Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes

Acta Biomaterialia xxx (xxxx) xxx

Contents lists available at ScienceDirect

Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat

Full length article

Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes B. Stojanovic´ a, C. Bauer b, C. Stotter b,c, T. Klestil c,d, S. Nehrer b, F. Franek a, M. Rodríguez Ripoll a,⇑ a

AC2T research GmbH, Viktor Kaplan-Straße 2/C, A-2700 Wiener Neustadt, Austria Danube University Krems, Faculty of Health and Medicine, Department for Health Sciences and Biomedicine, Center for Regenerative Medicine and Orthopedics, Dr. Karl-Dorrek-Str. 30, A-3500 Krems, Austria c LK Baden-Mödling-Hainburg, Department of Orthopedics and Traumatology, Waltersdorferstraße 75, A-2500 Baden, Austria d Danube University Krems, Faculty of Health and Medicine, Department for Health Sciences and Biomedicine, Center for Medical Specializations, Dr. Karl-Dorrek-Str. 30, A-3500 Krems, Austria b

a r t i c l e

i n f o

Article history: Received 5 February 2019 Received in revised form 11 June 2019 Accepted 13 June 2019 Available online xxxx Keywords: Cartilage Biotribology Metal ion release Partial implants CoCrMo alloy

a b s t r a c t Partial knee replacement and hemiarthroplasty are some of the orthopedic procedures resulting in a metal on cartilage interface. As metal implant material, CoCrMo based alloys are commonly used. The aim of the present study is to assess the role of biotribocorrosion on the CoCrMo-cartilage interface with an emphasis on metal release during sliding contact. The biotribocorrosion experiments were performed under controlled electrochemical conditions using a floating cell with a three electrode set up coupled to a microtribometer. Throughout the experiment the coefficient of friction and the open circuit potential were monitored. Analyses of the electrolyte after the experiment show that metal release can occur during sliding contact of CoCrMo alloy against articular cartilage despite the extraordinary low coefficient of friction measured. Metal release is attributed to changes in passive layer caused at the onset of sliding. The released metal was found to be forming compounds with potential cytotoxicity. Since the presence of metal ions in the cartilage matrix can potentially lead to cell apoptosis, the metabolic activity of human osteoarthritic chondrocytes (2D-cultures) was investigated in the presence of phosphate buffered saline containing metal ions using XTT-assay. The experiments indicate that critical concentrations of Co ions lead to a significant decrease in chondrocyte metabolic activity. Therefore, biotribocorrosion is a mechanism that can occur in partial replacements and lead to chondrocyte apoptosis thus playing a role in the observed accelerated degradation of the remaining cartilage tissue after the mentioned orthopedic procedures. Statement of Significance Partial replacements provide an alternative to total joint replacements. This procedure is less invasive, allows a faster rehabilitation and provides a better function of the joint. However, the remaining native cartilage experiences accelerated degradation when in contact with metallic implant components. This work investigates the role of tribocorrosion at the metal-cartilage interface during sliding. Tribocorrosion is a degradation process that can alter significantly the wear rates experienced by metallic implants and lead to the release of metal ions and particles. The released metal can form compounds with potential cytotoxicity on cartilage tissue. The knowledge gained in this work will serve to understand the mechanisms behind the failure of partial replacements and develop future biomaterials with an enhanced lifetime. Ó 2019 Acta Materialia Inc. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

⇑ Corresponding author. E-mail address: [email protected] (M. Rodríguez Ripoll).

Very complex physiology, limited regeneration and repair due to its non-vascular matrix requires specific treatments to compensate the cartilage defects. When regenerative cartilage treatments cannot be performed the joint is replaced with an artificial one.

https://doi.org/10.1016/j.actbio.2019.06.015 1742-7061/Ó 2019 Acta Materialia Inc. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

2

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

The most common standard procedure for people with severe cartilage defects is a total joint replacement (TJR) which would allow them to normally perform day-to-day activities. As the number of younger patients with cartilage injuries increases, there is a need for new inventions that would enable them a long-term solution and better quality of life. Besides tissue engineering, recent technologies introduced partial replacements as an alternative to TJR. The surgical procedure is less invasive, allowing faster rehabilitation, and providing better function of the joint. On the other side, it has been widely observed that the remaining native cartilage left after hemiarthroplasty, total knee arthroplasty without patella resurfacing and other partial resurfacing procedures which create a metal-on-cartilage interface, experiences accelerated wear when in contact against metallic implants [1,2]. Articular joints have impressive tribological properties able to reach coefficients of friction as low as 0.001 [3]. In hemiarthroplasty, the replacement of half of the affected joint by an orthopedic biomaterial alters this configuration. The selected biomaterial is expected to maintain the excellent frictional properties of the joint without degrading during its lifetime while simultaneously preserving the remaining native cartilage against degradation. Several studies have investigated in vitro the frictional properties of prospective implant materials, such as metals, ceramics, and polymers against articular cartilage [4,5]. Metallic implants are usually made of passive metals in order to prevent corrosion during their service life. One of the most commonly used biomaterials for cartilage resurfacing are CoCrMo alloys due to their superb corrosion resistance, good biocompatibility, and excellent mechanical properties. When in contact with body fluids, CoCrMo alloys spontaneously passivate by building up a thin oxide film on their surface, which prevents intergranular corrosion and improves their biocompatibility. The passivation behavior of CoCrMo alloys in reference electrolytes, such as simulated body fluid was studied in detail using electrochemical methods. Nowadays it is known that the passive film is primarily a mixture of Cr2O3 with small amounts of Co and Mo oxides [6]. In particular, Hodgson et al. showed that 90% of the passive film is formed by Cr oxides. In the passive region, the film is formed mostly by Cr(III) oxide and Cr(III) hydroxide [7]. Passive metals are prone to wear accelerated corrosion when they undergo sliding motion in a corrosive environment, such as in body fluids. The synergy between wear and corrosion, known as tribocorrosion, occurs when the passive film is ruptured by mechanical wear, leaving the unprotected metal exposed to the corrosive media [8–10]. The tribocorrosion behavior of CoCrMo alloys has attracted a great deal of attention due to their high popularity in metal-on-metal hip joints [11–15]. The synergy between wear and corrosion of CoCrMo alloys depends on the testing conditions and electrolyte, as highlighted by Mathew et al. using two test rig configurations and two reference electrolytes [11]. This means that surface degradation of CoCrMo during tribocorrosion is determined by the electrochemical conditions. In particular, a transition from low to high wear coefficients at an applied potential value of around 0.2 VSHE has been identified in the data published by Mischler and Igual Muñoz in a recent overview on the different testing methodologies [10]. The detrimental synergy between wear and corrosion is twofolded. Besides a higher wear rate, the major concern for prospective patients is that tribocorrosion also leads to the release of metal ions. The release of metal ions in CoCrMo alloys has been determined using Inductive Coupled Plasma-Mass Spectrometry (ICPMS) in several in vitro tribocorrosion experiments using model tests [15] and hip joint simulator [16]. Espallargas et al. noticed that despite the disparity of CoCrMo alloys investigated, tribocorrosion testing parameters and electrolyte used, all works reported

similar ion concentration ratios, with Co having the highest concentration (most of the values ranging from 60 to 85%), Cr having an intermediate concentration (10–35%) and Mo with the lowest concentration (typically 1–13%). Ion release and presence of wear particles may trigger an adverse response by the immune system, leading to inflammation and damage of soft tissues surrounding the implant. In particular, metal particles have been reported to trigger apoptosis or programmed cellular death in cells at the interface, thus impairing the biointegration process and eventually leading to total failure of the implant. The presence of a layer of dead cells has been reported at the surface of bovine cartilage articulation against cobalt chromium induced [17]. The release of metal ions has also been measured in patients having metal on metal (MoM) hip joint implants with Co concentrations in synovial fluid from patients with MoM hip prosthesis being 11.50–64 550 mg/l [18]. The concentration of Co and Cr was significantly higher in the metal on polyethylene than in the MoM group [19]. However, no data on metal ion concentration in hemiarthroplasty or partial surface replacement is reported until now. Most of the previous investigations dealing with the tribocorrosion performance of CoCrMo alloys used a three electrode setup, where the CoCrMo alloy is set as a working electrode. A ceramic ball, typically Al2O3, is rubbing the metal while immersed in simulated body fluid, such as Ringer or phosphate buffered saline solution. The contact pressures typically used in a ball-on-flat configuration by many authors exceed 1 GPa. Physiological loads experienced by cartilage in the knee have been measured to be about 1 MPa during standing with peaks of 5–10 MPa during walking [20]. The available literature reports that contact pressures in the range of 4.5 MPa at the cartilage-implant interface may induce apoptosis in addition to matrix degradation in articular cartilage [21]. There is little research done on the tribological behavior of articular cartilage and adjacent metal implants with focus on biotribocorrosion. The simultaneous influence of corrosion and wear on the metal implant can lead to metal release. As consequence, the degradation of the remaining cartilage by mechanical wear can be accompanied by cartilage degradation due to biological reactions at the metal - cartilage interface between corrosion products and organic compounds. Except for a few examples, biotribocorrosion experiments are often performed under highly idealized conditions that scarcely reflect the complexity of the in-vivo situation. Commonly, joint prostheses are tested under mechanical conditions mimicking the natural joint situations but mostly using simplified material couplings and chemical environments. Most of the studies dealing with the contact between implant materials and articular cartilage have set the focus on evaluating the frictional properties of the tribosystem and cartilage damage by several methodologies such as by measuring total protein and SZP evaluation [4,22]. However, the role of tribocorrosion on metallic implants sliding against articular cartilage is an aspect that to our best knowledge has not been addressed so far. Biotribocorrosion is a common failure mechanism in metallic implants, but the electrochemical conditions are rarely controlled in biotribological tests and real tissues are seldom used in laboratory experiments. There is a need to develop novel test methods that will allow to characterizing the tissue-biomaterial interface under well-defined mechanical loads, relative displacements, and chemical and electrochemical conditions representing the in-vivo complexity. This work aims to address this gap by investigating the biotribological behavior of CoCrMo alloys sliding against articular cartilage using a three electrode electrochemical cell. The emphasis is set on the role of sliding in ion release and the subsequent damage that critical ion concentrations may cause to chondrocytes.

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

3

2. Materials and methods

2.2. Biotribocorrosion experiments

2.1. Cartilage specimens and cobalt chromium molybdenum (CoCrMo) cylinder

The tribo-electrochemical set-up used in this study is shown in Fig. 1. It consists of a three electrode floating cell monitored using a potentiostat and coupled to a microtribometer. The set-up allows characterizing the synergistic behavior between corrosion and mechanical wear of passive metals. The CoCrMo cylinder serves as the working electrode, and its potential is monitored using a potentiostat VERSASTAT 3F from AMETEK GmbH (Meerbusch, Germany). The CoCrMo cylinder was electrically insulated from the metallic load cell by placing a polymeric PEEK insulating plate in-between to avoid galvanic coupling. A Ag/AgCl in 3 M saturated NaCl solution was used as reference electrode. A coiled platinum wire with 0.4 mm diameter and 500 mm length was used as the counter electrode. Throughout this paper, the potentials are reported against Ag/AgCl. The three-electrode electrochemical cell was coupled to a reciprocating sliding microtribometer FALEXMUST precision tester from Falex Tribology NV (Rotselaar, Belgium). The full set-up of the microtribometer is described elsewhere [24]. As an electrolyte, phosphate buffered saline (PBS, Sigma-Aldrich Chemie GmbH, Steinheim, Germany) was used. Sample preparation prior to testing was done in a bio-safety cabinet. All tribo-electrochemical experiments were conducted at room temperature (RT) under ambient conditions. The cylinderon-plane configuration with a stroke of 2 mm and an average sliding speed of 8 mm/s used for tribocorrosion experiments resulted in a migrating line contact that represents the situation in diarthrodial joints [25]. The selected speed is within the range of speeds experienced during a gait cycle [26]. A load of 1 N was applied at the beginning of sliding and it was constantly monitored throughout the sliding phase of the experiment. The resulting contact pressure of this testing configuration was determined using Fujifilm pressure measurement (Fujifilm PrescaleÒ, Fujifilm Corporation, Tokyo, Japan). The Fujifilm reacts to the applied pressure showing red discoloration of zones where the threshold pressure is reached or exceeded. The calibration was performed by applying a nominal force on a defined flat surface area (pin on disc) in order to obtain reference imprints for defined pressure values (0.5 MPa, 1 MPa, 2 MPa, 3 MPa, 5 MPa respectively). The contact pressure of the CoCrMo cylinder on cartilage configuration was estimated by placing the Fujifilm between the cylinder and cartilage (line contact). The selected load was applied and after 5 cycles of reciprocating sliding (2 mm stroke), the imprints were collected and photographed (Fig. 2). Due to the convex shape of the cartilage surface, the contact situation deviated from a perfect line contact, which resulted in a non-homogeneous contact pressure along the sliding area and a pressure concentration at the top of the specimen. These imprints were compared to the reference measurements done using a pin on disk configuration. By visual comparison, the contact pressure between the CoCrMo cylinder and the osteochondral OC plug when applying a load of 1 N was determined to be 2–2.5 MPa. This pressure range occurs in femoro-tibial contact during normal gait [20]. During the experiment, the CoCrMo cylinder slid against the fully immersed cartilage specimen. The open circuit potential

Osteochondral (OC) specimens were harvested (day 0) from a bovine medial femoral condyle (18–24 months old) supplied from a local butcher using Single-Use OATS punch (Arthrex Inc., Naples, USA). Single-Use OATS punch was selected for harvesting based on previous studies. It is known that chondrocyte survival in osteochondral plugs is influenced by the harvesting technique. The study [23] compared three different harvesting devices (Arthrex’s OATS; diamond cutter by Karl Storz, and hollow reamer with cutting crown) and samples harvested with OATS instruments showed a higher chondrocyte viability in the peripheral zone. Three OC plugs per animal (d = 8 mm, l = 10 mm) with visually flat, intact surface were chosen for experiments and stored in medium at 4 °C prior to testing in order to keep the tissue alive. The growth medium selected for storage (GIBCOÒ DMEM/F-12 GlutaMAXTM-I, Life Technologies, Carlsbad, CA, USA) contains antibiotics (penicillin 200 U/mL; streptomycin 0.2 mg/mL) and Amphotericin B 2.5 mg/mL (Sigma-Aldrich Chemie GmbH, Steinheim, Germany). Seven animals were tested in total and they were marked as Animal1 - Animal7 (A1-A7). The tribocorrosion experiments were performed within the following 2 days (in the following named day 1, day 2). CoCrMo alloy was selected for experiments in these studies due to its clinical role in hemiarthroplasty and other resurfacing procedures involving cartilage-metal interface. The material used (Acnis International, Chassieu, France) has standard specification for surgical implants ASTM F1537, ISO 5832-12 and for dentistry ISO 22674. The chemical composition of the alloy was determined using energy dispersive x-ray spectroscopy (EDX) and compared with the nominal values provided by the producer (Table 1). The lateral surface of the rod (6 mm diameter) was wet ground with SiC-paper starting with an initial grain size 500, followed by 1000, 2500 up to 4000. Each step was proceeded until achieving a homogeneous surface roughness. Afterwards the rod was polished with 3 mm and 1 mm paste resulting in a final Ra value of 15 ± 2 nm and was cut in small cylindrical samples (length = 10 mm) which were used for experiments. A small hole was drilled on the cylinder base approximately 1 mm away from the edge. The hole enabled electrical connection to a wire and usage of the cylinder as working electrode. After rinsing with water the samples were cleaned in an ultrasound bath with acetone and ethanol and assembled into the measurement cell. In order to initiate the experiments with a well-defined and reproducible protective passive layer on the metal surface, some samples (A6 and A7) were treated prior to testing with 20% HNO3 for 30 min in ultrasound bath at 50 °C. Afterwards they were rinsed with water and cleaned in an ultrasound bath with acetone. Before the test sample was cleaned with ethanol, dried by using N2 gas and assembled into the measurement cell. A conductive wire was attached to a cylinder and the opposite side was connected to the potentiostat.

Table 1 Elemental composition of the low-carbon CoCrMo alloy used in this study.

*

Element

Si

Mo

Cr

Mn

Co

C

Nominal, average wt.% Measured, average wt.% ± SD

0.67 0.89 ± 0.17

5.49 4.90 ± 0.42

27.21 27.51 ± 0.30

0.75 0.84 ± 0.21

65.55 64.41 ± 0.61

*

0.04

1.45 ± 0.05 wt.% concentration of C due to contamination.

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

4

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

Fig. 1. Procedure and experimental set-up for the tribo-electrochemical experiments: Image of bovine medial femoral condyle before (a) and after harvesting (b) in aseptic conditions. The osteochondral plugs were stored in growth medium and kept at 4 °C (c). Prior to testing, the OC plug was placed in the holder (d) and assembled in the tribocorrosion cell with a CoCrMo-on-cartilage configuration. Schematic diagram showing the positions of the electrodes during the tribocorrosion experiment (f).

Fig. 2. Fujifilm pressure determination. Reference imprints with a pin on disk configuration (polymer on metal) with defined contact pressures (a) and CoCrMo on cartilage configuration imprints (b). Red color density in the center of the imprint (for ±1 mm movement) for CoCrMo on cartilage system when 1 N is applied indicated the specific load (contact pressure) to be 2–2.5 MPa. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

(OCP) was monitored throughout the whole experiment. The experiments included three phases:
Phase 1: After assembling the samples in the three-electrode cell, the upper holder containing the CoCrMo cylinder was positioned approximately 1 mm distance (starting position) from the cartilage surface. In this position, the cylinder was immersed in PBS and stabilization of OCP was monitored. The phase duration was 2 h.
Phase 2: The load of 1 N was applied, and the cylinder started reciprocating sliding against the cartilage surface with 8 mm/s average velocity for 1 h. During this phase, the OCP and the coefficient of friction (COF) were continuously recorded.
Phase 3: When sliding was finished, the cylinder was placed back in the starting position for stabilization and allowing repassivation of the CoCrMo surface. The duration of the 3rd phase was 30 min. 2.3. Chemical analyses of the electrolyte and cartilage surface Immediately after the conclusion of the experiment, the PBS electrolyte was collected for investigating the presence of metal content. The analysis of the elemental composition was carried out utilizing an iCAPTM 7400 duo Inductively Coupled Plasma Optical Emission Spectrometry (ICP) spectrometer supplied by Thermo Fisher Scientific (Bremen, Germany). The PBS samples were measured directly, without dilution. The calibrating solutions were diluted from commercially available single-element ICP standards. For every element, 3 independent axial measurements were carried out on different emission wavelengths to minimize the

possibility of reporting false concentrations due to interferences. Accordingly, the reported values are the mean values of those independent measurements, where the standard deviation does not exceed 10% between the measured values. The following emission wavelengths were utilized:
Cobalt: 228.616 nm; 238.892 nm; 237.862 nm
Chromium: 283.563 nm; 267.716 nm; 357.869 nm
Molybdenum: 202.030 nm; 204.598 nm; 203.844 nm The limit of detection (LOD) and the limit of quantitation (LOQ) were determined based on the measured intensity (Ib) and its standard deviation (SDb) of the blank sample. In doing so, the measurement of the blank sample was repeated 20 times to achieve a statistically significant sample. The LOD was set Ib + 3 SDb and the LOQ to Ib + 10 SDb respectively. Table 2 shows the calculated LOD and LOQ values for the measurements. The presence of Co, Cr and Mo in the cartilage matrix after the experiment was investigated using high resolution mass spectrometry. This analytical technique is particularly suitable to detect the chemical composition of molecules on the cartilage surface after the tribo-corrosion experiment. Untested osteochondral plugs

Table 2 The limit of detection and the limit of quantitation for Co, Cr and Mo calculated for ICP measurements. Limits

Co

Cr

Mo

LOD (ppb) LOQ (ppb)

1 4

1 4

3 6

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

were used as control. Right after the experiment, the OC plug was removed from the test cell holder. A drop of methanol served as a solvent and was placed on the cartilage surface (tested and control) and left for 2 s. Subsequently, the drop with eventually solved organic molecules was collected and directly infused at a flow rate of 5 ml/min by utilizing a LTQ Orbitrap XL hybrid tandem mass spectrometer from ThermoFisher Scientific (Bremen, Germany). The instrument is fitted with an electrospray ionization (ESI) ion source and was operated in positive ion mode. The ESI ion source conditions were as follows: spray capillary temperature of 275 °C; spray voltage of 4.0 kV; capillary voltage of 48 V positive mode. Nitrogen was used as sheath gas, and helium was used both as buffer and collision gas in the linear ion trap section. Low-energy collision-induced dissociation (CID) was performed with a normalized collision energy optimized from 25 to 40%. CID generated product ions were transferred and detected by the high resolution orbitrap-section of the instrument at a resolution of 60,000 (full width at half maximum, FWHM). Data processing and interpretation were done using the software tools Xcalibur version 2.0.7 and Mass Frontier version 6.0 from ThermoFisher Scientific (Bremen, Germany). All mass measurements were acquired with a mass accuracy of 5 ppm or better. 2.4. CoCrMo cylinder surface characterization Following the experiment, the CoCrMo cylinder was gently rinsed in distilled water and examined using a scanning electron microscope (SEM) Zeiss Supra 40VP operated at an accelerated voltage of 10 kV and 20 kV. The elemental composition was measured using energy dispersive x-ray spectroscopy (EDX). The changes in chemical composition of the tribofilm after the test were analyzed using X-ray photoelectron spectroscopy (XPS). The analyses were conducted using a Theta Probe (Thermo Fisher Scientific, East Grinstead, UK) equipped with a monochromatic Al Ka X-ray source (hv = 1486.6 eV) and a hemispherical analyzer. The measurements were performed at a base pressure of 1010 mbar. The analyzed spot size diameter was 200 mm. For each sample, the high-resolution spectra were obtained at a pass energy of 50 eV. Afterwards, depth profiles of the respective surfaces were acquired by sequentially measuring and sputtering the surface. After each sputtering step, the composition of the surface was determined by using high resolution scans. The sputtered area was 2 mm  2 mm, with a sputter current of 1 mA at 3 kV. These parameters resulted in an estimated sputter rate of approximately 0.05 nm/s. The data analysis was performed using a Thermo Fisher Scientific Avantage Data System software. The results of both methods (SEM/EDX and XPS) were compared with an untested specimen (control). All specimens used for these measurements (both control and tested) were treated with HNO3 prior to testing as described in Section 2.1. 2.5. Co-ion electrolyte and cell cultures An electrolyte solution containing Co ions was prepared in order to determine the effect of Co ions on osteoarthritic human chondrocytes. A CoCrMo rod cylindrical sample was ultrasonically cleaned as earlier described and immersed in PBS in a three electrode electrochemical cell. The CoCrMo cylinder was used as the working electrode. As reference electrode we used Ag/AgCl in 3 M saturated NaCl solution. A platinum wire was used as counter electrode. The three electrodes were connected to a potentiostat for applying a potential of 0.7 V. The composition of the resulting stock solution after 5 h of potentiostatic exposure was determined using ICP (Co = 98 ppm, Cr = 40 ppm, Mo = 7 ppm), and it was very closely proportional to the alloy composition (67.6% cobalt, 27.6% chromium and 4.8% molybdenum).

5

Human articular cartilage was received from the local hospital from osteoarthritis patients undergoing total knee arthroplasty. In all cases, informed consent was obtained, and the study was approved by the Regional Ethical Committee (GS1-EK-4/4812017). For chondrocyte isolation, articular cartilage was minced into 2 mm3 pieces prior to enzymatic digestion with Liberase TM (0.2 WU/ml, Roche Diagnostics GmbH, Mannheim, Germany) in medium (GIBCOÒ DMEM/F-12 GlutaMAXTM-I, Life Technologies, Carlsbad, CA, USA) with antibiotics (penicillin 200 U/ml; streptomycin 0.2 mg/ml) and antifungal amphotericin B 2.5 mg/ml (Sigma-Aldrich Chemie GmbH, Steinheim, Germany)) under permanent agitation for 18–22 h at 37 °C. The resulting chondrocyte suspension was passed through a Cell Strainer with 40 mm pores (BD, Franklin Lakes, NJ, USA) to remove undigested debris, washed with phosphate-buffered saline (PBS), centrifuged (10 min, 500g, room temperature) and resuspended in growth medium (i.e. medium supplemented with antibiotics and antifungal (see above), 10% FCS (GIBCOÒ, Life Technologies, Carlsbad, CA, USA). Viability was determined via trypan blue (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) staining and cells were counted using a hemocytometer. The isolated cells were seeded in growth medium in 75 cm2 culture flasks (Nunc, Rochester, NY) at a density of 1x104 cells/cm2 and cultivated at 37 °C in a humidified environment with 5% CO2. Medium was changed every 2–3 days till 80% confluency. After expansion, cells were harvested by the use of accutase (1.5 ml/flask; Sigma-Aldrich Chemie GmbH, Steinheim, Germany) and counted and seeded onto 24 well plates at a seeding density of 10,000 cells/cm2. Followed cultivation was done in growth medium at 37 °C in a humidified environment with 5% CO2 for 3 days. Post 3 days medium was replaced by solutions (1 ml per well) with 16.3 ppm of Co ions. As a reference, these solutions were also added to wells without cells. The influence of metal ions on cell proliferation was measured in duplicates on day 0, day 1 and day 3 by the XTT based Cell Proliferation Kit II (Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer’s instructions. Briefly, the XTT labeling reagent was mixed with electron coupling reagent in the ratio 1:50. The XTT labeling mixture (500 ml per well) was then added to the cell culture medium (1 ml per well) and incubated for 4 h at 37 °C. After incubation, absorbance was measured at 492 nm with a background reference wavelength at 690 nm using a multi-mode microplate reader (SynergyTM 2, Winooski, Vermont, USA). The stock solution was diluted with PBS and the final solution for one 24 well was containing:
166.67 ml of solution with Co ions
733.33 ml medium without FCS and
100 ml FCS The mentioned components were mixed and added to the culture medium. The final concentration applied on the cells was 16.3 mg/kg of Co. PBS was serving as a control. Cell culture imaging on day 1 and 3 was performed using the EVOS FLOID Cell Imaging Station (Life Technologies, Carlsbad, CA, USA).

2.6. Statistical analysis A total number of 7 animals were tested using identical conditions. The data generated from the tribocorrosion experiments is presented as mean ± standard deviation (SD) from 3 repetitions per animal in case of CoF and metal content in PBS after the experiment. The only exception from this case is A1 where only one ICP value was counted as valid. Measurement of potential was included in experiment at least in 2 of the 3 repetitions.

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

6

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

Chondrocytes isolated from 3 osteoarthritic patients, were used in the cell culture experiments. The metabolic activity results are presented as median with error bar showing minimum and maximum value. 3. Results 3.1. Biotribocorrosion experiments The biotribocorrosion experiments were performed using a line contact with constant normal load of 1 N. The results of the tribocorrosion experiments show that the open circuit potential stabilizes during the initial 2 h (Fig. 3). The stabilization value of the potential lied in values ranging between 0.31 V and 0.24 V (Fig. 4a). For the samples treated with HNO3 prior to the experiments, the OCP values were higher (0.15 to 0.10 V). The OCP value of CoCrMo alloys in PBS reported in literature lies around 0.1 V after cathodic cleaning [27]. Repetitions of the test using the OC plugs from the same animal led to closer OCP values (Fig. 4b). Once the potential was stable, the CoCrMo cylinder started sliding against the osteochondral plug for 1 h. An approximately 4 mV drop of the open circuit potential was systematically detected in all experiments at the onset of sliding. The exception was in case of animal 4 where the drop was lower (1 mV). The drop in potential is attributed to changes in passive layer on the metal surface which may lead to metal release. After the drop, the potential increases steadily until the end of the test and recovers the pre-sliding value. In some experiments, the OCP value increases to higher anodic values than the ones measured during stabilization, suggesting a different passive layer characteristics after sliding. During rubbing, the coefficient of friction typically shows an initial peak at the onset of sliding and shortly after becomes lower and remains constant at a value of around 0.03 (Fig. 3). The stan-

dard deviation of the coefficient of friction was approximately 0.005 within one animal (Fig. 4c). The standard deviation of the coefficient of friction within one animal arises due to two main factors. The first one is due to small experimental inaccuracies that occur when repeating two nominally identical experiments. The second is caused by the necessarily different position of harvesting within the animal joint. The average value of the coefficient of friction for each one of the seven animals varied slightly. These variations are attributed to small physiological differences due to age and condition of the animal. Another source of variability is certainly the convex shape (curvature) of the surface which varies from one sample to another depending on the position where the cartilage was harvested. The average coefficient of friction for all 7 animals considering all test repetitions performed was 0.030 ± 0.011. No correlation could be found between the magnitude of the potential drop and the coefficient of friction. 3.2. Co concentration in electrolyte after tribocorrosion experiment The metal content in PBS after the biotribocorrosion experiments is shown in Table 3. In the table, the metal concentration in the solutions is shown in parts per billion (ppb) for the element Co. The other two elements, Cr and Mo, were sometimes detected, but it was not possible to quantify them. Significant quantities of Co were found in the electrolyte after the one hour experiment (up to 22 ppb) confirming the occurrence of tribocorrosion in the CoCrMo alloy during the rubbing process. As control, a CoCrMo cylinder was fully immersed in 25 ml PBS for 4 h (the same amount of PBS used in the tribocorrosion experiments). The amount of released metal according to ICP measurements was below the detection limit. Lower amounts of Co were found for slightly higher CoF (Fig. 4d). Nevertheless a significant trend between the coefficient of friction measured for each animal and the amount of Co release was not possible since the total amount of metal ions and particles detached from the CoCrMo alloy would require a quantitative analysis for both, PBS and OC plug, whereas in our case the concentration found refers uniquely to the PBS solution. Further, a correlation between the magnitude of the potential drop at the onset of sliding and the amount of Co release was not found. 3.3. Surface analyses of the CoCrMo cylinders after the experiments

Fig. 3. Evolution of the coefficient of friction (CoF) and the potential for the OCP experiments (A7).

The surface morphology of the CoCrMo cylinders before and after the tribocorrosion experiments are displayed in Fig. 5. The SEM investigations revealed a smooth surface morphology without any visible wear scars either before or after the experiments. The main feature differentiating the cylinders before and after the tests were scattered darker spots with an irregular shape present on the surface only after the experiment. The EDX measurements taken at the dark spots (indicated by orange arrows in Fig. 5) revealed significantly higher amounts of carbon (up to 73 at.%) and a higher content of oxygen (3–4 at.%) which accounts as an indication for corrosion. In contrast, the oxygen content outside the dark spots or in untested samples was lower than 0.2 at.% while the carbon content never exceeded 4 at.%. XPS analyses were performed on tested CoCrMo cylinders in order to investigate the chemical surface composition of the tribofilm and its thickness. An untested sample was used as a reference. The analyses revealed that the tribofilm formed after the biotribocorrosion experiments extended well below the complete sputtered depth. The tribofilm was mainly composed of chromium oxide, small amounts of molybdenum oxide, a considerable amount of carbon and nitrogen (Table 4). The layer formed by metal oxides and hydroxides had a higher thickness compared to the untested cylinder.

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

7

Fig. 4. Representative evolution of OCP with time for Animal 1 (A1) to Animal 7 (A7); drop of potential visible at the onset of sliding (a). The average values (mean ± SD) for stabilized OCP and OCP after load application (onset of sliding) including potential drop for A1 to A7 (b). Average coefficient of friction for A1-A7 (c). Average coefficient of friction as a function of Co amount released into the PBS (mean ± SD) for A1-A7 (d).

Table 3 Metal release after biotribocorrosion tests for animals A1-A7. Specimen

Control

A1

A2

A3

A4

A5

A6

A7

Co content (ppb)


2.0

3 ± 2.6

8 ± 2.1

3 ± 0.6

16 ± 6.7

7 ± 2.8

22 ± 4.5

Fig. 5. SEM micrograph of CoCrMo surface before (a) and after the tribocorrosion experiment (b, c). Cylinder after the experiment lightly rinsed with water and air dried. Accelerating voltage 20.00 kV (a, b) and 10 kV (c). Original magnification: 500 and 2000.

Table 4 Chemical composition (atomic %) according to XPS before (left columns) and after (right columns) the tribocorrosion experiments. Etch time (s)

C (carbon/carbide)

Co (metal)

0 10 20 30 40 65 90 115 140

58.4 6.1 1.8 1.6 1.6 1.2 1.0 0.8 0.7

2.5 59.6 70.3 71.8 70.0 68.5 69.4 67.3 68.8

Binding Energy

282.8; 284.4

71.7 8.3 7.5 7.3 7.0 6.8 6.0 5.9 6.1

778.4

Cr (metal)

1.1 33.2 33.7 33.5 33.5 33.1 34.4 34.6 35.1

1.3 18.1 16.9 15.7 17.1 18.5 18.1 18.7 18.7 574.3

Cr(III) (oxide)

0.5 13.1 14.1 13.1 13.0 12.8 13.0 13.2 12.9

6.3 0.3 0.2 0.1 0.1 0.0 0.1 0.0 0.0 576.3

Mo (metal)

1.5 18.5 18.0 19.0 19.3 20.2 19.0 19.4 19.1

0.8 7.4 8.6 8.1 7.7 8.3 8.7 10.0 8.5 227.9

Mo(II) (oxide)

0.4 4.7 5.9 5.4 5.9 5.1 5.7 5.6 5.5

0.8 1.8 0.0 0.6 1.5 1.6 1.0 1.4 1.3 228.3

O (metal oxides) 0.4 3.3 2.0 2.7 2.5 3.0 3.1 2.8 2.5

14.2 3.8 1.6 1.4 1.4 1.3 1.2 1.3 1.4 530.7

0.0 11.6 11.6 11.6 11.5 11.5 11.7 11.7 11.8

O (Metal hydroxides, carbonates)

N (organic matrix)

11.1 2.1 0.5 0.5 0.5 0.5 0.4 0.4 0.3

– – – – – – – – –

532.0

13.8 4.4 4.3 4.6 4.5 4.7 4.5 4.5 4.4

400.7

O (organic residue) 3.7 0.0 0.2 0.4 0.0 0.0 0.0 0.0 0.0

2.4 0.6 0.2 0.2 0.1 0.2 0.1 0.2 0.1

4.6 1.9 1.9 1.7 1.9 1.8 1.6 1.7 1.6

533.0

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

8

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

3.4. Identification of compounds present on the cartilage surface In order to obtain qualitative information about the molecules present on the cartilage surface before and after the tribocorrosion experiment, a mass spectra was acquired. Figs. 6 and 7 shows the changes in composition on the cartilage surface due to sliding against the metallic implant material. Fragment ions m/z 103.9555, 99.9606 and 101.9398 contained the element Cr in different oxidation states: +3, +6, +4 respectively (Fig. 6b). The element Co is known to form chelate-complex with amino acids. Fragment ions m/z 148.9890, 177.0204 and 191.0369 visible in the mass spectra (Fig. 7b) corresponds to organometallic complexes of the element Co with amino acids. More precisely it shows derived products of Co with Alanine, Valine, and Leucine respectively. The whole complex is visible in some cases in very low peak intensity and it’s not shown in the figures. 3.5. Co ion effect on osteoarthritic chondrocytes The electrolytic solution containing Co ions was obtained by maintaining a potentiostatic value of 0.7 V during 5 h on a CoCrMo rod half-immersed in PBS. The potential applied for metal dissolution is located in the transpassive region (Fig. 8a), which is characterized by a current increase due to the dissolution of the chromium oxide passivation layer accompanied by water oxidation. According to literature, the metal dissolution at this potential is proportional to the composition of the alloy, rather than dominated by chromium dissolution [28]. The potential presence of cytotoxic Cr6+ in the electrolyte was addressed using the following analytical method:

Cr2 O7 2 + 4 H2 O2 + 2 Hþ ! 2 [CrO(O2 )2 ] + 5 H2 O

ð1Þ

Chromium commonly adopts the +3 or +6 oxidation state in water solutions. The addition of sulfuric acid if Cr (VI) is present, leads to its conversion to bichromate. The following step includes addition of hydrogen peroxide, dropwise, under heating in a water bath. When bichromate is present, it reacts with hydrogen peroxide, and the ligation of peroxo moieties takes place and forms the complex according to reaction (1). The formed Cr (VI) complex

can be extracted by addition of diethyl ether, and a deep blueviolet color appears. The Co-ion electrolyte generated potentiostatically showed negative reaction on the presence of Cr (VI) using the method mentioned above (Fig. 8b). As a reference, the same trial was repeated with 5 M solution of K2Cr2O7 and the presence of Cr (VI) was confirmed with a dark blue color visible on top of the solution (Fig. 8c). The influence of Co ions on the metabolic activity of osteoarthritic human chondrocytes is shown in Fig. 9. When only PBS (control group) was present, the cells were proliferating in the culture resulting in an increased metabolic activity value (day 1 and day 3) when compared to the base line. Micrographs are also confirming a higher number of cells after 3 days in the control group. The presence of Co leads to a decrease in metabolic activity, which is significantly lower when 16.3 ppm of Co is present in the solution in comparison to the control group. Changes in cellular shape and their detachment from the surface for a high Co concentration is shown in Fig. 10. These analyses provide a first insight of Co ion effects on cell metabolic activity and morphology. Further studies to investigate the detailed mechanisms of Co effect on chondrocytes are currently underway.

4. Discussion The present contribution has addressed the potential role of biotribocorrosion during sliding contact between articular cartilage and a CoCrMo alloy as it may occur in hemiarthroplasty or other surgical procedures resulting in a metal on cartilage interface. To this end, a three electrode electrochemical cell coupled to a microtribometer was used. The results obtained reveal the occurrence of metal release during the rubbing process despite the extremely low coefficient of friction (0.030 ± 0.011) achieved during sliding. The coefficient of friction of cartilage on metal under comparable low contact pressures has been reported in literature. Chan et al. measured for OC plugs sliding against CoCrMo pins with a normal load of 1.2 N (average contact pressure of 0.1 MPa) in PBS an initial coefficient of friction of 0.07 after 1 min sliding that steadily rose during the 60-minute experiments up to values as high as 0.57 [4]. Bonnevie et al. investigated the

Fig. 6. Mass spectra of Cr-containing compounds identified on the cartilage surface before (a) and after the tribocorrosion experiment (b). A comparison of both spectra reveals differences in the molecules present on the cartilage surface. The element Cr was detected in different oxidation states (b): Cr (III) at 99.9606 m/z, Cr (VI) at 101.9398 m/z and Cr (IV) at 103.9555 m/z.

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

9

Fig. 7. Mass spectra of Co-containing compounds identified on the cartilage surface before (a) and after tribocorrosion test (b). A comparison of both spectra reveals differences in the molecules present on the cartilage surface. The element Co was detected in form of organometallic complexes with amino acids.

Fig. 8. Potentiodynamic polarization curve of CoCrMo alloy in PBS (a); Qualitative inorganic analysis of (b) Co-ion electrolyte and (c) control K2Cr2O7.

coefficient of friction as function of several parameters of OC plugs against 440C stainless steel in PBS using a microtribometer [5]. The reported friction values rose from 0.025 up to 0.14 with decreasing reciprocating sliding speed. For the largest reported sliding speed of 5 mm/s, comparable to the 8 mm/s used in the present work, the coefficient of friction had a value of 0.025, in agreement with our measurements. The open circuit potential was constantly monitored throughout all three phases of the experiment. The stabilization value of the potential after 2 h lay in values ranging between 0.31 V and 0.24 V. For the samples treated with HNO3 prior to the experiments, the OCP values were shifted towards the anodic region (0.15 to 0.10 V). In a study comparing the electrochemical response of CoCrMo in bovine calf serum (BCS) and extracted

synovial fluids from osteoarthritic patients [29], the OCP values found in BCS were significantly lower (in average 0.243 V vs Ag/AgCl) than the ones measured in synovial fluids (in average 0.001 V Ag/AgCl). The study also revealed an increase in OCP with increasing inflammation degree from 0.10 V Ag/AgCl (no inflammation) up to 0.15 V Ag/AgCl in a continuous trend. Comparing the mentioned results with ours, electrochemical response of CoCrMo under OCP was similar in the presence of BCS and bovine OC plugs in PBS (without pre-treatment with HNO3). Therefore, one of the factors that contribute to variations in stabilized OCP value between the animals is the health condition of the tested animals. For all experiments performed under identical contact conditions using OC plugs from different animals, a consistent drop of 4 mV was observed at the onset of sliding. Such a drop in potential

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

10

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

Fig. 9. Metabolic activity of osteoarthritic human chondrocytes presented as median from 3 patients (with min and max values) exposed to PBS with 16.3 ppm of Co and control group (PBS) after 1 day and after 3 days. Day 0 presents the base line value. Significant decrease in metabolic activity of the human osteoarthritic chondrocytes was noticed when 16.3 ppm of Co ions were present compared to control group.

is usually observed in passive metals [9] when the layer of the alloy is locally removed by wear, and as a consequence, a potential difference is established between depassivated areas (anodic) and still passive areas (cathodic). Depassivation of the metallic surface results in a subsequent release of metal ions. However, in many tribocorrosion experiments using model tests, the drop of potential during the experiment towards the cathodic region is much more pronounced. The reason is that in those experiments, the counterbody of the CoCrMo alloy is typically a ceramic ball with a much higher hardness compared to the cartilage, resulting in point contact and contact pressures exceeding 1 GPa for the typically reported normal force values ranging from 1 to 80 N [10,11,15,30,31]. The surface analyses of the CoCrMo cylinders after the experiments confirmed changes in the chemical composition of the surface layer. The significantly higher carbon content found after the test suggests the incorporation of cartilage derived organic matrix components in the formed tribolayer. The C rich dark spots observed in the scanning electron microscopy images have a higher oxygen content which accounts for corrosion of the alloy. The presence of C accompanied by N on the formed tribofilm could also be assessed by XPS. Besides depassivation, the presence of

cartilage-derived organic components on the CoCrMo surface can lead to a higher corrosion by metal dissolution. As reported in [32], the presence of proteins and adsorbed protein layers on a CoCrMo surface can accelerate corrosion in static conditions. Other studies also found that bovine serum albumin (BSA) increases corrosion in low carbon CoCrMo alloys, attributed to the formation of ion metal-protein complex formation with Cr that enhance passive dissolution [33]. This is in accordance with a previous study that reported that while a protein film can reduce the rate of corrosion by acting as a negatively charged barrier preventing corrosive anions such as Cl approaching the metal surface interface, the same proteins can act as a ligand-induced dissolution mechanism for increasing the ion content for Cr in relation to Co [34]. From another side, surface repassivation during sliding incorporates organic matrix compounds in the newly formed tribolayer, which has the potential to improve corrosion resistance [30,35]. The metal ion release resulted in concentrations of Co in the PBS ranging from 2 to 22 ppb, as measured using ICP after the experiments. The measured concentrations are much lower than the one found under much higher contact pressures (142.7 ppb of Co for 30 min rubbing time) for CoCrMo sliding against alumina ball in PBS under OCP conditions [15]. Nevertheless the underlying mechanism for ion release after removal of the passive layer might be the same in both cases. As summarised by Espallargas et al. [15], in vitro tribocorrosion experiments of CoCrMo alloys sliding against ceramic counter bodies lead to metal ion release of Co, Cr, and Mo in similar ratios fairly independent of the testing conditions. Co is present with the highest ratio of metal ions, with concentrations ranging from 60 to 85%, followed by Cr (10–35%) and finally Mo (1–13%). Higher concentrations of Co compared to the other two elements could be explained by the fact that Co forms dissolved ions, while Cr rather precipitates as solid oxide, and from the alloy stoichiometry [15]. Based on these observations it is highly likely that the observed presence of Co is accompanied by Cr and Mo in smaller concentrations. If we assume a ratio of Co, Cr and Mo of 70, 20 and 10 respectively, this assumption would lead to a concentration of Cr of 6 ppb and Mo of 3 ppb for the highest concentration of Co (22 ppb) detected in the present work. This best case scenario estimates Cr and Mo concentrations very close or even below the LOQ for Cr (4 ppb) and Mo (6 ppb) shown in Table 2. As a consequence, the outcome of this estimation is consistent with our observations that in some of the PBS solutions after the biotribocorrosion experiments, Cr and Mo could be detected but with concentrations too low to be quantified.

Fig. 10. Representative light microscopy images of control group cells (a, b) after 1 day and 3 days (e, f) and cells cultured with 16.3 ppm of Co in the culture medium and after 1 day (c, d) and 3 days (g, h). Treated chondrocytes show decreased cell growth, altered cell shapes and detachment from the surface.

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

The measured concentration of Co can not be straightforwardly compared to literature, since to our best knowledge, no comparable studies exist that report metal ion concentration under similar contact conditions. For instance, Yan et al. found Co concentrations of 2200 ppb in a self-mated CoCrMo contact (metal on metal) in bovine synovial fluid under OCP using a contact pressure of 112 MPa [36]. Bazzoni et al. reported Co concentrations of 94.7 ppb for CoCrMo sliding against Al2O3 in point contact (Hertzian contact pressure 4.7 GPa) in 0.9 wt.% NaCl [31]. Espallargas et al, found Co concentrations of 143 ppb in tribocorrosion experiments using CoCrMo against Al2O3 in PBS with a point contact pressure of 1.7 GPa [15]. Taking into account these values that were measured under much more severe contact conditions, the measured Co concentrations within the present work seem to be plausible. A rough estimate of the total Co concentration generated per year in a patient doing 5000 sliding cycles a day, if we assume an approximate Co concentration of 10 ppb in 3600 cycles based on the results obtained, would be 5 ppm. This estimation does not take into account different properties of a passive layer formed in the human body, due to the presence of synovial fluid components (eg. proteins) [37]. The shift of the OCP towards cathodic values at the onset of sliding and the Co concentration in the PBS after the experiment evidence the occurrence of biotribocorrosion. The finding of Co and Cr containing compounds on the OC plug after the experiments using high resolution mass spectrometry supports the previous observations. In particular Co was found in organometallic complexes with the amino acids Alanine, Valine, and Leucine. Of particular concern is also the observation of hexavalent chromium due to its cytotoxic and genotoxic effects of on human cells [38]. Articular cartilage is an avascular tissue, and consequently oxygen supply is reduced. Besides the influence of oxygen itself, reactive oxygen species (ROS) play a crucial role in the regulation of several basic chondrocyte activities such as cell activation, proliferation, and matrix remodeling. However, when ROS production exceeds the antioxidant capacities of the cell, an ‘‘oxidative stress” occurs leading to structural and functional cartilage damages like cell death and matrix degradation [39]. Exposing articular cartilage on bovine osteochondral explants to 1.0 MPa for 7 days resulted in elevated ROS production [40]. The occurrence of compounds containing Cr (VI) was unexpected in the studied range of potentials and the presence of ROS under aerobic conditions may contribute to its formation. Nevertheless, there are studies confirming that corrosion of implants can lead to the release of biologically active hexavalent chromium into the body [41]. This chromium is rapidly reduced to trivalent chromium in cells. In any case, further investigation is necessary to show the origin of Cr (VI) and conditions that determine its appearance under physiological conditions. The metabolic activity experiments performed on osteoarthritic human chondrocytes confirm the cytotoxicity of metal ions present in PBS. In particular it was observed that the used Co ions concentration hampered cell proliferation resulting in a decrease in metabolic activity. Hence, biotribocorrosion is a mechanism that could induce chondrocyte apoptosis by the release of metal ions and particles due to the initial mechanical depassivation, and the adhesion of organic compounds promoting passive dissolution. The subsequent reaction of the released metal with organic matrix components results in the formation of compounds with potential cytotoxicity (Fig. 11). An aspect that requires further research is the role of synovial fluid on biotribocorrosion. The introduction of an implant in the human body creates a new interface between synovial fluid, cartilage tissue, and the material. The interaction starts with protein adsorption on the solid implant surface, which is occurring spontaneously on its first contact with biological fluids [42]. In CoCrMo electrochemical studies it was shown that phosphate ions and

11

Fig. 11. Schematic of the proposed initial tribocorrosion mechanism for the system CoCrMo on cartilage.

albumin molecules competitively adsorb on the surface of alloy and in this way they influence the electrochemical behavior. Adsorption of phosphates reduces the corrosion rate of the alloy. Albumin acts in different ways: it limits adsorption of phosphate, thus accelerating corrosion, but can modify the passive film properties and, by acting as cathodic inhibitor, can reduce the corrosion rate [43]. The effects of synovial constituents, namely bovine serum albumin, hyaluronic acid induced a decrease of the friction coefficient, the amount of elastic accommodation of the displacement and the wear rate of the Ti–12.5Mo alloy. It has been shown that organometallic tribofilm has an important role in reducing friction and wear [30], but its long-term effects on cartilage tissue are still unknown. Whether biotribocorrosion is the dominant mechanism in the failure of preserved native cartilage after hemiarthroplasty and other surgical procedures resulting in a metal on cartilage interface can not be determined by the results obtained and is beyond the scope of the present work. Other mechanisms that may arrest apoptosis and produce a stressed chondrocyte phenotype is friction induced inflammation, as observed in Prg4 mutant mice [44]. More recently, Pitenis et al. investigated friction-induced inflammation using in vitro experiments in the context of contact lenses. Their experiments found an increased expression of proinflammatory cytokines and proteases in human corneal epithelial cell during friction testing as well as pro-apoptotic factors [45]. These studies point out that other mechanisms could also trigger chondrocyte apoptosis. Future studies need to determine the dominant mechanism leading to tissue degradation in preserved cartilage after medical procedures creating metal on cartilage interface in order to assess which materials and coatings could provide a lifetime extension in partial implants.

5. Limitations Osteochondral cylinders used in this study were harvested from skeletally mature cattle. Due to the high variability in the biologi-

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

12

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx

cal and mechanical properties of bovine and human tissues, caution should be exercised in extrapolating the present findings between species. The experiments reported here were performed under aerobic conditions. The occurrence of compounds containing Cr VI was not expected, and one of the factors which can contribute to this phenomena is oxygen from the air. Further investigation is necessary to address the origin of Cr VI in the system and possible occurrence in physiological conditions, which possibly would also influence the chosen approach of in-vitro-tests and/or the selected setup. 6. Conclusions The present work showed the occurrence of biotribocorrosion during sliding between a CoCrMo alloy and bovine articular cartilage. A 4 mV drop in open circuit potential was consistently observed in all experiments at the onset of sliding. The potential drop indicates depassivation of the CoCrMo alloy, which led to metal release. Surface analyses of CoCrMo cylinders after the experiments confirmed changes in passive layer and presence of a tribofilm. Co concentrations of up to 22 ppb were measured along with the presence of Co and Cr compounds on the cartilage surface. Co was bound to form organometallic complexes with amino acids while Cr was identified with several oxidation states, including hexavalent Cr. The cytotoxicity of the released metal ions was evaluated using experiments on 2D cell cultures. The experiments indicated that critical concentrations of Co ions led to chondrocyte morphology changes and their apoptosis. Ongoing studies of gene expression and cell viability will reveal in detail the effects of metal ions and particles on human chondrocytes. Biotribocorrosion experiments can be used to evaluate the stability of the passive film on metallic surfaces articulating against native cartilage and contribute to the development of more cartilage-friendly materials. A correlation between metal release during biotribocorrosion experiments and cell culture tests will contribute to assess the role of biotribocorrosion and determine the dominant failure mechanisms of partial replacement technologies with the ultimate goal of extending their lifetime. Future studies will include synthetic synovial fluid, elevated temperature, and anaerobic conditions in order to better simulate in-vivo conditions.

[3] [4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

Acknowledgments [19]

This work was funded by NÖ Forschungs- und Bildungsges.m.b. H. (NFB) with the project (Project ID: LSC15-019) and the provincial government of Lower Austria through the Life Science Calls (Project ID: LSC15-019) and has been carried out within the Austrian Excellence Centre of Tribology (AC2T research GmbH) in cooperation with Danube-University Krems – University for Continuing Education and Landesklinikum Baden-Mödling. The authors also gratefully acknowledge funds for parts of this work from the Austrian COMET-Program (project XTribology, no. 849109) via the Austrian Research Promotion Agency (FFG) with the project XTribology, no. 849109 and the Province of Lower Austria, Vorarlberg and Wien. Adam Agocs is gratefully acknowledged for his support with the ICP measurements. Dr. Christian Tomastik and Dr. Andjelka Ristic´ are acknowledged for performing the XPS measurements and MS analyses, respectively. References [1] S.R. Oungoulian, K.M. Durney, B.K. Jones, C.S. Ahmad, C.T. Hung, G.A. Ateshian, Wear and damage of articular cartilage with friction against orthopedic implant materials, J. Biomech. 48 (2015) 1957–1964, https://doi.org/10.1016/j. jbiomech.2015.04.008. [2] W. Figved, J. Dahl, F. Snorrason, F. Frihagen, S. Röhrl, J.E. Madsen, L. Nordsletten, Radiostereometric analysis of hemiarthroplasties of the hip – a

[20]

[21]

[22]

[23]

[24]

[25]

[26]

highly precise method for measurements of cartilage wear, Osteoarthr. Cartil. 20 (2012) 36–42, https://doi.org/10.1016/j.joca.2011.11.006. S. Jahn, J. Seror, J. Klein, Lubrication of articular cartilage, Annu. Rev. Biomed. Eng. 18 (2016) 235–258, https://doi.org/10.1063/PT.3.3898. S.M.T. Chan, C.P. Neu, K. Komvopoulos, A.H. Reddi, P.E. Di Cesare, Friction and wear of hemiarthroplasty biomaterials in reciprocating sliding contact with articular cartilage, J. Tribol. 133 (2011) 041201, https://doi.org/10.1115/1.4004760. E.D. Bonnevie, V.J. Baro, L. Wang, D.L. Burris, In situ studies of cartilage microtribology: roles of speed and contact area, Tribol. Lett. 41 (2011) 83–95, https://doi.org/10.1007/s11249-010-9687-0. ˇ . Donik, J.T. Grant, D. M. Jenko, M. Gorenšek, M. Godec, M. Hodnik, B.Š. Baticˇ, C Dolinar, Surface chemistry and microstructure of metallic biomaterials for hip and knee endoprostheses, Appl. Surf. Sci. 427 (2018) 584–593, https://doi.org/ 10.1016/j.apsusc.2017.08.007. A.W.E. Hodgson, S. Kurz, S. Virtanen, V. Fervel, C.O.A. Olsson, S. Mischler, Passive and transpassive behaviour of CoCrMo in simulated biological solutions, Electrochim. Acta 49 (2004) 2167–2178, https://doi.org/10.1016/ j.electacta.2003.12.043. D. Landolt, S. Mischler, M. Stemp, Electrochemical methods in tribocorrosion: a critical appraisal, Electrochim. Acta 46 (2001) 3913–3929, https://doi.org/ 10.1016/S0013-4686(01)00679-X. S. Mischler, Triboelectrochemical techniques and interpretation methods in tribocorrosion: a comparative evaluation, Tribol. Int. 41 (2008) 573–583, https://doi.org/10.1016/j.triboint.2007.11.003. S. Mischler, A. Igual Muñoz, Wear of CoCrMo alloys used in metal-on-metal hip joints: a tribocorrosion appraisal, Wear 297 (2013) 1081–1094, https://doi. org/10.1016/j.wear.2012.11.061. M.T. Mathew, M.J. Runa, M. Laurent, J.J. Jacobs, L.A. Rocha, M.A. Wimmer, Tribocorrosion behavior of CoCrMo alloy for hip prosthesis as a function of loads: a comparison between two testing systems, Wear 271 (2011) 1210– 1219, https://doi.org/10.1016/j.wear.2011.01.086. A. Igual Muñoz, L. Casabán Julián, Influence of electrochemical potential on the tribocorrosion behaviour of high carbon CoCrMo biomedical alloy in simulated body fluids by electrochemical impedance spectroscopy, Electrochim. Acta 55 (2010) 5428–5439, https://doi.org/10.1016/j.electacta.2010.04.093. Leandre Casabán Julián, Anna Igual Muñoz, Influence of microstructure of HC CoCrMo biomedical alloys on the corrosion and wear behaviour in simulated body fluids, Tribol. Int. 44 (3) (2011) 318–329, https://doi.org/10.1016/j. triboint.2010.10.033. V. Martinez-Nogues, J.M. Nesbitt, R.J.K. Wood, R.B. Cook, Nano-scale wear characterization of CoCrMo biomedical alloys, Tribol. Int. 93 (2016) 563–572, https://doi.org/10.1016/j.triboint.2015.03.037. N. Espallargas, C. Torres, A. Igual Muñoz, A metal ion release study of CoCrMo exposed to corrosion and tribocorrosion conditions in simulated body fluids, Wear 332–333 (2015) 669–678, https://doi.org/10.1016/j.wear.2014.12.030. J. Hesketh, Q. Meng, D. Dowson, A. Neville, Biotribocorrosion of metal-onmetal hip replacements: how surface degradation can influence metal ion formation, Tribol. Int. 65 (2013) 128–137, https://doi.org/10.1016/j. triboint.2013.02.025. R.L. Trevino, J. Stoia, M.P. Laurent, C.A. Pacione, S. Chubinskaya, M.A. Wimmer, Establishing a live cartilage-on-cartilage interface for tribological testing, Biotribology 9 (2017) 1–11, https://doi.org/10.1016/j.biotri.2016.11.002. A. Beraudi, S. Catalani, M. Montesi, S. Stea, A. Sudanese, P. Apostoli, A. Toni, Detection of cobalt in synovial fluid from metal-on-metal hip prosthesis: correlation with the ion haematic level, Biomarkers 18 (2013) 699–705, https://doi.org/10.3109/1354750X.2013.846413. F. Eltit, A. Assiri, D. Garbuz, C. Duncan, B. Masri, N. Greidanus, R. Bell, M. Sharma, M. Cox, R. Wang, Adverse reactions to metal on polyethylene implants: highly destructive lesions related to elevated concentration of cobalt and chromium in synovial fluid, J. Biomed. Mater. Res. Part A 105 (2017) 1876–1886, https://doi.org/10.1002/jbm.a.36057. S. Gilbert, T. Chen, I.D. Hutchinson, D. Choi, C. Voigt, R.F. Warren, S.A. Maher, Dynamic contact mechanics on the tibial plateau of the human knee during activities of daily living, J. Biomech. 47 (2014) 2006–2012, https://doi.org/ 10.1016/j.jbiomech.2013.11.003. D.D. Dlima, S. Hashimoto, P.C. Chen, C.W. Colwell, M.K. Lotz, Human chondrocyte apoptosis in response to mechanical injury, Osteoarthr. Cartil. 9 (2001) 712–719, https://doi.org/10.1053/joca.2001.0468. G. Verberne, Y. Merkher, G. Halperin, A. Maroudas, I. Etsion, Techniques for assessment of wear between human cartilage surfaces, Wear 266 (2009) 1216–1223, https://doi.org/10.1016/j.wear.2009.03.042. B. Hafke, M. Petri, E. Suero, C. Neunaber, S. Kwisda, C. Krettek, M. Jagodzinski, M. Omar, Chondrocyte survival in osteochondral transplant cylinders depends on the harvesting technique, Int. Orthop. 40 (2016) 1553–1558, https://doi. org/10.1007/s00264-015-3065-3. V. Pejakovic´, V. Totolin, H. Göcerler, J. Brenner, M. Rodríguez Ripoll, Friction and wear behaviour of selected titanium and zirconium based nitride coatings in Na2SO4 aqueous solution under low contact pressure, Tribol. Int. 91 (2015) 267–273, https://doi.org/10.1016/j.triboint.2015.04.047. M. Caligaris, G.A. Ateshian, Effects of sustained interstitial fluid pressurization under migrating contact area, and boundary lubrication by synovial fluid, on cartilage friction, Osteoarthr. Cartil. 16 (2008) 1220–1227, https://doi.org/ 10.1016/j.joca.2008.02.020. R.W. Bohannon, Comfortable and maximum walking speed of adults aged 20– 79 years: reference values and determinants, Age Age. 26 (1997) 15–19, https://doi.org/10.1093/ageing/26.1.15.

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015

B. Stojanovic´ et al. / Acta Biomaterialia xxx (xxxx) xxx [27] C. Valero Vidal, A. Igual Muñoz, Study of the adsorption process of bovine serum albumin on passivated surfaces of CoCrMo biomedical alloy, Electrochim. Acta. 55 (2010) 8445–8452, https://doi.org/10.1016/ j.electacta.2010.07.028. [28] E. Bettini, C. Leygraf, J. Pan, Nature of current increase for a CoCrMo alloy: ‘‘transpassive” dissolution vs. water oxidation, Int. J. Electrochem. Sci. 8 (2013) 11791–11804. [29] A. Igual Muñoz, J. Schwiesau, B.M. Jolles, S. Mischler, In vivo electrochemical corrosion study of a CoCrMo biomedical alloy in human synovial fluids, Acta Biomater. 21 (2015) 228–236, https://doi.org/10.1016/j.actbio.2015.03.008. [30] Y. Yan, A. Neville, D. Dowson, Tribo-corrosion properties of cobalt-based medical implant alloys in simulated biological environments, Wear 263 (2007) 1105–1111, https://doi.org/10.1016/j.wear.2007.01.114. [31] A. Bazzoni, S. Mischler, N. Espallargas, Tribocorrosion of pulsed plasmanitrided cocrmo implant alloy, Tribol. Lett. 49 (2013) 157–167, https://doi.org/ 10.1007/s11249-012-0047-0. [32] Y. Yan, A. Neville, D. Dowson, Biotribocorrosion of CoCrMo orthopaedic implant materials—assessing the formation and effect of the biofilm, Tribol. Int. 40 (2007) 1492–1499, https://doi.org/10.1016/j.triboint.2007.02.019. [33] C. Valero-Vidal, L. Casabán-Julián, I. Herraiz-Cardona, A. Igual-Muñoz, Influence of carbides and microstructure of CoCrMo alloys on their metallic dissolution resistance, Mater. Sci. Eng. C. 33 (2013) 4667–4676, https://doi. org/10.1016/j.msec.2013.07.041. [34] A.C. Lewis, M.R. Kilburn, I. Papageorgiou, G.C. Allen, C.P. Case, Effect of synovial fluid, phosphate-buffered saline solution, and water on the dissolution and corrosion properties of CoCrMo alloys as used in orthopedic implants, J. Biomed. Mater. Res. Part A. 73A (2005) 456–467, https://doi.org/10.1002/jbm. a.30368. [35] M.T. Mathew, C. Nagelli, R. Pourzal, A. Fischer, M.P. Laurent, J.J. Jacobs, M.A. Wimmer, Tribolayer formation in a metal-on-metal (MoM) hip joint: an electrochemical investigation, J. Mech. Behav. Biomed. Mater. 29 (2014) 199– 212, https://doi.org/10.1016/j.jmbbm.2013.08.018.

13

[36] Y. Yan, D. Dowson, A. Neville, In-situ electrochemical study of interaction of tribology and corrosion in artificial hip prosthesis simulators, J. Mech. Behav. Biomed. Mater. 18 (2013) 191–199, https://doi.org/10.1016/j. jmbbm.2012.08.009. [37] N. Diomidis, S. Mischler, N.S. More, M. Roy, S.N. Paul, Fretting-corrosion behavior of b titanium alloys in simulated synovial fluid, Wear 271 (2011) 1093–1102, https://doi.org/10.1016/j.wear.2011.05.010. [38] S.S. Wise, A.L. Holmes, J.P. Wise, Particulate and soluble hexavalent chromium are cytotoxic and genotoxic to human lung epithelial cells, Mutat. Res. - Genet. Toxicol. Environ. Mutagen. 610 (2006) 2–7, https://doi.org/10.1016/j. mrgentox.2006.06.005. [39] Y. Henrotin, B. Kurz, T. Aigner, Oxygen and reactive oxygen species in cartilage degradation: friends or foes?, Osteoarthr Cartil. 13 (2005) 643–654, https:// doi.org/10.1016/j.joca.2005.04.002. [40] M.C. Coleman, P.S. Ramakrishnan, M.J. Brouillette, J.A. Martin, Injurious loading of articular cartilage compromises chondrocyte respiratory function, Arth. Rheumatol. 68 (2016) 662–671, https://doi.org/10.1002/art.39460. [41] K. Merritt, S.A. Brown, Release of hexavalent chromium from corrosion of stainless steel and cobalt—chromium alloys, J. Biomed. Mater. Res. 29 (1995) 627–633, https://doi.org/10.1002/jbm.820290510. [42] A. Neville, Y. Yan, Biotribocorrosion: surface interactions in total joint replacement (TJR), Tribocorr. Passiv. Met. Coat. (2011) 337–367, https://doi. org/10.1016/B978-1-84569-966-6.50012-9. [43] A. Igual Muñoz, S. Mischler, Interactive effects of albumin and phosphate ions on the corrosion of CoCrMo implant alloy, J. Electrochem. Soc. 154 (2007) C562, https://doi.org/10.1149/1.2764238. [44] K. Waller, L. Zhang, G. Jay, Friction-induced mitochondrial dysregulation contributes to joint deterioration in Prg4 knockout mice, Int. J. Mol. Sci. 18 (2017) 1252, https://doi.org/10.3390/ijms18061252. [45] A.A. Pitenis, J.M. Urueña, S.M. Hart, C.S. O’Bryan, S.L. Marshall, P.P. Levings, T.E. Angelini, W.G. Sawyer, Friction-induced inflammation, Tribol. Lett. 66 (2018) 1–13, https://doi.org/10.1007/s11249-018-1029-7.

Please cite this article as: B. Stojanovic´, C. Bauer, C. Stotter et al., Tribocorrosion of a CoCrMo alloy sliding against articular cartilage and the impact of metal ion release on chondrocytes, Acta Biomaterialia, https://doi.org/10.1016/j.actbio.2019.06.015