Acta Biomaterialia xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat
Full length article
Release of zirconia nanoparticles at the metal stem–bone cement interface in implant loosening of total hip replacements Antje Schunck a, Andreas Kronz b, Cornelius Fischer c, Gottfried Hans Buchhorn a,⇑ a
University Hospital Göttingen, Department of Orthopedics/Biomaterials Research Laboratory, Robert-Koch-Str. 40, 37075 Göttingen, Germany University of Göttingen, Department of Geochemistry, Goldschmidtstr. 1, 37077 Göttingen, Germany c University of Bremen, MARUM/Department of Geosciences, Klagenfurter Str., 28359 Bremen, Germany b
a r t i c l e
i n f o
Article history: Received 28 May 2015 Received in revised form 16 November 2015 Accepted 19 November 2015 Available online xxxx Keywords: Bone cement Zirconia Contrast medium Abrasion Polishing Surface roughness
a b s t r a c t In a previous failure analysis performed on femoral components of cemented total hip replacements, we determined high volumes of abraded bone cement. Here, we describe the topography of the polished surface of polymethyl methacrylate (PMMA) bone cement containing zirconia radiopacifier, analyzed by scanning electron microscopy and vertical scanning interferometry. Zirconia spikes protruded about 300 nm from the PMMA matrix, with pits of former crystal deposition measuring about 400 nm in depth. We deduced that the characteristically mulberry-shaped agglomerates of zirconia crystals are ground and truncated into flat surfaces and finally torn out of the PMMA matrix. Additionally, evaluation of in vitro PMMA-on-PMMA articulation confirmed that crystal agglomerations of zirconia were exposed to grain pullout, fatigue, and abrasion. In great quantities, micron-sized PMMA wear and zirconia nanoparticles accumulate in the cement–bone interface and capsular tissues, thereby contributing to osteolysis. Dissemination of nanoparticles to distant lymph nodes and organs of storage has been reported. As sufficient information is lacking, foreign body reactions to accumulated nanosized zirconia in places of longterm storage should be investigated. Statement of Significance The production of wear particles of PMMA bone cement in the interface to joint replacement devices, presents a local challenge. The presence of zirconia particles results in frustrated digestion attempts by macrophages, liberation of inflammatory mediators, and necrosis leading to aseptic inflammation and osteolyses. Attempts to minimize wear of articulating joints reduced the attention to the deterioration of cement cuffs. We therefore investigated polished surfaces of retrieved cuffs to demonstrate their morphology and to measure surface roughness. Industrially admixed agglomerates of the radiopacifier are abraded to micron and nano-meter sized particles. The dissemination of zirconia particles in the reticulo-endothelial system to storage organs is a possible burden. Research to replace the actual contrast media by non-particulate material deserves more attention. Ó 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
1. Introduction 1.1. Background Resorption of anchoring bone due to a foreign body reaction at the device-bone interface is a major complication in aseptic loosening of cemented total joint replacements [1,2]. The loss of con⇑ Corresponding author. E-mail addresses: [email protected] (A. Schunck), [email protected] (A. Kronz), [email protected] (C. Fischer), [email protected] (G.H. Buchhorn).
straint [3] resulting from an intermediate layer of soft tissue and mechanical overload have been ascribed to fragmentation of the cement mantle (polymethyl methacrylate, PMMA). Among others, Wang and co-workers based their criticism of the clinical performance of cement anchorage on the production of cement debris. They concluded that significant synergistically pathological reactions to the polymer containing customary radiopacifier (barium sulfate or zirconium dioxide (zirconia)) not only impair articulation of the joint, but also contribute to bone resorption [4]. In the granulation tissues of the cement–bone interface, macrophages with foamy cytoplasm have been described as containing micron and submicron wear particles. Additionally, giant cells engulf
http://dx.doi.org/10.1016/j.actbio.2015.11.044 1742-7061/Ó 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: A. Schunck et al., Release of zirconia nanoparticles at the metal stem–bone cement interface in implant loosening of total hip replacements, Acta Biomater. (2015), http://dx.doi.org/10.1016/j.actbio.2015.11.044
2
A. Schunck et al. / Acta Biomaterialia xxx (2015) xxx–xxx
greater particles representing PMMA pearls and matrices, fragments thereof, and conglomerates of zirconia contrast media [5]. During mechanical loosening, instability and relative motion of anchoring stems under loading create abrasion of both metal component and contact surfaces of bone cement within a relatively short time. The loss of submicron material from cement mantles may reach volumes as high as of up to 1 or 2 cm3 [6]. In previous studies, we observed cases of implant loosening over several years with displacement and rotation of the stem resulting in moderate but persistent abrasion of bone cement particles [7]. Light microscopic examinations of interface granulation tissues have demonstrated predominant storage of zirconia contrast medium that were measurable within the detection limit range of light microscopy (diameter of about 0.5 lm). No fragments of PMMA pearls and matrix were present. The welldescribed mulberry- or cauliflower-shaped agglomerates of the admixed zirconia radiopacifier or its larger fragments were not found in these granulation tissues by our studies. Repetition of foreign body reactions to nanoparticles of the zirconia radiopacifier after apoptosis of resting macrophages may initiate and sustain mediator-induced activation of osteolysis (frustrated digestion) [8]. Widespread dissemination of particulate foreign material and accumulation in distant organs has been well described [9,10]. Foreign body reactions to micron- and nanoscaled zirconia particles in particle accumulation sites represent possible health hazards that deserve closer follow-up of total joint replacements with longterm function and post mortem histopathology whenever possible.
1.2. Aims Limited information is available on the morphology of worn PMMA surfaces, the fate and behavior of the zirconia conglomerates, and on the generation of wear products. Herein, we continue a retrieval analysis of cemented total hip replacements. In the previous publication [6], failure induction and propagation was attributed to special implant design features, positioning in the cement dough, and sandblasted surfaces of the implanted femoral device. The metal stem rubbed against the PMMA bone cement and abraded the polymer matrix as well as the zirconia conglomerates
added as radiopacifier. Our analysis demonstrated cement surfaces polished to a high finish quality [6]. Here we analyzed the topography of polished zones, and sizes of structures within zirconia agglomerates and of crystals were measured. We observed in vitro preparations of the basic material for comparison of surface characteristics with retrieved samples. As no information on the bonding strength of the zirconia within the PMMA and the coherence of agglomerated crystals was available, we surmised that a moderate in vitro process of abrasion would demonstrate whether disintegration and polishing of zirconia agglomerates is probable. Our aim was to test the hypothesis that movements of the femoral shaft relative to the cement cuff liberate nanoparticles and additionally deteriorate agglomerates of the radiopacifier zirconia.
2. Materials and methods 2.1. Origin of materials Over a 9-year period, 393 total hip joint replacements (‘‘CF-30” femoral stems, stainless steel – ISO 5832-9 and all-polyethylene sockets made of UHMW-PE – ISO 5834/1 type A and manufactured by Sulzer Orthopedics AG, Winterthur, Switzerland) were implanted in patients. Throughout all procedures, Palacos RÒ bone cement (plain PMMA cement with ZrO2 radiopacifier, Heraeus AG, Wehrheim, Germany) was prepared by vacuum mixing and inserted by antegrade syringe application. Alumina ceramic balls (Biolox Ò, CeramTec AG, Plochingen, Germany) completed the joint replacement (Fig. 1a). A high, above-average number of femoral components (about 9.5% after an implantation period of 5 years) had to be removed due to aseptic loosening following mechanical failure of the cement mantles and overload of bone. During replacement surgery, intermediate layers of soft tissue facilitated extraction of cement mantles. The surgeons attempted to extract segments as large and unimpaired as possible (Fig. 1a). In a previous retrieval analysis [6] of 37 retrieved total joint replacements (stems and cement mantles), the fragments were cleaned in detergent, sonicated and rinsed with tap water, and finally dried and stored in the dark at room temperature. The original position of the fragments was mapped.
Fig. 1. (a) Retrieved CF-30 femoral stem with rearranged fragments of cement mantle. Note fragmentation, inappropriate thinning of cement mantle and cementing defect leading to metal-bone contact. Ball head diameter 28 mm. (b) PMMA macro: part of cement mantle fragment from the ventro-lateral side at the lower proximal third of the mantle, lower (distal) and upper (proximal) cuts are rectangular to the axis of the stem. The left border corresponded with the medial bow of the stem, the right with the lateral side. Two zones, (a) of original primary contact and (b) deteriorated by abrasion and polishing, depict the surface of metal contact. The bar measures 5 mm.
Please cite this article in press as: A. Schunck et al., Release of zirconia nanoparticles at the metal stem–bone cement interface in implant loosening of total hip replacements, Acta Biomater. (2015), http://dx.doi.org/10.1016/j.actbio.2015.11.044
A. Schunck et al. / Acta Biomaterialia xxx (2015) xxx–xxx
Out of these, bone cement samples from 13 cases were selected for scanning electron microscopy (SEM) and from 5 for vertical infrared scanning (VIS). An additional femoral component with clearly differentiated surface characteristics (area with (a) no contact with cement, (b) cement contact, and (c) mirror polished) was used for hardness measurements. In accordance with Good Laboratory Practice guidelines, all specimens were pseudonymized. A zirconia powder representative of radiopacifier in Palacos RÒ bone cement provided by Heraeus Medical GmbH, Wehrheim, Germany (lot No. 10/313/01) was used for comparison. 2.2. Methods A binocular lens with magnification up to 30 (M5 APO Wild, Herbrugg AG, Switzerland) was used for reflected-light macroscopy. To inspect the size and volume distribution of zirconia powder, we employed Laser Scintillation Particle Size Analysis (LS-PSA) with LS13320 (Beckmann Coulter GmbH, Krefeld, Germany; Universal Liquid Module, Optical Model: Fraunhofer.rf780d PIDS included, measurement range 0.04 lm to 2 mm). Prior to measurements, the particles were sonicated in ethanol (RK255H, BANDELIN Electronic GmbH, Berlin, Germany); all measurements were taken at least in duplicate. The Rockwell hardness (HRC on the C scale) of the metal femoral stem was investigated (MicroDur II with Mic21 readings recorder, GE Inspection Technologies GmbH, Hürth, Germany) in both the as-delivered shot-peened and in the abraded state for comparison with the zirconia reference data. Five measurements of each surface characteristic were performed on the representative femoral stem. A sample of the Heraeus representative of zirconia powder was embedded in epoxy resin (EpoFix, Struers GmbH, Willich, Germany) and cured in a vacuum to ensure vesicle-free contact. After hardening overnight at 50 °C, the sample was ground (plasticbound diamond P240, P1200), lapped (silicon carbide, grain size 1200), and polished (diamond 3 lm, 1 lm, corundum 0.05 lm). The fine polishing enabled cross-sections of zirconia powder to be even smaller than 1 lm. The surfaces were gold-coated to a thickness of about 20 nm for SEM analysis. A cement-on-cement low-pressure articulation was established (modification of method previously described [11]). The intent was to avoid metal wear as well as high-temperature smear of the bone cement polymer during in vitro wear particle generation. A magnetic stirrer agitated a magnet within a cylindrical rotor in a depression at about 1 Hz (Fig. 3a). The difference in inner/outer diameter of the statorrotator assembly allowed for tumbling movements. The rotor was top-loaded by an external magnet to apply pressure to the contact surfaces of about 1.5 kPa. After 5 days of continuous rotation, the lubricant was collected and the underside of the rotator was cut off (Exakt grinding band saw, Exakt GmbH, Hamburg, Germany) for SEM inspection of the polished surfaces. Samples of cement cuffs were cut into halves perpendicular to the stem axis with the water-cooled, low-speed diamond grinding band saw for SEM, electron probe microanalysis (EPMA) as well as VSI. Bone cement surfaces opposing the metal stem of the joint replacements were the regions of interest. After carbon, gold–palladium or gold coating, we applied both secondary element mode and backscatter electron mode (DSM 960, Zeiss, Oberkochen, Germany). Secondary electron images (SEIs) revealed the topography, while the corresponding backscatter image (BSI) depicted material contrast. Magnification was limited to up to 5000 due to thermal decomposition of the PMMA. Gold–palladium- or carbon-coated bone cement specimens were examined for element detection using an EPMA Superprobe (JXA 8900 RL, Jeol Ltd. Tokyo, Japan) in WDS (wavelengthdispersive X-ray spectrometry)/EDS (energy-dispersive X-ray spec-
3
trometry) mode. Here too, thermal damage of the bone cement material PMMA limited resolution to 1000, especially when beam currents typical for elemental analysis of 2–12 nA were used at an accelerating voltage of 15 kV. Decreasing the beam current to 20– 50 pA allowed magnifications up to 20,000 and displayed only moderate beam-induced sample damage. For topography of zirconia particles, powder was dusted on carbon-conductive adhesive tape (Plano GmbH, Wetzlar, Germany) and carbon-coated. Element analysis was concentrated on zirconium and hafnium. Drops of water (lubricant) containing the laboratory-produced bone cement wear were applied to glass cover slips, air dried, and gold-coated. The vertical scanning interferometry (VSI) technique (ZeMapper, Zemetrics Inc., Tucson, AZ, USA) was employed to obtain surface topography data of the bone cement material with high spatial resolution and a large field of view (FOV) [12]). The light source in white-light mode was a light-emitting diode (LED) with center wavelength k = 560 nm. VSI-samples were sputter-coated with Au (layer thickness: 80 nm) to homogenize the light reflectivity of the complex material surfaces. Owing to overall surface waviness and heterogeneity in the millimeter range of the investigated material, the analyses were limited to surface sections with FOV < 150 lm 150 lm. The resultant height map data provided information about the height differences of the zirconia–bone cement transitions and grain boundaries within the zirconia agglomerates. The small FOV provided background topographies with limited impact of large-scale and irregular height differences. This strategy enabled detailed analysis of investigation-relevant small height differences between zirconia crystals and cement material. Measurements of length were performed using a 100 Mirau interferometer objective with and without a 1.6 magnification lens. According to the analysis of converged roughness parameters [13] the maximum FOVs were 150 lm 150 lm and 93 lm 93 lm, respectively. The virtual pixel sizes of the resulting data were 73 nm 73 nm and 46 nm 46 nm. Surface topography data were analyzed using SPIP software (Image Metrology A/S, Hørsholm, Denmark), version 5.0.5. 3. Results After examination of clinical data and the retrieved cement mantles, we investigated the control specimens and bone cement samples prepared in the laboratory for better interpretation of findings. 3.1. Clinical data Twelve patients with unilateral degenerative disease of the hip and one with a femoral neck fracture had received a total hip replacement. After an average age of 80 years (range 73–87) and average 4 years and 9 months of functional implantation time (range 2.8–7.10), the femoral components were removed. The body mass index at implantation ranged between 20.6 and 37.3; three patients had grade I, and one grade II obesity. Fourteen specimens from the bone cement cuffs originated from Gruen zones 1 and 7, six from Gruen zones 2 and 6, and five from Gruen zones 3 and 5 [14]. The straight conical stem, a close taper angle, and a rectangular cross-section with a small radius of rounded corners of the component presented here (Fig. 1a) had caused fractures of the cement cuff. The frequent displacement of the femoral stem was a characteristic of mediodorsal rotation over a longitudinal axis and a cranial tilt over a tilt point variable in height. The deflection of the femoral stem from cement surfaces resulted in a relatively sharp transition from fairly well-preserved replicas of the shot-peened
Please cite this article in press as: A. Schunck et al., Release of zirconia nanoparticles at the metal stem–bone cement interface in implant loosening of total hip replacements, Acta Biomater. (2015), http://dx.doi.org/10.1016/j.actbio.2015.11.044
4
A. Schunck et al. / Acta Biomaterialia xxx (2015) xxx–xxx
metal surface to abraded and polished surfaces (Fig. 4b) [15]. Soft tissue depositions filled permanent gaps between stem and cement.
undisturbed rough surfaces, a hardness of HR 34 to 43 was found. For comparison, the hardness of zirconia was given as HR 71, and of CoCr alloy as about 60 HR.
3.2. Macroscopy
3.3.2. Pure zirconia: size distribution The particle size distribution (LS-PSA) of original contrast medium powder (ZrO2) of Palacos RÒ bone cement is shown in Fig. 2a and Table 1. The volume share (%) showed a sharp peak in the diameter range of about 20 lm. Thus, big agglomerates dominated the volume distribution. This was the main impression given by light microscopy. Nevertheless, a considerably smaller mean diameter (
Contents lists available at ScienceDirect
Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat
Full length article
Release of zirconia nanoparticles at the metal stem–bone cement interface in implant loosening of total hip replacements Antje Schunck a, Andreas Kronz b, Cornelius Fischer c, Gottfried Hans Buchhorn a,⇑ a
University Hospital Göttingen, Department of Orthopedics/Biomaterials Research Laboratory, Robert-Koch-Str. 40, 37075 Göttingen, Germany University of Göttingen, Department of Geochemistry, Goldschmidtstr. 1, 37077 Göttingen, Germany c University of Bremen, MARUM/Department of Geosciences, Klagenfurter Str., 28359 Bremen, Germany b
a r t i c l e
i n f o
Article history: Received 28 May 2015 Received in revised form 16 November 2015 Accepted 19 November 2015 Available online xxxx Keywords: Bone cement Zirconia Contrast medium Abrasion Polishing Surface roughness
a b s t r a c t In a previous failure analysis performed on femoral components of cemented total hip replacements, we determined high volumes of abraded bone cement. Here, we describe the topography of the polished surface of polymethyl methacrylate (PMMA) bone cement containing zirconia radiopacifier, analyzed by scanning electron microscopy and vertical scanning interferometry. Zirconia spikes protruded about 300 nm from the PMMA matrix, with pits of former crystal deposition measuring about 400 nm in depth. We deduced that the characteristically mulberry-shaped agglomerates of zirconia crystals are ground and truncated into flat surfaces and finally torn out of the PMMA matrix. Additionally, evaluation of in vitro PMMA-on-PMMA articulation confirmed that crystal agglomerations of zirconia were exposed to grain pullout, fatigue, and abrasion. In great quantities, micron-sized PMMA wear and zirconia nanoparticles accumulate in the cement–bone interface and capsular tissues, thereby contributing to osteolysis. Dissemination of nanoparticles to distant lymph nodes and organs of storage has been reported. As sufficient information is lacking, foreign body reactions to accumulated nanosized zirconia in places of longterm storage should be investigated. Statement of Significance The production of wear particles of PMMA bone cement in the interface to joint replacement devices, presents a local challenge. The presence of zirconia particles results in frustrated digestion attempts by macrophages, liberation of inflammatory mediators, and necrosis leading to aseptic inflammation and osteolyses. Attempts to minimize wear of articulating joints reduced the attention to the deterioration of cement cuffs. We therefore investigated polished surfaces of retrieved cuffs to demonstrate their morphology and to measure surface roughness. Industrially admixed agglomerates of the radiopacifier are abraded to micron and nano-meter sized particles. The dissemination of zirconia particles in the reticulo-endothelial system to storage organs is a possible burden. Research to replace the actual contrast media by non-particulate material deserves more attention. Ó 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
1. Introduction 1.1. Background Resorption of anchoring bone due to a foreign body reaction at the device-bone interface is a major complication in aseptic loosening of cemented total joint replacements [1,2]. The loss of con⇑ Corresponding author. E-mail addresses: [email protected] (A. Schunck), [email protected] (A. Kronz), [email protected] (C. Fischer), [email protected] (G.H. Buchhorn).
straint [3] resulting from an intermediate layer of soft tissue and mechanical overload have been ascribed to fragmentation of the cement mantle (polymethyl methacrylate, PMMA). Among others, Wang and co-workers based their criticism of the clinical performance of cement anchorage on the production of cement debris. They concluded that significant synergistically pathological reactions to the polymer containing customary radiopacifier (barium sulfate or zirconium dioxide (zirconia)) not only impair articulation of the joint, but also contribute to bone resorption [4]. In the granulation tissues of the cement–bone interface, macrophages with foamy cytoplasm have been described as containing micron and submicron wear particles. Additionally, giant cells engulf
http://dx.doi.org/10.1016/j.actbio.2015.11.044 1742-7061/Ó 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: A. Schunck et al., Release of zirconia nanoparticles at the metal stem–bone cement interface in implant loosening of total hip replacements, Acta Biomater. (2015), http://dx.doi.org/10.1016/j.actbio.2015.11.044
2
A. Schunck et al. / Acta Biomaterialia xxx (2015) xxx–xxx
greater particles representing PMMA pearls and matrices, fragments thereof, and conglomerates of zirconia contrast media [5]. During mechanical loosening, instability and relative motion of anchoring stems under loading create abrasion of both metal component and contact surfaces of bone cement within a relatively short time. The loss of submicron material from cement mantles may reach volumes as high as of up to 1 or 2 cm3 [6]. In previous studies, we observed cases of implant loosening over several years with displacement and rotation of the stem resulting in moderate but persistent abrasion of bone cement particles [7]. Light microscopic examinations of interface granulation tissues have demonstrated predominant storage of zirconia contrast medium that were measurable within the detection limit range of light microscopy (diameter of about 0.5 lm). No fragments of PMMA pearls and matrix were present. The welldescribed mulberry- or cauliflower-shaped agglomerates of the admixed zirconia radiopacifier or its larger fragments were not found in these granulation tissues by our studies. Repetition of foreign body reactions to nanoparticles of the zirconia radiopacifier after apoptosis of resting macrophages may initiate and sustain mediator-induced activation of osteolysis (frustrated digestion) [8]. Widespread dissemination of particulate foreign material and accumulation in distant organs has been well described [9,10]. Foreign body reactions to micron- and nanoscaled zirconia particles in particle accumulation sites represent possible health hazards that deserve closer follow-up of total joint replacements with longterm function and post mortem histopathology whenever possible.
1.2. Aims Limited information is available on the morphology of worn PMMA surfaces, the fate and behavior of the zirconia conglomerates, and on the generation of wear products. Herein, we continue a retrieval analysis of cemented total hip replacements. In the previous publication [6], failure induction and propagation was attributed to special implant design features, positioning in the cement dough, and sandblasted surfaces of the implanted femoral device. The metal stem rubbed against the PMMA bone cement and abraded the polymer matrix as well as the zirconia conglomerates
added as radiopacifier. Our analysis demonstrated cement surfaces polished to a high finish quality [6]. Here we analyzed the topography of polished zones, and sizes of structures within zirconia agglomerates and of crystals were measured. We observed in vitro preparations of the basic material for comparison of surface characteristics with retrieved samples. As no information on the bonding strength of the zirconia within the PMMA and the coherence of agglomerated crystals was available, we surmised that a moderate in vitro process of abrasion would demonstrate whether disintegration and polishing of zirconia agglomerates is probable. Our aim was to test the hypothesis that movements of the femoral shaft relative to the cement cuff liberate nanoparticles and additionally deteriorate agglomerates of the radiopacifier zirconia.
2. Materials and methods 2.1. Origin of materials Over a 9-year period, 393 total hip joint replacements (‘‘CF-30” femoral stems, stainless steel – ISO 5832-9 and all-polyethylene sockets made of UHMW-PE – ISO 5834/1 type A and manufactured by Sulzer Orthopedics AG, Winterthur, Switzerland) were implanted in patients. Throughout all procedures, Palacos RÒ bone cement (plain PMMA cement with ZrO2 radiopacifier, Heraeus AG, Wehrheim, Germany) was prepared by vacuum mixing and inserted by antegrade syringe application. Alumina ceramic balls (Biolox Ò, CeramTec AG, Plochingen, Germany) completed the joint replacement (Fig. 1a). A high, above-average number of femoral components (about 9.5% after an implantation period of 5 years) had to be removed due to aseptic loosening following mechanical failure of the cement mantles and overload of bone. During replacement surgery, intermediate layers of soft tissue facilitated extraction of cement mantles. The surgeons attempted to extract segments as large and unimpaired as possible (Fig. 1a). In a previous retrieval analysis [6] of 37 retrieved total joint replacements (stems and cement mantles), the fragments were cleaned in detergent, sonicated and rinsed with tap water, and finally dried and stored in the dark at room temperature. The original position of the fragments was mapped.
Fig. 1. (a) Retrieved CF-30 femoral stem with rearranged fragments of cement mantle. Note fragmentation, inappropriate thinning of cement mantle and cementing defect leading to metal-bone contact. Ball head diameter 28 mm. (b) PMMA macro: part of cement mantle fragment from the ventro-lateral side at the lower proximal third of the mantle, lower (distal) and upper (proximal) cuts are rectangular to the axis of the stem. The left border corresponded with the medial bow of the stem, the right with the lateral side. Two zones, (a) of original primary contact and (b) deteriorated by abrasion and polishing, depict the surface of metal contact. The bar measures 5 mm.
Please cite this article in press as: A. Schunck et al., Release of zirconia nanoparticles at the metal stem–bone cement interface in implant loosening of total hip replacements, Acta Biomater. (2015), http://dx.doi.org/10.1016/j.actbio.2015.11.044
A. Schunck et al. / Acta Biomaterialia xxx (2015) xxx–xxx
Out of these, bone cement samples from 13 cases were selected for scanning electron microscopy (SEM) and from 5 for vertical infrared scanning (VIS). An additional femoral component with clearly differentiated surface characteristics (area with (a) no contact with cement, (b) cement contact, and (c) mirror polished) was used for hardness measurements. In accordance with Good Laboratory Practice guidelines, all specimens were pseudonymized. A zirconia powder representative of radiopacifier in Palacos RÒ bone cement provided by Heraeus Medical GmbH, Wehrheim, Germany (lot No. 10/313/01) was used for comparison. 2.2. Methods A binocular lens with magnification up to 30 (M5 APO Wild, Herbrugg AG, Switzerland) was used for reflected-light macroscopy. To inspect the size and volume distribution of zirconia powder, we employed Laser Scintillation Particle Size Analysis (LS-PSA) with LS13320 (Beckmann Coulter GmbH, Krefeld, Germany; Universal Liquid Module, Optical Model: Fraunhofer.rf780d PIDS included, measurement range 0.04 lm to 2 mm). Prior to measurements, the particles were sonicated in ethanol (RK255H, BANDELIN Electronic GmbH, Berlin, Germany); all measurements were taken at least in duplicate. The Rockwell hardness (HRC on the C scale) of the metal femoral stem was investigated (MicroDur II with Mic21 readings recorder, GE Inspection Technologies GmbH, Hürth, Germany) in both the as-delivered shot-peened and in the abraded state for comparison with the zirconia reference data. Five measurements of each surface characteristic were performed on the representative femoral stem. A sample of the Heraeus representative of zirconia powder was embedded in epoxy resin (EpoFix, Struers GmbH, Willich, Germany) and cured in a vacuum to ensure vesicle-free contact. After hardening overnight at 50 °C, the sample was ground (plasticbound diamond P240, P1200), lapped (silicon carbide, grain size 1200), and polished (diamond 3 lm, 1 lm, corundum 0.05 lm). The fine polishing enabled cross-sections of zirconia powder to be even smaller than 1 lm. The surfaces were gold-coated to a thickness of about 20 nm for SEM analysis. A cement-on-cement low-pressure articulation was established (modification of method previously described [11]). The intent was to avoid metal wear as well as high-temperature smear of the bone cement polymer during in vitro wear particle generation. A magnetic stirrer agitated a magnet within a cylindrical rotor in a depression at about 1 Hz (Fig. 3a). The difference in inner/outer diameter of the statorrotator assembly allowed for tumbling movements. The rotor was top-loaded by an external magnet to apply pressure to the contact surfaces of about 1.5 kPa. After 5 days of continuous rotation, the lubricant was collected and the underside of the rotator was cut off (Exakt grinding band saw, Exakt GmbH, Hamburg, Germany) for SEM inspection of the polished surfaces. Samples of cement cuffs were cut into halves perpendicular to the stem axis with the water-cooled, low-speed diamond grinding band saw for SEM, electron probe microanalysis (EPMA) as well as VSI. Bone cement surfaces opposing the metal stem of the joint replacements were the regions of interest. After carbon, gold–palladium or gold coating, we applied both secondary element mode and backscatter electron mode (DSM 960, Zeiss, Oberkochen, Germany). Secondary electron images (SEIs) revealed the topography, while the corresponding backscatter image (BSI) depicted material contrast. Magnification was limited to up to 5000 due to thermal decomposition of the PMMA. Gold–palladium- or carbon-coated bone cement specimens were examined for element detection using an EPMA Superprobe (JXA 8900 RL, Jeol Ltd. Tokyo, Japan) in WDS (wavelengthdispersive X-ray spectrometry)/EDS (energy-dispersive X-ray spec-
3
trometry) mode. Here too, thermal damage of the bone cement material PMMA limited resolution to 1000, especially when beam currents typical for elemental analysis of 2–12 nA were used at an accelerating voltage of 15 kV. Decreasing the beam current to 20– 50 pA allowed magnifications up to 20,000 and displayed only moderate beam-induced sample damage. For topography of zirconia particles, powder was dusted on carbon-conductive adhesive tape (Plano GmbH, Wetzlar, Germany) and carbon-coated. Element analysis was concentrated on zirconium and hafnium. Drops of water (lubricant) containing the laboratory-produced bone cement wear were applied to glass cover slips, air dried, and gold-coated. The vertical scanning interferometry (VSI) technique (ZeMapper, Zemetrics Inc., Tucson, AZ, USA) was employed to obtain surface topography data of the bone cement material with high spatial resolution and a large field of view (FOV) [12]). The light source in white-light mode was a light-emitting diode (LED) with center wavelength k = 560 nm. VSI-samples were sputter-coated with Au (layer thickness: 80 nm) to homogenize the light reflectivity of the complex material surfaces. Owing to overall surface waviness and heterogeneity in the millimeter range of the investigated material, the analyses were limited to surface sections with FOV < 150 lm 150 lm. The resultant height map data provided information about the height differences of the zirconia–bone cement transitions and grain boundaries within the zirconia agglomerates. The small FOV provided background topographies with limited impact of large-scale and irregular height differences. This strategy enabled detailed analysis of investigation-relevant small height differences between zirconia crystals and cement material. Measurements of length were performed using a 100 Mirau interferometer objective with and without a 1.6 magnification lens. According to the analysis of converged roughness parameters [13] the maximum FOVs were 150 lm 150 lm and 93 lm 93 lm, respectively. The virtual pixel sizes of the resulting data were 73 nm 73 nm and 46 nm 46 nm. Surface topography data were analyzed using SPIP software (Image Metrology A/S, Hørsholm, Denmark), version 5.0.5. 3. Results After examination of clinical data and the retrieved cement mantles, we investigated the control specimens and bone cement samples prepared in the laboratory for better interpretation of findings. 3.1. Clinical data Twelve patients with unilateral degenerative disease of the hip and one with a femoral neck fracture had received a total hip replacement. After an average age of 80 years (range 73–87) and average 4 years and 9 months of functional implantation time (range 2.8–7.10), the femoral components were removed. The body mass index at implantation ranged between 20.6 and 37.3; three patients had grade I, and one grade II obesity. Fourteen specimens from the bone cement cuffs originated from Gruen zones 1 and 7, six from Gruen zones 2 and 6, and five from Gruen zones 3 and 5 [14]. The straight conical stem, a close taper angle, and a rectangular cross-section with a small radius of rounded corners of the component presented here (Fig. 1a) had caused fractures of the cement cuff. The frequent displacement of the femoral stem was a characteristic of mediodorsal rotation over a longitudinal axis and a cranial tilt over a tilt point variable in height. The deflection of the femoral stem from cement surfaces resulted in a relatively sharp transition from fairly well-preserved replicas of the shot-peened
Please cite this article in press as: A. Schunck et al., Release of zirconia nanoparticles at the metal stem–bone cement interface in implant loosening of total hip replacements, Acta Biomater. (2015), http://dx.doi.org/10.1016/j.actbio.2015.11.044
4
A. Schunck et al. / Acta Biomaterialia xxx (2015) xxx–xxx
metal surface to abraded and polished surfaces (Fig. 4b) [15]. Soft tissue depositions filled permanent gaps between stem and cement.
undisturbed rough surfaces, a hardness of HR 34 to 43 was found. For comparison, the hardness of zirconia was given as HR 71, and of CoCr alloy as about 60 HR.
3.2. Macroscopy
3.3.2. Pure zirconia: size distribution The particle size distribution (LS-PSA) of original contrast medium powder (ZrO2) of Palacos RÒ bone cement is shown in Fig. 2a and Table 1. The volume share (%) showed a sharp peak in the diameter range of about 20 lm. Thus, big agglomerates dominated the volume distribution. This was the main impression given by light microscopy. Nevertheless, a considerably smaller mean diameter (
