CLINICALLY RELEVANT

Nanomedicine: Nanotechnology, Biology, and Medicine7 (2011) 674–681

Case Study

Characterization of metal-wear nanoparticles in pseudotumor followingmetal-on-metal hip resurfacing

Zhidao Xia, PhDa,b,⁎, Young-Min Kwon, PhD, FRCS(Orth), FRACS(Orth)c,Shahid Mehmoodb, Clive Downing, PhDd,

Kerstin Jurkschat, PhDc, David W. Murray, MD, FRCS(Orth)baCentre for Nanohealth, School of Medicine, Swansea University, Singleton Park, Swansea, United Kingdom

bBotnar Research Centre, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford,Headington, Oxford, United Kingdom

cDepartment of Orthopedic Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USAdDepartment of Materials, University of Oxford, Begbroke Science Park, Oxford, United Kingdom

Received 18 May 2011; accepted 5 August 2011

nanomedjournal.com

Abstract

Biopsies from a typical case of pseudotumor following metal-on-metal hip resurfacing (MoMHR) were analyzed using light andtransmission electron microscopy, backscatter scanning electron microscopy and energy dispersive x-ray spectrometry (EDS). Heavymacrophage infiltration was observed in all black pigmented specimens. Metal nanoparticles (NPs) were observed exclusively withinphagosomes of living macrophages and fragments of dead macrophages. Although dead fibroblasts were found to be juxtaposed withdead and disintegrated macrophages, the NPs were not seen within either live or dead fibroblasts. Chromium (Cr) but not cobalt (Co) wasthe predominant component of the remaining wear NPs in tissue. The current study finding suggests that corrosion of Co in phagosomesof macrophages and resultant Co ion release lead to tissue necrosis and adverse soft tissue reactions (pseudotumors). Further studies arerequired to elucidate the precise mechanism of intracellular corrosion of metal NPs and the long-term toxicity of the Cr remaining in theperi-prosthetic tissues.

From the Clinical Editor: In this study of metal-on-metal hip resurfacing-related tissue necrosis and pseudotumor formation, corrosion anddecomposition of metallic cobalt in phagosomes of macrophages and resultant cobalt ion release were demonstrated to be the key elementsof pathogenesis.© 2011 Elsevier Inc. All rights reserved.

Key words: Metal-on-metal hip resurfacing; Metal wear nanoparticles; Cobalt; Chromium; Macrophages

Contemporary metal-on-metal hip resurfacing (MoMHR) isan alternate option for the treatment of end-stage osteoarthritisin young and active adults.1 The potential advantages of theMoM resurfacing procedure include improved wear charac-teristics, relative conservation of bone stock, restoration ofanatomic hip mechanics and enhanced stability.2,3 The totalnumber of MoMHR procedures accounted for 46% of allprimary hip arthroplasty procedures performed in the patientgroup younger than 55 years of age in the United Kingdom,

This project was partly supported by theNIHRMusculoskeletal BiomedicalResearch Unit and Botnar Research Fellowship. No conflict of interest wasreported by the authors of this article.

⁎Corresponding author: Centre for Nanohealth, School of Medicine,Swansea University, Singleton Park, Swansea, SA2 8PP UK.

E-mail address: z.xia@swansea.ac.uk (Z. Xia).

1549-9634/$ – see front matter © 2011 Elsevier Inc. All rights reserved.doi:10.1016/j.nano.2011.08.002

Please cite this article as: Z. Xia, Y.-M. Kwon, S. Mehmood, C. Downing, K.pseudotumor following metal-on-metal hip resurfacing. Nanomedicine: NBM 2

and one-third of 250,000 hip replacements annually in theUnited States. The majority of patients receiving MoMHRhave well-functioning hips and are thought to be at low riskof developing serious problems. However, concerns havearisen as the devices have been shown to cause an adversereaction to metal debris (described as “pseudotumor”)4 insome patients.

Pseudotumor is a mass that can be cystic or solid and it cancause nerve damage, dislocation and extensive soft tissuedestruction.5-8 Although the prevalence of pseudotumor isthought to be rare, the consequence of soft tissue damage canbe catastrophic,4 as the majority of pseudotumor patients requireadditional revision surgery, and the outcomes of revision in casesof pseudotumor are reported to be poor.9 Risk factors forpseudotumor include female gender, hip dysplasia, age b 40years and use of small components.10 Progressive soft tissue

Jurkschat, D.W. Murray, Characterization of metal-wear nanoparticles in011;7:674-681, doi:10.1016/j.nano.2011.08.002

Figure 1. X-ray radiography showing the hip prosthesis before (A) and after (B) revision surgery.

Figure 2. Light micrographs of toluidine blue-stained thin sections. (A) a group of macrophages with phagocytosed particles. (B) a focal area of necroticmacrophages and nearby fibroblasts. MΦ, macrophages. Fb, fibroblasts. Original magnification, 1,000×.

675Z. Xia et al / Nanomedicine: Nanotechnology, Biology, and Medicine 7 (2011) 674–681

reactions to the wear debris associated with MoM articulationshave been suggested to be the main cause of pseudotumor.11

We have previously reported in vitro assessment of toxicityof Co and Cr nanoparticles (NPs) on macrophages anddemonstrated that these toxic effects are dose dependent.11

Although metal wear NPs in retrieved tissue from failed MoMtotal hip replacements have been characterized using transmis-sion electron microscopy (TEM),12,13 the critical morphologicalevidence demonstrating the presence of these particles in peri-prosthetic tissues in patients with pseudotumor is lacking. Thismay be due to the size of these particles, which is in thenanometer scale, making them difficult to visualize usingroutine histology and analytic tools. The aim of this study wasto perform ultrastructural characterization of wear metal NPsand to obtain morphological evidence of cellular responses inpseudotumor following MoM hip resurfacing. High-resolutionTEM and backscatter scanning electron microscopy in combi-nation with energy-dispersive x-ray spectroscopy (EDS) were

used for analysis of metal wear NPs in biopsies retrieved fromrevision surgery.

Methods

Subject

A 39-year-old female patient with hip dysplasia underwenta left Birmingham DDH hip resurfacing procedure. Thefemoral component head size was 42 mm and the acetabularcomponent orientation was 38 degrees of inclination and 32degrees of anteversion. At 3 years post-operatively, she beganto complain of groin pain. A work-up for infection wasnegative. However, markedly elevated levels of Co-Cr ionswere measured (serum Cr concentration was 174.00 nmol/Land serum Co was 14.70 nmol/L). An exploration and revisionsurgery was performed. During the revision operation,extensive metallosis, copious fluid and marked lysis of the

Figure 3. TEM micrographs showing the presence of metal wear NPs in living macrophages and in tissue around a dead macrophage. Nu, cell nucleus; Col,collagen fibres. (A) a group of macrophages that have phagocytosed large numbers of metal wear particles. All cells show the presence of an intact plasmamembrane enveloping phagosomes containing metal wear particles. (B) high magnification shows the nanoparticles are located mainly within phagosomes. Twotypes of nanoparticles can be seen; larger, high density needle-like particles, and smaller, low-density, smooth-surfaced particles. Notice that metal particle-containing phagosomes show high electron-density matrix staining. (C) High-density NPs. White arrows indicate the intact membrane of phagosomes. (D)Metal wear NPs in tissue around a dead macrophage. Notice the disintegrated plasma membrane and particles scattered in the extracellular matrix, mainlycollagen fibres. (E)Nucleus of a dead macrophage that has lost staining of nucleoli and chromatin. (F) Breakdown of phagosomemembrane. Metal particles areno longer enclosed within phagosomes.

676 Z. Xia et al / Nanomedicine: Nanotechnology, Biology, and Medicine 7 (2011) 674–681

femoral head and neck were observed. Neither component wasloose. Following explant of the MoMHR components, a totalhip replacement with a ceramic-on-ceramic bearing wasreimplanted (Figure 1). The patient reported near-completeresolution of her symptoms post-operatively.

Histopathological examination and TEM

Biopsies were collected from the operating room during therevision operation with institutional research ethical approval andinformed consent from the patient. Hip capsule, synovium,

Figure 4. Electron micrographs of fibroblastic cell death near disintegrated macrophages and released NPs. (A) Four fibroblastic cells undergoing apoptosis. Cell1 and cell 3 show condensed nuclei and shrinkage and condensation of the cytoplasm. Cell 2 shows chromatin condensation with heavily vacuolized cytoplasm;Cell 4 shows condensed cytoplasm and a typical apoptotic nucleus. (B) high-density particles and disintegrated phagosomes released into the collagen matrixfrom a dead macrophage. Notice the disappearance of the plasma membrane.

677Z. Xia et al / Nanomedicine: Nanotechnology, Biology, and Medicine 7 (2011) 674–681

pseudotumor tissue, femoral bone and acetabular membrane werefixed in 10% formaldehyde and processed by routine histopath-ological methods. Biopsies were also fixed immediately in 4%glutaraldehyde in 0.1M phosphate buffered saline (PBS) (pH 7.2–7.4). A series of slices of approximately 1 mm thickness werefashioned using a No. 4 scalpel handle and No. 21 blade. Thesewere observed under a Zeiss stereomicroscope to locate areaswhere therewere darkened patches. These patches were trimmed to1mm×1mm×2mmand then dehydrated through a gradedwater-ethanol series and embedded in Spurr's resin. Thin sections werestained with Toluidine blue and observed by light microscopy(Carl-Zeiss M1, Carl-Zeiss Ltd., Welwyn Garden City, UnitedKingdom). Suitable samples were selected for ultrathin sectioningand contrasting with lead citrate. Micrographs were taken with aJEOL 1200EXII TEM for morphological observation.

TEM in combination with energy dispersive X-ray spectrometer(EDS) analysis

For element analysis, ultrathin sections were analyzed by aJEOL 2010 analytical TEM. This TEM has a LaB6 electron gun(80 – 200 kV) with a resolution of 0.19 nm, an electron probesize down to 0.5 nm and a maximum specimen tilt of ± 10degrees along both axes. It has facilities for point analysis as wellas mapping and line scanning through the SemiStem controller.

Backscatter scanning electron microscopy and EDX analysis

Tissue samples of 3 mm × 5 mm × 1 mm were embeddedin Bright Cryo-M-Bed embedding compound (Bright Instru-

ment Company Ltd., Huntingdon, United Kingdom) at -30°C.The frozen tissue samples were cryosectioned (Reichert-JungCryostat 2800, Cambridge Instruments Ltd., Cambridge,United Kingdom) at a thickness of 50 μm and mounted on5 mm × 5 mm pieces of single crystal silicon (Agar ScientificLtd., Stansted, United Kingdom). A JEOL 6480 SEM wasused for the analysis. To remove the need for sample coating,this was operated in a low vacuum mode that also helped todissipate heat and reduce sample damage. The microscope wasused at 15 KV. Backscattered imaging mode was used. AnOxford Instruments Inca x-ray analysis system (OxfordInstruments, Abingdon, United Kingdom) was used for theEDX analysis.

Results

Within the samples of retrieved tissue, areas of blackpigmentation were seen in gross specimens from synovium,pseudotumor, femoral bone and acetabular membrane. Heavymacrophage infiltration was observed in all specimens withpigmentation. Routine histopathological examination showedthat macrophages and scattered giant cells contained metallicwear particles. Focal areas of necrotic tissue and degeneratecollagen were also present in which there were necroticmacrophages containing wear particles (Figure 2). A fewscattered lymphocytes were observed. There was no markedpolymorph or plasma cell response. In femoral bone tissue there

Figure 5. TEM EDX analysis of metal nanoparticles within an intact phagosome. (A) a macrophagic phagosome with intact membrane containing many nano-sized particles. (B) analysis of the EDX spectrum revealed a Cr signal (highlighted in circle) of the 50 nm needle-like metal particles shown in (C). The Cu signalis from the copper grid, and U is from staining. (C) high magnification of the NPs in A.

678 Z. Xia et al / Nanomedicine: Nanotechnology, Biology, and Medicine 7 (2011) 674–681

were areas of necrotic and degenerative changes within the bonemarrow that also contained numerous macrophages containingwear particles.

TEM observation confirmed that there were both living anddead macrophages in the tissue (Figure 3). Living macrophageswere packed with high electron density metal particles (Figure 3,A). These cells had intact plasma membranes and nuclearenvelopes, with normal nuclear staining (centred nucleoli andtypical nuclear membrane depressions). The cytoplasm was fullof phagosomes containing nanometer-sized metal particleswithin intact membranes (Figures 3, B and C). The NPs wereof two types; either high-density needle-like particles, or low-density, smooth-surfaced particles.

The plasma membrane of dead macrophages had disinte-grated (Figure 3, D) and released the metal NPs into theextracellular matrix. In the nuclei, the normal staining of nucleoliand chromatin were lost (Figure 3, E). As a consequence ofphagosome membrane breakdown, metal particles were nolonger enclosed within lysosomes (Figure 3, F).

No metal NPs were observed within fibroblasts. Deadfibroblasts were mainly seen in close proximity to deadmacrophages (Figure 4). The dead fibroblasts had typicalapoptotic morphology, including nuclear and chromatin con-densation, shrinkage and condensation of the cytoplasm orvacuolization of the cytoplasm.

Under TEM with EDX analysis, the tissue was seen to have anumber of areas containing mainly small (20 – 200 nm), longishparticles, which contained Cr. No Co could be detected. Manymetal NPs were observed within an intact phagosome with intactmembrane (Figure 5, A). The EDX analysis spectrum revealed aCr signal (Figure 5, B). There was also a small amount (b 1%) ofuranium (U) and osmium (Os) contamination. Different parts ofthe tissue also contained phosphorus, carbon or oxygen. In areaswithout metal particles, neither Cr nor Co could be detectedusing EDX.

To confirm the composition of the metal particles,backscatter SEM in combination with EDX was used toanalyze tissue sections. After extensive searching, clusters ofparticles were observed (Figure 6, A). These particles appearedto be buried in the tissue because they did not appear bright onthe backscattered images. EDX analysis was performed onareas where particles were observed and mapped within a largerarea (Figure 6, B, C). Cr was detected in these areas (Figure 6spectrum 1, 2 and Figure 6, B), but no Co was found (Figure 6spectrum 1, 2 and Figure 6, C). Neither Cr nor Co was detectedin areas where no particles were observed (Figure 6 spectrum 3and 4).

Discussion

This report describes a typical clinical case of adverse softtissue reaction to metal device wear or “pseudotumor” in apatient with MoMHR who had multiple known risk factors.These risk factors include female gender, hip dysplasia, b 40years old at the time of hip resurfacing and use of a smallfemoral head (42 mm) component of the prosthesis. Metal ionconcentrations were found to be elevated in the blood.Diagnosis of pseudotumor was confirmed following patholog-ical examination. Following the revision surgery in whichmetal-on-metal articulation was revised with a ceramic-on-ceramic bearing, the patient reported near complete resolution ofher symptom at one-year follow-up.

We recently reported that pseudotumors were associated withsignificantly higher Co and Cr levels in blood and hip aspirates incomparison with patients who underwent MoMHR but did notdevelop pseudotumor.14 However, our previous in vitro studyindicated that Co ion concentrations present in serum and hipaspirates may not be sufficient to cause widespread cytotoxicity,as the toxic dose of Co ions in vitro is 2 orders of magnitudehigher than the concentration found in hip aspirates.11 Therefore,

Figure 6. EDX analysis by backscatter scanning electron microscopy. (A) notice the high electronic density of particles on the surface of the section. EDXanalysis was performed at 4 individual points (white arrows). Spectra 1 and 2 were from the high-density particles; both showed the presence of Cr, but nodetectable Co. Spectra 3 and 4 were from normal density tissue, where neither Cr nor Co was detected. (B) mapping of Cr distribution in the selected area in A.High-density dots show Cr signals. (C) mapping of Co distribution. Co signal is undetectable.

679Z. Xia et al / Nanomedicine: Nanotechnology, Biology, and Medicine 7 (2011) 674–681

necrosis and soft tissue destruction are seen in focal areas in vivo,which is likely to be mediated predominantly by Co/Cr NPs,whereby phagocytosed NPs lead to direct effects on cells viaintracellular corrosion.

Commercially available MoM prostheses contain 62% Co and28% Cr, a Co/Cr ratio of 2.21.15 However, in this study, Co wasnot detectable in themetal debris using EDS element analysis. Thiswas an unexpected finding because Co was detected in the serumof this patient. However, this finding would be consistent withDoorn's study, in which metal wear particles were observed byTEM in 8 of 12 patients who underwent revision following

MoMHR. Although cellular structure was lacking in Doorn'sreport, the sizes and shapes of the particles observed were verysimilar to our findings, and the composition of the particles byEDSanalysis also demonstrated that the metal wear particles in tissuehad relatively higher Cr and oxygen (O) peakswith no or lowCo.12

One possibility is that Co is more soluble than Cr; therefore,Co corroded faster and leached from the local environment.Although Co is almost insoluble in water, its solubility increases500-fold in serum.16 Haynes et al reported that by incubatingCo/Cr alloy metal particles of between 500 nm and 3 μm indiameter at pH 4 (the same pH level as that found within

680 Z. Xia et al / Nanomedicine: Nanotechnology, Biology, and Medicine 7 (2011) 674–681

lysosomes in macrophages) for 48 hours in culture medium at37°C, 30-fold more Co ions than Cr ions were released.17 Thisprovides strong evidence that Co is more soluble than Cr underphysiological conditions.18

Macrophages are believed to be the dominant cells that respondto foreign materials in small particle form via phagocytosis.19 It isnotable that in this case metal wear NPs were exclusively collectedby macrophages and stored in phagosomes, cell organellesdesigned for ingestion of engulfed foreign bodies.

A study of Co and Cr ions in macrophages by Catelas et alshowed that both Co and Cr are toxic to macrophages at highconcentrations. After 24 hours in contact with these ions,macrophages underwent apoptosis, but at 48 hours cell death wasdominated by necrosis.13 Co2+ was found to be more toxic thanCr3+, with 15% cell death observed at 8 ppm, whereas 250 ppmof Cr3+ was required to cause 13% cell death at 24 hours.13 Co6+

is also believed to be more toxic than Cr3+.Although Co ions are more toxic than Cr, the toxic dose of Co

ions in vitro is higher than the in vivo concentration of Co ions inthe blood and synovial fluid in MoMHR patients withpseudotumors.11 Therefore it is unlikely that Co ions at theconcentration levels found in tissue cause direct toxicity. In fact,tissue necroses are seen in focal areas, rather than widespreadthroughout the tissue.

In light of these findings, it is important to understand how theions are released into tissue. A plausible mechanism of metalcorrosion would need to take into account the acidic environmentin which themetal particle is found after phagocytosis. Haynes et alreported that Co/Cr alloy wear particles had cytotoxic effects onmouse peritoneal macrophages. By using bafilomycin A, aninhibitor of H+-ATPase, to inhibit acidification of macrophagelysosomes, toxicity was significantly reduced.17 When the Co/Crwear particles were incubated at a low pH similar to that foundwithin the phagosomes of macrophages, soluble products inducedthe same toxic effects as the wear particles.17 This suggests that it isthe soluble products released from wear-particle corrosion, ratherthan the particles themselves, that lead to cytotoxicity.

Co has been shown to be transported from tissue to the bloodand eliminated in urine; however, Cr is not rapidly eliminated butaccumulates in tissues and remains within cells.20 This findingprovides a possible explanation for our results as we found that,using SEM and TEM in combination with EDX analysis, Co wasnot detectable in phagocytosed metal NPs.

Utilizing ultrastructural analysis of the wear particles at ananometer scale, we were able to observe that as a phagocyte, themacrophage engulfs metal NPs and stores them in lysosomes. Itis well known that macrophages can produce an acidicenvironment in lysosomes and digest phagocytosed materialsand minerals. As Co is more soluble than Cr, Co is likely to bedissolved more rapidly within lysosomes. It is still unknown howCo ions are transported from phagosomes to the cytoplasm andextracellular spaces. However, it is logical to postulate that asphagosomes or lysosomes possess the most suitable localenvironment for Co corrosion, there would be a gradient of Coions from the phagosomes to the extracellular spaces.

As large numbers of NPs were phagocytosed by themacrophages in this study, this suggests that Co ions releasedfrom phagosomes can reach a toxic concentration level sufficient

to kill macrophages. Dead macrophages were seen to beoverloaded with metal particles, and showed disintegration ofthe lysosomes (or phagosomes) and the plasma membrane. Toxiclevels of Co released from dead macrophages may form a Co“ion-wave” that also kills nearby fibroblasts. Metal wear NPswere not seen within either live or dead fibroblasts. However, thedead fibroblasts were always observed to be in close proximity tothe dead and disintegrated macrophages, without morphologicalevidence of engulfment of or contact with metal particles.

High concentrations of Co in “ion-waves” would be dilutedand reach equilibrium with extracellular body fluid, andsubsequently transported from the tissue to the blood.20 Thiswould be consistent with the elevated Co concentrationmeasured in the study patient's blood. Furthermore, macro-phage death may not all occur at the same time or at the samelocation; therefore, it should not cause a dramatic and long-termelevation of Co in local tissue. Because Cr is less soluble and istransported from tissue less easily than Co is, Cr is more likelyto remain in the cells after degradation20 and thus can bedetected by various techniques.

The current study is the first to morphologically characterizemetal wear NPs in pseudotumor tissue, in particular, the presenceof metal NPs in phagosomes of macrophages. Although thisstudy is limited to the findings of one typical pseudotumor case,the relative uniformity of histological findings and morphologyof pseudotumor tissues, which are characterized by necrosis inthe presence of dead macrophages, has been previously reportedin multiple studies.5,8,11,14

Our future research direction is to pursue a larger scale,multicenter study, to provide quantitative results of the NPcharacterization reported in this study, and to ascertain theunderlying cellular and molecular mechanism. Apart fromdeveloping techniques to reduce the wear of prostheses, under-standing the mechanism of pseudotumor formation will pave theway to the development of therapies that will improve patient care.

In conclusion, the current study provides the first ultrastruc-tural morphological evidence of cellular response to metal wearNPs in pseudotumor tissue following MoMHR. The findingssuggest that Co/Cr wear NPs originating from excessive wear ofMoMHR are taken up mainly by macrophages, within whichthey corrode in the acidic environment of the phagosomes,causing death of individual NP-overloaded macrophages andnearby fibroblasts (Supplementary Figure 1). Further studies arerequired to elucidate the precise mechanism of intracellularcorrosion of metal NPs and the long-term toxicity of the Crremaining in the peri-prosthetic tissues.

Appendix A. Supplementary data

Supplementary materials related to this article can be foundonline at doi:10.1016/j.nano.2011.08.002.

References

1. Amstutz HC, Beaule PE, Dorey FJ, Le Duff MJ, Campbell PA, GruenTA. Metal-on-metal hybrid surface arthroplasty: two to six-year follow-up study. J Bone Joint Surg Am 2004;86-A:28-39.

681Z. Xia et al / Nanomedicine: Nanotechnology, Biology, and Medicine 7 (2011) 674–681

2. Grigoris P, Roberts P, Panousis K, Bosch H. The evolution ofhip resurfacing arthroplasty. Orthop Clin North Am 2005;36:125-34, vii.

3. Mabilleau G, Kwon YM, Pandit H, Murray DW, Sabokbar A. Metal-on-metal hip resurfacing arthroplasty: a review of periprosthetic biologicalreactions. Acta Orthop 2008;79:734-47.

4. Langton DJ, Jameson SS, Joyce TJ, Hallab NJ, Natu S, Nargol AV. Earlyfailure of metal-on-metal bearings in hip resurfacing and large-diametertotal hip replacement: A consequence of excess wear. J Bone Joint SurgBr 2010;92:38-46.

5. Pandit H, Glyn-Jones S, McLardy-Smith P, Gundle R, Whitwell D,Gibbons CL, et al. Pseudotumours associated with metal-on-metal hipresurfacings. J Bone Joint Surg Br 2008;90:847-51.

6. Harvie P, Giele H, Fang C, Ansorge O, Ostlere S, Gibbons M, et al. Thetreatment of femoral neuropathy due to pseudotumour caused by metal-on-metal resurfacing arthroplasty. Hip Int 2008;18:313-20.

7. Clayton RA, Beggs I, Salter DM, Grant MH, Patton JT, Porter DE.Inflammatory pseudotumor associated with femoral nerve palsyfollowing metal-on-metal resurfacing of the hip. A case report. J BoneJoint Surg Am 2008;90:1988-93.

8. Mahendra G, Pandit H, Kliskey K, Murray D, Gill HS, Athanasou N.Necrotic and inflammatory changes in metal-on-metal resurfacing hiparthroplasties. Acta Orthop 2009;80:653-9.

9. Grammatopolous G, Pandit H, Kwon YM, Gundle R, McLardy-Smith P, Beard DJ, et al. Hip resurfacings revised for inflammatorypseudotumour have a poor outcome. J Bone Joint Surg Br2009;91:1019-24.

10. Glyn-Jones S, Pandit H, Kwon YM, Doll H, Gill HS, Murray DW. Riskfactors for inflammatory pseudotumour formation following hipresurfacing. J Bone Joint Surg Br 2009;91:1566-74.

11. Kwon YM, Xia Z, Glyn-Jones S, Beard D, Gill HS, Murray DW. Dose-dependent cytotoxicity of clinically relevant cobalt nanoparticles andions on macrophages in vitro. Biomed Mater 2009;4:025018.

12. Doorn PF, Campbell PA, Worrall J, Benya PD, McKellop HA, AmstutzHC. Metal wear particle characterization from metal on metal total hipreplacements: transmission electron microscopy study of periprosthetictissues and isolated particles. J Biomed Mater Res 1998;42:103-11.

13. Catelas I, Campbell PA, Bobyn JD, Medley JB, Huk OL. Wear particlesfrom metal-on-metal total hip replacements: effects of implant designand implantation time. Proc Inst Mech Eng H 2006;220:195-208.

14. Kwon YM, Ostlere SJ, McLardy-Smith P, Athanasou NA, Gill HS,Murray DW. qAsymptomaticq Pseudotumors After Metal-on-Metal HipResurfacing Arthroplasty Prevalence and Metal Ion Study. J Arthro-plasty 2011;26:511-8.

15. Saikko V, Nevalainen J, Revitzer H, Ylinen P. Metal release from totalhip articulations in vitro: substantial from CoCr/CoCr, negligible fromCoCr/PE and alumina/PE. Acta Orthop Scand 1998;69:449-54.

16. Harding HE. Notes on the toxicology of cobalt metal. Br J Ind Med1950:76-8.

17. Haynes DR, Rogers SD, Howie DW, Pearcy MJ, Vernon-Roberts B.Drug inhibition of the macrophage response to metal wear particles invitro. Clin Orthop Relat Res 1996:316-26.

18. Heath JC, Freeman MA, Swanson SA. Carcinogenic properties of wearparticles from prostheses made in cobalt-chromium alloy. Lancet1971;1:564-6.

19. Xia Z, Triffitt JT. A review on macrophage responses to biomaterials.Biomed Mater 2006;1:R1-9.

20. Merritt K, Brown SA. Distribution of cobalt chromium wear andcorrosion products and biologic reactions. Clin Orthop Relat Res 1996(329 Suppl):S233-43.