Contact Dermatitis • Original Article CODContact Dermatitis

Failure of total hip implants: metals and metal release in 52 cases

Stig S. Jakobsen1, Carola Liden2, Kjeld Søballe1, Jeanne D. Johansen3, Torkil Menne4,Lennart Lundgren2,5, David Bregnbak3, Per Møller6, Morten S. Jellesen6 and Jacob P. Thyssen3

1Department of Orthopaedic Surgery, Aarhus University Hospital, DK-8000 Aarhus, Denmark, 2 Institute of Environmental Medicine, Karolinska Institutet,171 77 Stockholm, Sweden, 3Department of Dermato-Allergology, National Allergy Research Centre, Gentofte University Hospital, DK-2900 Hellerup,Denmark, 4Department of Dermato-Allergology, Gentofte University Hospital, DK-2900 Hellerup, Denmark, 5Department of Applied Environmental Science,Stockholm University, SE-10691 Stockholm, Sweden, and 6Department of Mechanical Engineering, Technical University of Denmark, DK-2800 KongensLyngby, Denmark

doi:10.1111/cod.12275

Summary Background. The pathogenesis of total joint replacement failure is multifactorial. Onehypothesis suggests that corrosion and wear of alloys result in metal ion release, whichmay then cause sensitization and even implant failure, owing to the acquired immunereactivity.Objectives. To assess cobalt, nickel and chromium(VI) release from, and the metalcomposition of, failed metal-on-ethylene total hip replacements.Materials/methods. Implant components from 52 revision cases were evaluated withspot tests for free nickel, cobalt, and chromium (VI) ions. Implant composition wasdetermined with X-ray fluorescence spectroscopy, and information on the reason forrevision and complications in relation to surgery was collected from the medical chartswhen possible (72%). For 10 implants, corrosion was further characterized with scanningelectron microscopy.Results. We detected cobalt release from three of 38 removed femoral heads and fromone of 24 femoral stems. Nickel release was detected from one of 24 femoral stems. Nochromium(VI) release was detected.Conclusions. We found that cobalt and nickel were released from some failed total hiparthroplasties, and corrosion was frequently observed. Metal ions and particles corrodedfrom metal-on-polyethylene may play a role in the complex aetiopathology of implantfailure.

Key words: chromium; cobalt; metal allergy; nickel; spot test; total joint replacement.

Total joint replacement surgery has traditionally beenreserved for elderly patients with advanced arthrosis and

Correspondence: Stig S. Jakobsen, Department of Orthopaedic Surgery,Aarhus University Hospital, DK-8000 Aarhus, Denmark. Tel: +45 78467471; Fax: +45 7846 5271. E-mail: Stig.Jakobsen@ki.au.dk

Disclosures: Jacob P. Thyssen is a Lundbeck Foundation fellow and issupported financially by an unrestricted grant.Conflict of interests: Jacob P. Thyssen and Morten S. Jellesen sold the cobaltspot test to SmartHealth (Phoenix, AZ, USA) after its development, and willreceive annual royalty fees for net sales of the spot test. The other authorshave no conflict of interest to disclose.

Accepted for publication 20 May 2014

who postoperatively would be less active. However, thisscenario has substantially changed, owing to increasedlife-expectancy. Pertinently, many patients now receivetotal hip arthroplasty at a young age. The clinical outcomeis generally excellent, but many young patients still needimplant replacement within 10–15 years (1, 2), and somemay experience complications, including implant failure.

The pathogenesis of implant failure is multifactorial,and several hypotheses have been raised (3). Recently,excessive metal ion release from total hip arthroplas-ties has been suggested and investigated (4, 5). Thus,high levels of metal ions released from second-generationmetal-on-metal bearings have been associated with

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons LtdContact Dermatitis, 71, 319–325 319

FAILURE OF TOTAL HIP IMPLANTS • JAKOBSEN ET AL.

clinically serious complications, including aseptic lym-phocytic vasculitis-associated lesions and pseudotumourformation (6). Histopathological examination has shownunusual lymphocytic aggregates in the periprosthetic tis-sue of patients with high systemic metal levels, whichdiffers from the tissue reactions observed in patients withlow metal levels (7–9). Moreover, traditional metal-on-polyethylene total hip arthroplasties may release cobalt,chromium, and nickel, which putatively may cause a typeIV delayed hypersensitivity reaction (10) and perhapseven induce pseudotumour formation (11).

It is currently unclear whether metal allergy beforeinsertion of a metal implant increases the risk ofpostoperative implant failure. Corrosion and wear ofalloys result in metal ion and particle release, whichmay then cause sensitization and even implant failure,owing to the acquired immune reactivity. It is interestingto evaluate the release of metal ions from failed total hiparthroplasties, as this may be indicative of a possible role ofmetal ions in the aetiopathogenesis of implant failure. Theaim of this study was to assess the metal composition of,and release of cobalt, nickel and chromium(VI) from,consecutively collected implants from failed total hiparthroplasties removed during revision surgery.

Methods/Materials

At Aarhus University Hospital in Denmark, all totalhip athroplasties removed from patients during revisionsurgery performed between 2008 and 2009 wereautoclaved and stored for later analysis. Information onthe reason for revision and complications in relationto surgery was collected from the medical records. Asimplant components from 15 patients were stored withouta personal identification number, the medical historycould not be retrieved in these cases. All joints weremetal-on-polyethylene, except for four hemi-alloplastics,which, by definition, were ‘metal-on-cartilage’. All non-metallic parts were excluded from the study, as theywere not believed to be a likely source of metalrelease.

Free cobalt, nickel and chromium(VI) ions on theimplant surfaces were assessed with, respectively, thecobalt spot test based on disodium-1-nitroso-2-naphthol-3,6-disulfonate (12), the nickel spot test based ondimethylglyoxime (13), and a chromium(VI) spot testbased on diphenylcarbazide (14). These screening testshave not been developed or evaluated for this purpose,but, typically, a visible colour change (indicating a positivetest result) occurs when there is approximately 10 ppm(μg/ml) nickel, 8 ppm cobalt and 0.5 ppm chromium(VI)in a solution of the respective reagent.

Fig. 1. X-ray of a total hip arthroplasty with markings on the testsites (A, B, C, D, E, and F) for assessment by X-ray fluorescencespectrometry and spot tests for the release of nickel, cobalt, andchromium(VI). There are specially designed rough ingrowth areason the implants (coated areas).

The implant composition was then determined withX-ray fluorescence (XRF) spectroscopy. The compositionsof the outer layers of all implants (n = 52), in differentparts (n = 152) of the separate components (n = 88),were determined with a handheld XRF analyzer (Innov-X, Alpha 4000; Innov-X Europe, Hertogenbosch, TheNetherlands). Areas for assessment were indicated inphotographs of each item by one of the authors (S.J.,orthopaedic surgeon): (A) head-and-neck junction headside, (B) weight-bearing area of the femoral head, (C)coating on the acetabular shell, (D) coating on the femoralstem, (E) the femoral neck, (F) head-and-neck junctionfemoral side, and (G) screw holes on the acetabular

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd320 Contact Dermatitis, 71, 319–325

FAILURE OF TOTAL HIP IMPLANTS • JAKOBSEN ET AL.

Table 1. Compositions of 52 failed hip implants. Materials are based on X-ray fluorescence spectrometry, patient files, and visual identification.Numbers of components with release of nickela, cobaltb or chromium(VI)c, as shown by spot tests, are shown. Implants were often composedof more than one material. Details are shown in Table S1

Material category(no. of implants)

Examples of alloyname (typicalcomposition)

Femoral head(A and B in Fig. 1)

(n = 38)

Acetabular cup(C and G in Fig. 1)

(n = 19)

Femoral stem(D, E and F in Fig. 1)

(n = 25)

Femoral head fromhemiarthroplasties(H in Fig. 1) (n = 6)

CoCrMo alloy (n = 23) F1537 (Co/Cr/Mo/Mn) 19 (3 Co-positive)b 0 3 (1 Co-positive)b 4Stainless steel (n = 23) 316L (Fe/Cr/Ni/Mo/Mn) 1 1 2 (1 Ni-positive)a 2

F1586 (Fe/Cr/Ni/Mn/Mo) 17 1 14 0Titanium alloy (n = 21) Ti 6–4 (Ti90/Al6/V4) 1 15 6 0Titanium (n = 3) cpTi (Ti100) 0 3 0 0Tantalum–tungsten

alloy (n = 1)(Ta/W) 0 1 0 0

aNi spot test: implant no. 25 was posistive on the femoral stem.bCo spot test: implants no. 12, 41 and 44 were positive on femoral heads, and implant no. 35 was positive on the femoral stem.cCr(VI) spot test: no positives.

shell (Fig. 1 and Table S1). The femoral neck (E) waschosen additionally because this area commonly doesnot have a coating and represents the bulk material ofthe implant. X-ray analysis was conducted for 300 s.Elements with a smaller atomic mass than that oftitanium, such as aluminium, were not detected by theXRF instrument used. Only elements with an estimatedelemental content of > 0.5% are reported, but content of< 0.5% was also assessed (Table S1). It should be notedthat the composition determined by XRF is indicative andsemiquantitative.

We then collected 10 implants for further surfacecomposition analyses with scanning electron microscopy(SEM) (JEOL 5900, JEOL Ltd, Tokyo, Japan) in combina-tion with energy dispersive spectroscopy (EDS) (OxfordLink ISIS analyzer; Oxford Instruments, Oxford, UK).

Results

Metal content and alloys

A summary of the compositions of the 52 removed hipimplants is shown in Table 1; details for all implantsare shown in Table S1. We found that stainless steelwas the most frequently used material, being present in23 (44%) of the 52 implants, and in 38 (43%) of the88 implant components, generally in femoral heads andfemoral stems. The XRF results indicated that the stainlesssteel alloy F1586 was more frequent than 316L. CoCrMoalloy was present in 23 (44%) of the implants, and in 26(36%) of the implant components, generally in femoralheads. Titanium alloys were found in 21 (40%) of the 52implants, and in 22 (25%) of the 88 implant components,most often in acetabular cups, but also in femoral headsand stems. Pure titanium metal was detected in single hipimplants.

Fig. 2. Nickel spot test-positive distal femoral stem tip of stainlesssteel with clear signs of material degradation and local corrosion.Implant no. 29.

Metal release determined with spot tests

We found that three (8%) of 38 removed femoral headsand one (4%) of 25 femoral stems released cobalt,all from CoCrMo alloys. Furthermore, one (4%) of 25femoral stems released nickel, the culprit being a 316Lalloy (Table 1) (Fig. 2). We did not find any positivechromium(VI) spot test reactions.

We were not able to establish any correlations betweenimplant failure and positive cobalt and/or nickel spot testresults. A positive cobalt spot test result was found on afemoral head in a patient revised for polyethylene wear.The reason for revision was unknown for the rest of thepositive spot test results.

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons LtdContact Dermatitis, 71, 319–325 321

FAILURE OF TOTAL HIP IMPLANTS • JAKOBSEN ET AL.

12

Fig. 3. Scanning electron microscopy/energy dispersivespectroscopy (EDS) analysis of a stainless steel distal femoral stemtip (close-up of Fig. 2). The surface topography appears degraded,and EDS analysis of a local corrosion spot (spectrum 2) indicatesthat iron and nickel have been locally released as compared withthe reference site (spectrum 1).

Spectrum O Si P Ca Cr Mn Fe Ni Mo

1 (wt%) – 0.6 – – 17.6 1.5 63.2 13.4 3.42 (wt%) 19.9 – 1.7 1.3 54.5 – 4.8 3.3 14.2

Fig. 4. Specimen from the concave part of a stainless steel femoralstem, showing deposits and metal degradation. Implant no. 51.

Surface characteristics

The stainless steel femoral stem resulting in a positivenickel spot test result (no. 25) was analysed with opticalmicroscopy and SEM/EDS, which clearly showed signsof material degradation (Figs. 2–5). Sixteen of the 25investigated femoral stems were made of stainless steel.Even naked eye inspection clearly showed deposits onthe concave side of the stem for the majority of these(Fig. 4).

Fig. 5. Scanning electron micrograph (close-up of Fig. 4) showingthe degraded region of a stainless steel femoral stem with anunaffected region (top left corner) and a material-degraded region.

Table 2. Reason for primary surgery (n = 52) (top row) andexpected distribution according to the Danish Hip ArthroplastyRegistry (bottom row) (15)

Primaryarthrosis Unknown Fracture

Congenitaldislocated

hip

Asepticbone

necrosis

22 (42%) 17 (33%) 11 (21%) 1 (2%) 1 (2%)78% – 12% 1% 3%

Reasons for primary and revision surgery

We were not able to determine the reason for primarysurgery in 17 (33%) patients or the reason for revisionsurgery in 20 (40%) patients. We found comparablereasons for primary and revision surgery in the DanishHip Register (Tables 2 and 3). The amount of material wastoo small to allow any conclusions to be drawn concerningstatistically significant differences between the materialsand the reasons for revision (Table 4).

Discussion

This study utilized cobalt, nickel and chromium spottests to screen for metal ion release from the surfacesof failed total hip arthoplasties retrieved over a 1-yearcollection period. We found that cobalt was releasedfrom three removed femoral heads, and nickel from oneitem, but that no chromium(VI) was released at all.Interestingly, chromium is present in both the stainlesssteel and CoCrMo alloys that were frequently retrieved.It is known that CoCrMo alloy implants corrode andrelease both chromium and cobalt ions (17, 18). Finally,

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd322 Contact Dermatitis, 71, 319–325

FAILURE OF TOTAL HIP IMPLANTS • JAKOBSEN ET AL.

Table 3. Reason for revision surgery (n = 52) (top row) and expected distribution according the Danish Hip Arthroplasty Registry(bottom row) (16)

Aseptic looseningRepeatedluxation Pain Infection

Polyethylenewear

Loosening causedby to metastasis

Peri-acetabularossification Unknown

12 (23%) 8 (15%) 5 (9%) 3 (6%) 2 (4%) 1 (2%) 1 (2%) 20 (40%)55% 17% 3% 8% 1% 0–2% 0–2% –

Table 4. Reasonfor revision surgery and compositions of 52 failed hip implants. Implants were often composed of several components and ofmore than one material. Details are shown in Table S1

Reason for revision; n = number of patients

Material category (n = ni/nc/np)aAseptic

loosening (n = 12)Repeated

luxation (n = 8)Pain

(n = 5)Other reasonb

(n = 7)Unknown(n = 20)

CoCrMo alloy (n = 23/26/44) 3 5 3 4 (1 Co-positive) 8 (3 Co-positive)Stainless steel 316L (n = 5/6/7) 1 0 1 0 3 (1 Ni-positive)Stainless steel F1586 (n = 19/32/58) 7 2 4 2 4Titanium (n = 3/3/3) 1 0 0 0 2Titanium alloy (n = 21/22/39) 4 3 1 3 10Tantalum–tungsten alloy (n = 1/1/1) 0 0 0 0 1

ani/nc/np = number of implants/number of components/number of tested parts.bInfection (n = 3), polyethylene wear (n = 2), loosening caused by metastasis (n = 1), and peri-acetabular ossification (n = 1).

we found that the reasons for primary and revisionsurgery were comparable with registered informationfrom the nationwide Danish Hip Arthroplasty Register(16) (Tables 2 and 3), but acknowledge that a majorproportion of these were categorized as ‘unknown’.

The spot tests have not been developed or evaluatedfor the purpose of screening implants for metal release.In dermatology, the nickel spot test is primarily used toscreen for metal release from occupational and consumermetallic items, as a preventive measure to protect againstallergic contact dermatitis. The cobalt spot test hasrecently been introduced for the same purpose. A positivereaction to these tests indicates that metal ions are releasedin sufficient amounts to elicit allergic skin reactions inalready metal-sensitized individuals. Diphenylcarbazidehas long been used for quantitative determination ofchromium(VI) in products, among them cement andleather products. We note that the more simple spot testfor chromium(VI) still is under evaluation, and that it hasnot yet been introduced for use by dermatitis patients.The International Organization for Standardization andthe American Society for Testing and Materials havepublished several standards concerning medical devices,some of them concerning the degradation of metals andalloys (19) and metallic implants (20).

Released metal ions normally reach the immunesystem only following transepidermal penetration, andamino acids and proteins in a competent barriermay prevent penetration, owing to chelation. In theperiprosthetic tissue, it is possible, and even likely, that

lower metal concentrations may cause sensitization andinflammatory reactions. Deposits of metals are also seenin tissues around implants, suggesting accumulation ofsensitizing metals. Thus, the relatively high proportion ofcobalt spot test-positive implants (four of 23 CoCrMoimplants) could possibly translate into a higher riskof causing sensitization in the peri-implant tissue andeven play a role in implant failure (21–24). Hence, it isconsidered likely that metal release below the detectionlimit of the spot tests could also be of clinical concern.

Prior to spot testing, implants were autoclaved indemineralized water at 121◦C, which is expected tocause some level of repassivation of the surfaces unlesssevere material degradation has occurred. Most implantswith positive cobalt spot test results were femoralheads (three of 19 CoCrMo femoral heads). The reasonfor some CoCrMo alloys resulting in negative cobaltspot test results could be related to variations in theformation of a stable surface oxide layer of the testedCoCrMo alloys. Depending on the implant lifetime andits local biochemical environment during use, somealloys could have developed a more homogeneouschromium-enriched oxide layer, thus minimizing cobaltand chromium ion release when they were spot tested. Apositive nickel spot test result was seen with a femoral stemmade of SAE standard 316L stainless steel. Nickel releasefrom stainless steel is not expected if the alloy is allowedto form and maintain its passive oxide film. Corrosioncombined with fretting (small-amplitude movements) isexpected to be the main cause of disruption of the passive

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons LtdContact Dermatitis, 71, 319–325 323

FAILURE OF TOTAL HIP IMPLANTS • JAKOBSEN ET AL.

film on stainless steels at the concave side of the femoralstems.

Titanium alloys are currently utilized in 70–80% ofall primary total hip replacements in Denmark (16). Theway in which they are used varies between countries, butthere is growing evidence to support the idea that youngerpatients should receive uncemented components, forexample a titanium alloy, meaning that the implant doesnot have to be embedded in the bone with polymethylmethacrylate (25). Titanium is very biocompatible, andrarely induces hypersensitivity reactions, but evidencesuggests that TiO2 and titanium ions form haptensthat can, indeed, induce hypersensitivity and create aproinflammatory cellular response (26). As expected, wedid not find any release of cobalt, chromium or nickelfrom the titanium alloys.

Metal-on-metal total hip arthroplasties often releasecobalt together with chromium (27). This is mainlysuspected to be in the form of chromium(III), Cr2O3,or Cr(OH)3, and not chromium(VI) (28). However,the chromium spot test only detects chromium(VI),and not chromium(III). Thus, chromium release wasnot sufficiently evaluated in this study. In a recentexperimental study, mice received intra-articularinjections of cobalt chromium particles separated intonanosize and micrometre-size; nanosized particles arethe most prevalent in tissue surounding metal-on-metaltotal hip arthroplasties (29) Only mice injected withmicrometre-sized particles were sensitized to cobaltand chromium (29). Micrometre-sized particles werephagocytized, presented on the major histocompatibilitycomplex, and carried away from the site of release.Micrometre-sized and nanosized wear particles arereleased from orthopaedic implants, and from debrisresulting from galvanic and fretting corrosion. Cobalt[Co(II)], titanium [Ti(IV)], aluminium [Al(III)], iron[Fe(III)], nickel [Ni(II)] and chromium [Cr(III)] can bedetected predominantly as metal oxides (Cr2O3, CoO,TiO2, Al2O3, etc.) and hydroxides [Cr(OH)3 and Co(OH)2],all with the potential to cause adverse events (28).

Up to 17% of women and 3% of men in the generalpopulation are metal-allergic, mainly to nickel but alsoto cobalt and chromium (30). In patients with totalhip arthroplasty, the prevalence of patch test-verified orlymphocyte transformation assay-verified metal allergy to

one of these metals reaches 25%, and it even rises to 60%in patients with poorly functioning total hip arthroplasties(21, 31). This may be a non-causative association, andthe increased prevalence of metal allergy can be explainedby excessive secondary metal release from an alreadyfailing implant, as micromotion of implant componentslead to metal release. In contrast, implants not subjectedto any mechanical stress form very stable oxide surfacelayers (chromium and titanium), which act as kineticbarriers, thereby reducing the release of metal ions(32). If an implant experiences mechanical wear (e.g.micromotion), disrupting the thin oxide surface layer(‘fretting corrosion’), metal ion release will increase untilthe oxide film is restored (33).

Despite the relatively small study size, we found thatcobalt was released from failed metal-on-polyethylenetotal hip arthroplasties made of CoCrMo alloy, andthat corrosion was frequently observed in stainlesssteel and CoCrMo alloy implants. Wear, corrosion andrelease of metal ions can be important in the complexaetiopathology of implant failure. However, we could notfind any association between the reason for revision, metalrelease, and material category, perhaps owing to lowstudy power. The clinical impact of metals released fromimplant corrosion is still unproven, but large, prospectiveclinical trials may be useful in establishing associationsbetween metal release, implant failure, and reason forrevision.

AcknowledgementsWe thank Birthe Thykjær for careful collection andstorage of the implant components, and Lizbet Skarefor skilled technical assistance with analysing all sampleswith the XRF instrument.

Supporting Information

Additional Supporting Information may be found in theonline version of this article:

Table S1. Detailed results on metal compositionaccording to X-ray fluorescence (XRF) spectroscopy,metal release according to spot tests for Ni, Co and Cr(VI),suggested alloy, and reason for revision surgery for 52failed total hip implants.

References

1 Malchau H, Herberts P, Eisler T, Garellick

G, Soderman P. The Swedish total hip

replacement register. J Bone Joint Surg Am

2002: 84-A (Suppl. 2): 2–20.

2 Malchau H, Herberts P, Ahnfelt L.Prognosis of total hip replacement inSweden. Follow-up of 92,675 operationsperformed 1978–1990. Acta Orthop Scand1993: 64: 497–506.

3 Sundfeldt M, Carlsson L V, Johansson C B,Thomsen P, Gretzer C. Aseptic loosening,not only a question of wear: a review ofdifferent theories. Acta Orthop 2006: 77:177–197.

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd324 Contact Dermatitis, 71, 319–325

FAILURE OF TOTAL HIP IMPLANTS • JAKOBSEN ET AL.

4 Caicedo M S, Samelko L, McAllister K,Jacobs J J, Hallab N J. Increasing bothCoCrMo-alloy particle size and surfaceirregularity induces increased macrophageinflammasome activation in vitropotentially through lysosomaldestabilization mechanisms. J Orthop Res2013: 31: 1633–1642.

5 Samelko L, Caicedo M S, Lim S J, Della-ValleC, Jacobs J, Hallab N J. Cobalt-alloy implantdebris induce HIF-1alpha hypoxiaassociated responses: a mechanism formetal-specific orthopedic implant failure.PLoS ONE 2013: 8: e67127.

6 Pandit H, Glyn-Jones S, McLardy-Smith Pet al. Pseudotumours associated withmetal-on-metal hip resurfacings. J BoneJoint Surg Br 2008: 90: 847–851.

7 Davies A P, Willert H G, Campbell P A,Learmonth I D, Case C P. An unusuallymphocytic perivascular infiltration intissues around contemporarymetal-on-metal joint replacements. J BoneJoint Surg Am 2005: 87: 18–27.

8 Jacobs J J, Urban R M, Hallab N J, Skipor AK, Fischer A, Wimmer M A.Metal-on-metal bearing surfaces. J AmAcad Orthop Surg 2009: 17: 69–76.

9 Willert H G, Buchhorn G H, Fayyazi A et al.Metal-on-metal bearings andhypersensitivity in patients with artificialhip joints. A clinical andhistomorphological study. J Bone Joint SurgAm 2005: 87: 28–36.

10 Hallab N J, Caicedo M, Finnegan A, Jacobs JJ. Th1 type lymphocyte reactivity to metalsin patients with total hip arthroplasty. JOrthop Surg Res 2008: 3: 6-799X-3-6.

11 Bisseling P, Tan T, Lu Z, Campbell P A,Susante J L. The absence of ametal-on-metal bearing does not precludethe formation of a destructivepseudotumor in the hip – a case report.Acta Orthop 2013: 84: 437–441.

12 Thyssen J P, Menne T, Johansen J D et al. Aspot test for detection of cobaltrelease – early experience and findings.Contact Dermatitis 2010: 63: 63–69.

13 Thyssen J P, Skare L, Lundgren L et al.Sensitivity and specificity of the nickel spot

(dimethylglyoxime) test. Contact Dermatitis2010: 62: 279–288.

14 Nilsson H, Malmgren-Hansen B,Christensen I, Lassen C. Development andUse of Screening Methods to DetermineChromium (VI) and Brominated FlameRetardants in Electrical and ElectronicEquipment: Copenhagen, DanishEnvironmental Protection Agency, 2009.

15 Overgaard S, Retpen J, Ulrich-Vinter Met al. Dansk Hoftealloplastik Register:Arsrapporten 2009, 2009: 1.

16 Overgaard S, Retpen J, Ulrich-Vinter Met al. Dansk Hoftealloplastik RegisterArsrapporten 2013, 2013: 1.

17 Hodgson A W, Mischler S, VonRechenberg B, Virtanen S. An analysis ofthe in vivo deterioration of Co-Cr-Moimplants through wear and corrosion. ProcInst Mech Eng [H] 2007: 221: 291–303.

18 Hedberg Y, Odnevall Wallinder I. Metalrelease and speciation of releasedchromium from a biomedical CoCrMoalloy into simulated physiologicallyrelevant solutions. J Biomed Mater Res BAppl Biomater 2014: 102: 693–699.

19 ISO. Biological Evaluation of MedicalDevices – Part 15: Identification andQuantification of Degradation Products fromMetals and Alloys. Geneva,10993–15:2000, 1988. ISO.

20 ASTM. Standard Specification for FemoralProstheses – Metallic Implants. F2068-09,2009. West Conshohocken, ASTMInternational.

21 Hallab N, Merritt K, Jacobs J J. Metalsensitivity in patients with orthopaedicimplants. J Bone Joint Surg Am 2001: 83-A:428–436.

22 Hallab N J, Jacobs J J. Biologic effects ofimplant debris. Bull NYU Hosp Jt Dis 2009:67: 182–188.

23 Gawkrodger D J. Metal sensitivities andorthopaedic implants revisited: thepotential for metal allergy with the newmetal-on-metal joint prostheses. Br JDermatol 2003: 148: 1089–1093.

24 Long M, Rack H J. Titanium alloys in totaljoint replacement – a materials scienceperspective. Biomaterials 1998: 19:1621–1639.

25 Troelsen A, Malchau E, Sillesen N,Malchau H. A review of current fixationuse and registry outcomes in total hiparthroplasty: the uncemented paradox.Clin Orthop Relat Res 2013: 471:2052–2059.

26 Gallo J, Kontinen Y, Goodman S et al.Aseptic loosening of total hip arthroplastyas a result of local failure of tissuehomeostasis. In: Recent Advances inArthroplasty, Fokter S (ed.): Rijeka, InTech,2012: p. 614.

27 MacDonald S J, McCalden R W, Chess D Get al. Metal-on-metal versus polyethylenein hip arthroplasty: a randomized clinicaltrial. Clin Orthop Relat Res 2003: 406:282–296.

28 Keegan G M, Learmonth I D, Case C P.Orthopaedic metals and their potentialtoxicity in the arthroplasty patient: areview of current knowledge and futurestrategies. J Bone Joint Surg Br 2007: 89:567–573.

29 Brown C, Lacharme-Lora L,Mukonoweshuro B et al. Consequences ofexposure to peri-articular injections ofmicro- and nano-particulatecobalt–chromium alloy. Biomaterials2013: 34: 8564–8580.

30 Thyssen J P, Linneberg A, Menne T,Johansen J D. The epidemiology of contactallergy in the generalpopulation – prevalence and mainfindings. Contact Dermatitis 2007: 57:287–299.

31 Granchi D, Cenni E, Tigani D, Trisolino G,Baldini N, Giunti A. Sensitivity to implantmaterials in patients with total kneearthroplasties. Biomaterials 2008: 29:1494–1500.

32 Ohnsorge J, Holm R. Surface investigationsof oxide layers oncobalt–chromium-alloyed orthopedicimplants using ESCA technique. Med ProgTechnol 1978: 5: 171–177.

33 Gilbert J L, Zarka L, Chang E, Thomas C H.The reduction half cell in biomaterialscorrosion: oxygen diffusion profiles nearand cell response to polarized titaniumsurfaces. J Biomed Mater Res 1998: 42:321–330.

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons LtdContact Dermatitis, 71, 319–325 325