel concentration was above11.7 ppm. The corrosion products also changed cell mor-phology and induced cell necrosis. The cell growth inhibi-tion occurred at the G0/G1 to S transition phase. Similar toour recent study of nitinol stent wire, the present investiga-tion also demonstrated the cytotoxicity of corrosion prod-ucts of 316 L stainless steel stent wire on smooth musclecells, which might affect the poststenting vascular response.© 2001 John Wiley & Sons, Inc. J Biomed Mater Res 57:200–207, 2001

Key words: corrosion; 316 L stainless steel; cytotoxicity;nickel; smooth muscle cell

INTRODUCTION

Biocompatibility has evolved from the previous no-tion of inert material to a more recent concept basedon the ability of a material to perform with an appro-priate host response in a specific environment. Thereare several important factors involved in the design of

an optimal intravascular stent. Most of the mechanicalproperties are related to the bulk characteristics of themetal, and those linked to surface properties. Earlybiocompatibility problems with intravascular stentsare associated with thrombosis, inflammation, andneointima formation. Late problems with stents aremechanical failure because of material fatigue result-ing from the considerable stress imposed to the stentby cardiac contractions, and chemical failure wherecorrosion or depolymerization can release potentiallytoxic substances such as nickel ions, degradationproducts, or contaminants.1

Intimal hyperplasia is one of the major mechanismsresponsible for poststenting restenosis, and is mainlycaused by the proliferation of vascular smooth musclecells.1,2 But the effects of electrochemical corrosion ofstent materials on the tissue response, especially on

Correspondence to: S.-J. Lin; e-mail: sjlin@vghtpe.gov.twContract grant sponsor: National Science Council, Taiwan;

contract grant number: NSC-89-2314-B-010-024-M55Contract grant sponsor: Yen-Tjing-Ling Medical Founda-

tion, Taiwan; contract grant number: CI-89-7-1Contract grant sponsor: Taipei Veterans General Hospital,

Taiwan; contract grant numbers: VGH-218, VGH-254,VGH-362

© 2001 John Wiley & Sons, Inc.

the growth of vascular smooth muscle cells, had notbeen well studied. The release of oxides and metalions could be detected in the first few days immedi-ately after the stent implantation.2–4 However, corro-sive breakdown of stents might take place any time,even several years after implantation. At the tissuecontact interface, the released ions may be accumu-lated locally and then taken up by the adjacent vascu-lar cells.5 Therefore, the cytotoxicity of these corrosionproducts on vascular smooth muscle cells needed tobe evaluated.

The cell growth around pieces of metal in culturedishes could be observed and compared with that incontrol cultures. High concentrations (15–30 ppm) ofeither cobalt or nickel have been shown to cause mor-phological changes and inhibit cell growth of culturedfibroblasts.6,7 In a recent study, we demonstrated thecytotoxicity of corrosion products of nitinol stent wireon cultured smooth muscle cells.8 In the present in-vestigation, we aimed to study the effects of corrosionproducts of 316 L stainless steel wire on the growthand morphology of cultured rat aortic smooth musclecells. In addition, the affected phases of cell replicationwere analyzed by cell cycle sorter.

MATERIALS AND METHODS

Corrosion occurrence by usingpotentiostatic control

We used a three-electrode system and recorded with acomputer program of Potentiostat (EG & G model 273; NJ),as described previously.8–12 The corrosion of the 316 L stain-less steel wire was enhanced by applying a constant poten-tial higher than the predicted breakdown potential.

Preparation of culture media by collectingcorrosion products with or withoutapplied potentials

Sample wires were degreased with 70% ethanol in an ul-trasonic vibrobath and sterilized by exposure to a dose of 2.5× 104 Gy radiation before each experiment. The 316 L stain-less steel wires used in this study have a polycrystallineoxide surface. The surface oxide of 316 L stainless steel wireshas been determined by transmission electron microscopy,selected area diffraction, and scanning electron micros-copy.8–12 The chemical composition of the polycrystallineoxide film was determined by the energy dispersion X-rayanalysis or energy dispersion spectrum. Five 316 L stainlesssteel wires of 0.25-mm diameter and 15-cm length were usedin this study.

After filling the plastic centrifugal tubes with 50 mL ofDulbecco’s modified Eagle’s medium (DMEM), three elec-trode wires were immersed separately. Different potentials

were applied to the samples and corresponding current den-sities were recorded. The mild leaching medium was pre-pared by immersing 316 L stainless steel wire in DMEMwithout any applied potential for 24 h at 37°C. The severecorrosion media were prepared with 316 L stainless steelwires immersed and a constant potential applied at +1.0 Vversus AgCl electrode by potentiostatic control for 6 h at37°C. As described previously,8–12 the wires of stainless steelor nitinol covered with polycrystalline oxide usuallyshowed corrosion breakdown at relatively low potentialsbetween +0.2 to +0.6 V versus AgCl reference electrode. Po-tential was held continuously until the 316 L stainless steelwire had been completely dissolved. These media containedboth the solid precipitated corrosion products and the dis-solved ion solutes. After centrifugation, the supernatant andthe precipitates were separated by centrifugation and col-lected, respectively. The corrosion precipitates were resus-pended with standard DMEM. The DMEM, free from anycorrosive metallic ions, was used as the control assay.

Nickel ion concentrations in these specially prepared me-dia were measured by a graphite atomic absorption spectro-photometer. To all the media were added the same amountof 10% fetal calf serum and 1% penicillin (100 IU/mL) plusstreptomycin (100 mg/mL), and pH value was maintainedbetween 7.30–7.35 after sterilization.

Aortic smooth muscle cell cultures

Rat aortic smooth muscle cells were isolated and culturedaccording to a modified method claimed by Shih et al.8 andKwok et al.13 About 105 cells, between the 6th and 16thpassage, were cultured in 12-well Falcon tissue cultureplates (2.9 cm2/well) filled with DMEM in a humidifiedincubator at 37°C in the presence of 5% CO2. The cell growthand morphology were examined regularly with a phase con-trast microscope. The cells were characterized as smoothmuscle cells by morphological criteria,14 and with a murineanti-a-actin antibody to recognize a unique epitope ofa-smooth muscle actin.

The effects of corrosion products on the growth ofsmooth muscle cells

After the cells had been growing to occupy above 90% ofthe wells, serum-free condition was maintained for 12 h.One negative control (polystyrene patch-immersed DMEM)and three specially prepared culture media (namely, mediawith wire immersed and without any applied potential, me-dia with supernatant portion of corrosion products after ap-plied 1.0 V potential, and media with precipitate portion ofcorrosion products after applied 1.0 V potential) werepoured into 12-well plates.

Twelve hours later, [3H]-methyl thymidine was addedwith some plates of cultured cells for the DNA incorporationtest. The others were trypsinized down the cells from eachwell and stored in 10% dimethylsulfoxide with fetal calfserum and DMEM at −70°C for further cell cycle analysis.

201CYTOTOXICITY OF 316 L STAINLESS STEEL CORROSION

[3H]-thymidine uptake test

The radiolabeled [3H]-thymidine uptake was used as anindex of cell proliferation. DNA synthesis was quantified bymeasuring [3H]-thymidine uptake as described previ-ously.8,13 Wells with standard DMEM added were used ascontrols. The negative control media, mild leaching media,and different dilution media of severe corrosion were pre-pared and tested simultaneously. The smooth muscle cells,grown in 12-well plates, were incubated with the addition of1 mCi/mL [3H]-thymidine at 37°C for 2 h. The exposure to[3H]-thymidine was terminated by washing the cells twice,replacing the media with ice-cold phosphate buffer saline(pH 7.2 without Ca2+ and Mg2+). After solubilizing the cellsin 0.1% sodium dodecyl sulphate for 30 min, aliquots weretaken for protein assay and liquid scintillation counting. TheDNA was precipitated with ice cold 10% trichloroacetic acidand filtered with GF/C glass microfibre filters (Whatman 24mm w circle). The radioactivity of aliquots was measured ina Beckman LS-6500 liquid scintillation counter. Protein con-tent of the solubilized cells in each well was determined bythe method suggested by Shih et al.8 and Kwok et al.13 withsome modifications. Briefly, 0.1-mL aliquots of solubilizedcell were mixed and reacted with 0.8 mL of working solutionA (2% Na2CO3 in 0.1N NaOH: 1% CuSO4: 2% NaK = 100:1:1)for 10 min. Then, after adding 0.1 mL of solution B (Folin-phenol: H2O = 1:1) and standing for 30 min, the resultingsolutions were checked for absorbance at 562 nm using aBeckmann DU-64 spectrophotometer. Standards weremixed using bovine serum albumin from 0 to 0.6 mg/mL.

Cell cycle analysis

DNA ploidy and content were analyzed by flow cytom-etry (Epics Elite; Coulter Company, Hialeah, FL) followingthe method of Dressler et al. with some modifications.8,15

The frozen cells from each well, which were stored at −70°C,were defrosted to 37°C immediately, and then centrifugedfor 30 s and the supernatant solution discarded. A 0.5-mLstaining solution A [50 mg/mL propidium iodide (no.p-4170; Sigma Chemical Co., St. Louis, MO), 3% polyethyl-eneglycol 6000 (ART 807491; Merck, Darmstadt, Germany),0.1% triton X-100, 180 U/mL RNAase (no. R-5503; Sigma), 4mM citrate buffer, pH 7.2] was added, incubated at 37°C for20 min, followed by the addition of 0.5 mL of staining solu-tion B [50 mg/mL propidium iodide, 3% polyethyleneglycol6000, 0.1% triton X-100, 0.4M NaCl, pH 7.2]. Samples werethen kept at 4°C for 1 h and filtered before being analyzedusing a flow cytometer.

The time-course and dose-response studies of[3H]-thymidine uptake inhibition by severecorrosion media

Five culture media with the supernatant of severe corro-sion of 316 L stainless steel wires were mixed together andthe final nickel concentration was detected. This culture me-

dium was then diluted further and separated into 1×, 1/2×,1/4×, and 1/8× dilution groups for time-course and dose-response studies. During these studies, the incubation pe-riod for each different dilution of corrosion medium was 12h. At time intervals of 0, 6, 8, 10, and 11 h, culture media ofnear confluent smooth muscle cells in every two wells of12-well plates were replaced with standard DMEM, negativecontrol polyethylene media, or severe corrosion media ofdifferent dilutions simultaneously and sequentially with du-plicate individual samples. [3H]-thymidine uptake was de-termined at each time point of different dilutions.

The detection of nickel dissolution

The nickel ion concentrations in the corrosion media wereanalyzed by using a graphite atomic absorption spectropho-tometer, GFAAS (Hitachi Z 8100, Japan), with deuteriumand polarized Zeeman-effect background correction system.Medium samples were analyzed according to the thermalprograms described in Table I with the reference setting at232.0-nm wavelength, 10-mA lamp current, 0.4-nm slit, and20-mL sample volume.

STATISTICS

Nickel ion concentrations were expressed as ppm, anddata of [3H]-thymidine uptake were calculated as absolutevalue in c.p.m. per mg of protein. The effects of negativecontrol, and corrosion products of different treatments oncell proliferation were expressed as the ratio of [3H]-thymidine uptake compared with control uptake of smoothmuscle cells cultured in standard DMEM. Six experimentswere performed on triplicate individual samples.

In the time-course and dose-response curves, the [3H]-thymidine uptake tests of different dilutions of corrosionmedia at different time intervals were expressed as indi-vidual points. Each point was the mean of six measurementson duplicate individual samples and expressed as mean ±standard deviation of the percentage of uptake in the controlgroup. The cell cycle phase analysis was performed on du-plicate individual samples with a total of six experiments.

The difference among the different dilution media andnegative control (DMEM with polyethylene) were comparedby analysis of variance, using a Scheffe test for post hoc com-

TABLE IThe Thermal Programs for Nickel Ion Detection in

Culture Media by Graphite AtomicAbsorption Spectrophotometry

Steps Stages

Temperature (°C)Time

(s)Carrier Gas(mL/min)Start End

1 Dry 80 120 30 2002 Ash 1200 1200 30 2003 Atom 2850 2850 10 304 Clean 3000 3000 4 200

202 SHIH ET AL.

parisons. A p < 0.05 was considered as statistically signifi-cant.

RESULTS

Severe corrosion of 316 L stainless steel wires wasinduced by applying potentiostatic control at +1.0 VAgCl, and the response current density was recordedas shown in Figure 1. Nickel concentrations were clas-sified at different test media: 1.84 ppm for the wire-immersed mild leaching media, 93.2 ppm for the su-pernatant portion of severe corrosion, and 87.5 ppmfor the resuspended precipitated portion of severe cor-rosion, respectively. The concentration of nickel ionsin the supernatant of mixed media before dilution fortime-course and does-response studies was 93.2 ppm.

The corrosion products of 316 L stainless steel weretoxic for the primary culture of rat aortic smoothmuscle cells. Morphologically, under a phase contrastmicroscope (Fig. 2), cellular death appeared after in-cubation with the corrosion media and the affectedcells increased along with the nickel concentration in-crease and exposure time.

The wire-immersed mild leaching media with lessthan 1.84 ppm of nickel ion had no inhibition effect on[3H]-thymidine uptake by cultured smooth musclecells. However, corrosion products of severe corrosiondisplayed significant inhibition of cell replication asshown in Figure 3. Both the supernatant and precipi-tated products have similar inhibition effects on thecell replication.

The time-course and dose-response curves of [3H]-thymidine uptake ratio of cultured smooth musclecells are shown in Figure 4. The replication inhibitioninitiated immediately after changing media and

reached the maximal effect 4–6 h later. This inhibitionphenomenon could be observed even in the 1/8×dilution media with a nickel concentration of 11.7ppm.

Cell numbers in severe corrosion media were sig-nificantly lower than that in wire-immersed mildleaching media, which showed no significant differ-ence compared with standard DMEM and negativecontrol media (Table II). Cell growth in severe corro-sion media was inhibited during the transition ofG0-G1 phase to S1 phase by cell cycle analysis (Ta-ble II).

DISCUSSION

Currently, few cellular biocompatibility data of 316L stainless steel are available for arterial smoothmuscle cells. Histological studies16,17 have shown thatthe constituent elements of alloys could be detected inthe local tissues and the tissue reaction around an al-loy was related to the concentration of metal ions re-leased into the tissues. Local tissues at the site of animplanted prosthetic alloy are exposed continuouslyto gradually accumulated concentrations of the metal-lic ions comprising the alloy. However, the exact iden-tity and concentration of such metallic products areusually unknown, thus limiting the possibilities forquantifying any observed tissue response to the met-als.6 Despite the satisfactory clinical use of 316 L stain-less steel alloy in some situations, its biocompatibilityis still not well documented and conflicting conclu-sions are derived from in vivo experiments undertakenfor 316 L stainless steel alloy prosthetic components.5

The in vitro cell cultivation with alloy immersion is

Figure 1. The occurrence of 316 L stainless steel corrosion recorded for 4–6 h by Potentiostat. (a) Flat linear curve withoutconduction of any potential voltage and nickel dissolution in media with a concentration of 1.02 ppm. (b) Curve of severecorrosion after conduction of 1.0 V and nickel dissolution in media with a concentration of 91.8 ppm.

203CYTOTOXICITY OF 316 L STAINLESS STEEL CORROSION

a widely used investigation method for biocompatibil-ity study because it is simultaneous, comparative, andrapidly repeatable. However, the immersion corrosionrate is unreliable even on the unique specimen and in

the same immersion fluid, and it tends to be decreasedwith the time after immersion.18,19 On the contrary,culturing cells with specially prepared culture mediaof alloy corrosion products is a reliable approach for in

Figure 2. Rat aortic smooth muscle cells observed under a phase contrast microscope after incubation with DMEM con-taining corrosion products of 316 L stainless steel wires for 12 h. (A, B) Confluence culture of smooth muscle cells incubatedwith negative control polystyrene patch-immersed medium. Original magnification: (A) ×100; (B) ×200. (C, D) Confluenceculture of smooth muscle cells incubated with wire-immersed mild leaching media without any applied potential (nickel: 1.84ppm). Original magnification: (C) ×100; (D) ×200. (E, F) Large numbers of dead smooth muscle cells appeared after incubationwith supernatant portion of severe corrosion products (nickel: 93.2 ppm). Original magnification: (E) ×100; (F) ×200. (G, H)Large numbers of dead smooth muscle cells appeared after incubation with resuspended precipitate portions of severecorrosion products (nickel: 87.5 ppm). Original magnification: (G) ×100; (H) ×200.

204 SHIH ET AL.

vitro cellular response evaluation. By using thismethod in a recent study, we demonstrated the cyto-toxicity of corrosion products of nitinol stent wire oncultured smooth muscle cells.8 In the present investi-gation, the corrosion products of 316 L stainless steelwire have also been shown to inhibit the growth ofprimary culture of rat aortic smooth muscle cells.

Because thrombosis, inflammation, and neointimalhyperplasia are all involved in the development ofpoststenting restenosis, further studies would be nec-essary to look into the poststening vascular adverseeffects from 316 L stainless steel stents if they corrode.

Mechanism of 316 L stainless steel corrosion

Whenever metal is exposed to a fluid, a film of cor-rosion products is formed on the surface. If the film isrelatively impermeable and adherent, it will inhibitfurther corrosion. If it is permeable to ions either in thefluid or in the metal beneath the film or to both, and ifthe electrochemical conditions are favorable, there willbe continuing corrosion. Initially, the corrosion prod-

ucts remain on the metal surface and inhibit continu-ing corrosion. The corrosion progress depends ontheir solubility and chemical reaction with the envi-ronment. If the saturated corrosion products havebeen precipitated, then the corrosion progress de-pends on the site of precipitation and the permeabilityof these products.4

This protective layer is mainly composed of metaloxides. The formation of two types of metal oxide,namely, polycrystalline oxide, and amorphous oxide,depending on the compositions of the metal or alloymaterials and passivated processes used in surface fin-ishing. The rate of chemical and electrochemical reac-tions relies heavily on the properties of this oxide layeron the surface of metal or alloys. The current commer-cial intravascular stents of 316 L stainless steel or niti-nol with polycrystalline oxide have been proven to beless corrosion resistant.6,8–12 Our previous observationhas also revealed that 316 L stainless steel wire cov-ered with polycrystalline oxide usually shows break-down at the relatively low potentials between +0.2 to+0.6 V versus AgCl reference electrode.8–12 Therefore,under the consistently applied potential of 1.0 V, wireswere completely dissolved in the DMEM medium.Thus, the corrosion products can be produced expe-ditiously and controllably by electrochemical poten-tial. The matrix of 316 L stainless steel with low carboncontent (0.03% weight) is predominantly iron (60 to

Figure 3. [3H]-thymidine uptake ratio in rat aortic smoothmuscle cells incubated with different corrosion culture me-dia. (A) Negative control polystyrene thread in DMEM for24 h. (B) 316 L stainless steel wires with polycrystalline oxide(PO) surface in DMEM without applied potential for 24 h(nickel: 1.84 ppm). (C) 316 L stainless steel wires with POsurface in DMEM with applied potential of 1.0 V (the su-pernatant of corrosion products) (nickel: 93.2 ppm). (D) 316L stainless steel wires with PO surface in DMEM with ap-plied potential of 1.0 V (the resuspension of precipitatedcorrosion products with DMEM) (nickel: 87.5 ppm). Cellswere incubated with media A, B, C, or D for 12 h, then[3H]-thymidine uptake was measured. Results are expressedas mean ± SD of [3H]-thymidine uptake ratio as comparedwith control uptake in standard DMEM. Each bar representsthe average of six determinations performed on triplicateindividual samples. Both the supernatant and the precipi-tates of severe corrosion significantly reduced the [3H]-thymidine uptake by smooth muscle cells. *p < 0.01

Figure 4. The time-course and dose-response curves of[3H]-thymidine uptake ratio in rat aortic smooth muscle cellsafter incubation with different dilutions of corrosion mediaof 316 L stainless steel [nickel: 1× = 93.2 ppm (closed circle),1/2× = 46.6 ppm (open circle), 1/4× = 23.3 ppm (closedtriangle), 1/8× = 11.7 ppm (open triangle), negative controlpolyethylene media (closed square)] for up to 12 h. [3H]-thymidine uptake was measured at different time points anddata are expressed as mean ± standard deviation of [3H]-thymidine uptake ratio as compared with control uptake atstart point. Each point represents the average of six deter-minations performed on duplicate individual samples. *p <0.05 as compared with that of each corresponding time pointof negative control polystyrene media.

205CYTOTOXICITY OF 316 L STAINLESS STEEL CORROSION

65%) mixed with chromium (17 to 18%) and nickel (12to 14%).18 In accord with the chemical compositionsand electrochemical characteristic of surface oxidesand the matrix of 316 L stainless steel (Table III), thesupernatant consisted of saturated solution of nickelions after centrifugation and the precipitates are solidcorrosion products, which comprise ferrous and chro-mium oxides.

Several studies have found nickel ions to be cyto-toxic as well as carcinogenic.7,20–23 The released nickelions from corrosion may act as cofactors or inhibitorsin enzymatic processes involved in protein synthesisand cell replication,1 disrupt intracellular organelles,alter morphology, and decrease cell numbers.24–29

The dose response and time course of growthinhibition by 316 L stainless steelcorrosion products

In cases of mild leaching or severe corrosion, thereleased ions of metal corrosion would influence thecell growth if the local nickel concentration was higherthan the cell tolerance level. As one of the major con-stituents of 316 L stainless steel matrix, though little insurface polycrystalline oxide (Table III), nickel ion dis-solution was proportional to the severity of corrosionand was identified to be hazardous to vascular smoothmuscle cells. Nickel concentrations higher than 11.7ppm (1/8× dilution) displayed growth inhibition oncultured smooth muscle cells, as demonstrated by the[3H]-thymidine uptake test and cell cycle analysis.

The dose-response and time-course studies (Fig. 4)have clearly shown that the degree of cytotoxicity of

316 L stainless steel corrosion increased with exposuretime and dose increment. The growth inhibitionreached maximal effect at 4 to 6 h after incubation.

The inhibition on transition of G0-G1 phase to Sphase of cell cycle

Shown from the [3H]-thymidine uptake test, the cellreplication of aortic smooth muscle cells was signifi-cantly inhibited by the corrosion products of 316 Lstainless steel wire. Cell cycle analyzed by flow cytom-etry had further displayed that the inhibition processoccurred during the transition of G0-G1 phase to Sphase, which caused the diminution of cell numbers(Table II). A number of growth factors and cytokinesmay be involved during this transition period and af-fect the cell cycle manifestation, the cell viability, andproliferation. To avoid potential nickel cytotoxicity,enhancement of corrosion resistance of 316 L stainlesssteel by surface treatment with either preliminary oxi-dation, or melt-spraying with biocompatible materialor surface morphological modification during the finalmanufacturing process of cardiovascular stent is veryimportant and needs further investigation.

CONCLUSION

Similar to nitinol stent wire, both the supernatantand the precipitated corrosion products of current 316L stainless steel wire with a polycrystalline oxide sur-face are potentially toxic to vascular smooth musclecells, especially when the released nickel concentra-tion is higher than 11.7 ppm. The growth of culturedsmooth muscle cells was inhibited during the transi-tion of G0-G1 phase to S phase. The cell number dimi-nution with morphology alteration, or even cell deathwas more prominent along with the time exposureand dose increment of corrosion media.

The authors are indebted to Miss Ruei-Chi Chou at theDepartment of Chemistry, National Taiwan Normal Univer-

TABLE IIIThe Compositions (Weight Percentage) of 316 L

Stainless Steel9,12

Elements Polycrystalline Oxide Matrix

Chromium 5.77 17–18Nickel 2.05 12–14Molybdenum 1.57 MinimalOxygen 32.69 MinimalIron Balance 60–65

TABLE IICell Numbers and Cell Cycle Analysis of Rat Aortic Smooth Muscle Cells Incubated with Different Corrosion Media

Total Cell CountG0/G1

Phase (%) S Phase (%)G2 + M

Phase (%)

Standard DMEM 120235 ± 9154 75.1 ± 8.3 18.9 ± 6.5 5.7 ± 2.8Negative control polystyrene media 124174 ± 19721 79.6 ± 5.3 9.2 ± 2.8 11.3 ± 3.5Mild leaching media (0 V) 128296 ± 21313 76.8 ± 8.4 17.4 ± 6.7 5.8 ± 2.7Media of supernatant of severe corrosion (1.0 V) 80987 ± 12878* 86.7 ± 6.3 7.8 ± 3.2* 5.4 ± 4.3Media of resuspended precipitates of severe corrosion (1.0 V) 80185 ± 10257* 89.6 ± 2.4* 6.0 ± 1.5* 4.4 ± 2.2

Six experiments were done on duplicate individual samples.*p < 0.05 as compared with negative control.

206 SHIH ET AL.

sity, for her invaluable advice and assistance with the analy-sis of graphite atomic absorption spectrophotometry. Theauthors are also indebted to Ms. Lee-Wha Wu of the Divi-sion of General Surgery, Taipei Veterans General Hospital,Taipei, Taiwan for her invaluable advice and assistance inusing flow cytometry.

References

1. Bertrand OF, Sipehia R, Mongrain R, Rodes J, Tardif J-C, Bilo-deau L, Cote G, Bourassa MG. Biocompatibility aspects of newstent technology. J Am Coll Cardiol 1998;32:562–571.

2. Ryhanen J, Niemi E, Serlo W, Niemela E, Sandvik P, Pernu H,Salo T. Biocompatibility of nickel-titanium shape memorymetal and its corrosion behavior in human cell culture.J Biomed Mater Res 1997;35:451–457.

3. Putters JLM, Sukul K, Zeeuw GR, Bijma A, Besselink PA. Com-parative cell culture effects of shape memory metal (Nitinol),nickel, and titanium: A biocompatibility estimation. Eur SurgRes 1992;24:378–382.

4. Pappas AM, Cohen J. Toxicity of metal particles in tissue cul-ture. Part I. A new assay method using cell counts in the phaseof replication. J Bone Joint Surg 1968;50:535–547.

5. Oshida Y, Miyazaki S. Corrosion and biocompatibility of shapememory alloys. Corr Eng 1991;40:1009–1025.

6. Bearden LJ, Cooke FW. Growth inhibition of cultured fibro-blasts by cobalt and nickel. J Biomed Mater Res 1980;14:289–309.

7. Wataha JC, Hanks CT, Craig RG. The in vitro effects of metalcations on eukaryotic cell metabolism. J Biomed Mater Res1991;25:1133–1149.

8. Shih CC, Lin SJ, Chen YL, Su YY, Lai ST, Wu JW, Kwok CF,Chung KH. The cytotoxicity of corrosion products of nitinolstent wire on cultured smooth muscle cells. J Biomed MaterRes 2000;52:395–403.

9. Su YY, Raman V. The quest for wire surface quality for medicalapplications. Wire J Int 1998;31:106–113.

10. Su YY, Raman V. The quest for nitinol wire surface quality formedical applications. Proceedings of the Second InternationalConference on Shape Memory and Superelastic Technologies.Conference; 1997 March; Asilomar, California. p 389–394.

11. Su YY, Marek M. Applications of electrochemistry in the wireindustry. Wire J Int 1996;29:78–83.

12. Shih CC, Lin SJ, Chung KH, Chen YL, Su YY. Increased cor-rosion resistance of stent materials by converting current sur-face film of polycrystalline oxide into amorphous oxide.J Biomed Mater Res 2000;52:323–332.

13. Kwok CF, Goldstein BJ, Muller-Weiland D, Lee T-S, Kahn CR,King GL. Identification of persistent defects in insulin receptorstructure and function in capillary endothelial cells from dia-betic rats. J Clin Invest 1989;83:127–136.

14. Ross R. The smooth muscle cell. II. The growth of smoothmuscle in culture and formation of elastic fibers. J Cell Biol1971;50:172–178.

15. Dressler LG, Seamer L, Owens MA, Clark GM, McGuire WL.Evaluation of a modeling system for S-phase estimation inbreast cancer by flow cytometry. Cancer Res 1987;47:5294–5302.

16. Ferguson AB, Laing PG, Hodge ES. The ionization of metalimplants in living tissues. J Bone Joint Surg 1960;42:77–90.

17. Ferguson AB, Akahoshi Y, Yoshihiko K, Laing PG, Hodge ES.Characteristics of trace ions released from embedded metalimplants in the rabbit. J Bone Joint Surg 1962;44:323–336.

18. Lu S. Medical applications of Ni-Ti alloys in China. In: DuerigTW, editor. Engineering aspects of shape memory alloys. Lon-don: Butterworth-Heinemann; 1990. p 445.

19. Lemons JE. General characteristics and classification of implantmaterial. In: Lin OCC, editor. Perspectives on biomaterials.Amsterdam: Elsevier; 1986. pp. 1–13.

20. Webb M, Weinzierl SM. Uptake of Ni++ from its complexeswith proteins and other ligands by mouse dermal fibroblasts invitro. Br J Cancer 1972;26:292–298.

21. Swierenga SHH, Basrur PK. Effect of nickel on cultured ratembryo muscle cells. Lab Invest 1968;19:663–674.

22. Bensimon J, Rosenfield C. Influence of du sulfate de nickel surla croissance de deux lignees lymphoblastoides humainesD’origine normale et Leucemique. C R Acad Sci Hebd SeancesAcad Sci D 1974;278:345–348.

23. Sunderman FW. Carcinogenic effects of metals. Fed Proc 1978;37:40–46.

24. Rae T. A study on the effects of particulate metals of orthope-dic interest on murine macrophage in vitro. J Bone Joint Surg1975;57B:444–450.

25. Jacobsen N. Epithelial-like cells in culture derived from humangingiva: Response to nickel. Scand J Dent Res 1977;85:567–573.

26. Kawahara H. Cellular response to implant materials: Biologi-cal, physical, and chemical factors. Int Endod J 1983;33:350–375.

27. Evans EJ, Thomas IT. The in vitro toxicity of cobalt-chrome-molybdenum alloy and its constituent metals. Biomaterials1986;7:25–29.

28. Craig RG, Hanks CT. Cytotoxicity of experimental casting al-loys evaluated by cell culture tests. J Dent Res 1990;69:1539–1542.

29. Burstein GT. The effects of anions on the behaviors ofscratched iron electrodes in aqueous solution. Corros Sci 1980;20:1143–1155.

207CYTOTOXICITY OF 316 L STAINLESS STEEL CORROSION