Sensitivity and specificity of blood cobalt and chromium metal ions for predicting failure of metal-on-metal hip replacement

Transcription

1 HIP Sensitivity and specificity of blood cobalt and chromium metal ions for predicting failure of metal-on-metal hip replacement A. J. Hart, S. A. Sabah, A. S. Bandi, P. Maggiore, P. Tarassoli, B. Sampson, J. A. Skinner From Imperial College London, London, United Kingdom A. J. Hart, MA, MD, FRCSG(Orth), Senior Lecturer, Consultant Orthopaedic Surgeon S. A. Sabah, BSc, MBBS, Academic Foundation Doctor A. S. Bandi, BSc, MBBS, Foundation Year 1 Doctor P. Maggiore, MBBS, Senior House Officer P. Tarassoli, MBBS, Senior House Officer B. Sampson, MRSC, CChem, Consultant in Clinical Chemistry Imperial College London, Department of Musculoskeletal Surgery, Charing Cross Hospital Campus, Fulham Palace Road, London W6 8RF, UK. J. A. Skinner, FRCS(Orth), Consultant Orthopaedic Surgeon Royal National Orthopaedic Hospital, Brockley Hill, Stanmore, Middlesex HA7 4LP, UK. Correspondence should be sent to Mr A. J. Hart; a.hart@imperial.ac.uk 2011 British Editorial Society of Bone and Joint Surgery doi: / x.93b $2.00 J Bone Joint Surg Br 2011;93-B: Received 9 November 2010; Accepted after revision 1 June 2011 Blood metal ions have been widely used to investigate metal-on-metal hip replacements, but their ability to discriminate between well-functioning and failed hips is not known. The Medicines and Healthcare products Regulatory Agency (MHRA) has suggested a cut-off level of 7 parts per billion (ppb). We performed a pair-matched, case-control study to investigate the sensitivity and specificity of blood metal ion levels for diagnosing failure in 176 patients with a unilateral metal-on-metal hip replacement. We recruited 88 cases with a pre-revision, unexplained failed hip and an equal number of matching controls with a well-functioning hip. We investigated the 7 ppb cut-off level for the maximum of cobalt or chromium and determined optimal mathematical cut-off levels from receiver-operating characteristic curves. The 7 ppb cut-off level for the maximum of cobalt or chromium had a specificity of 89% and sensitivity 52% for detecting a pre-operative unexplained failed metal on metal hip replacement. The optimal cut-off level for the maximum of cobalt or chromium was 4.97 ppb and had sensitivity 63% and specificity 86%. Blood metal ions had good discriminant ability to separate failed from well-functioning hip replacements. The MHRA cut-off level of 7 ppb provides a specific test but has poor sensitivity. Hip replacement is undetaken on one million hips annually worldwide and in 2006 in the United States metal-on-metal (MoM) hips were used in 35% of cases. 1 In the United Kingdom the revision rate of MoM hip resurfacing is double that of other hip replacements, with unexplained pain being the most common reason. 2 One type of MoM hip has been withdrawn due to unacceptably high failure, which in some patients has been associated with adverse biological reactions. 3 Had these been detected earlier their management might have been easier. The question arises whether blood metal ion levels could have been used in a predictive manner to identify potential failures with sufficient sensitivity and specificity. The Medicines and Healthcare products Regulatory Agency (MHRA) safety alert in April drew attention to all types of MoM hips. The subsequent document from the Expert Advisory Group of the MHRA 4 explained the details behind the safety alert and included four situations in which measurement of blood metal ions were recommended: 1) patients who have symptoms associated with MoM bearings; 2) radiological features associated with adverse outcomes including component position or small component size; 3) if the patient or surgeon are concerned regarding the MoM bearing; 4) if there is concern about patients with higher than expected rates of failure. The MHRA have suggested that whole blood cobalt or chromium levels of greater than 7 ppb are associated with significant soft-tissue reactions and failed MoM hips. Blood cobalt and chromium ions levels in patients with unexplained painful MoM hips are double those of well-functioning MoM hips. 5 High levels of metal ions are related to a high inclination angle of the acetabular component, which is linked to an increased risk of revision. 6 However, in a recent report on the risk factors for failure due to pseudotumour, the mean inclination angle of the acetabular component was only Therefore we investigated the sensitivity and specificity of blood metal ions as a predictor of failure of a MoM hip replacement. Patients and Methods We designed a pair-matched, case-control study to investigate patients with current generation, large diameter unilateral MoM hip replacements who had undergone blood metal ion testing. The pre-operative diagnosis was osteoarthritis in all patients. Cases were 1308 THE JOURNAL OF BONE AND JOINT SURGERY

2 SENSITIVITY AND SPECIFICITY OF BLOOD METAL IONS FOR PREDICTING FAILURE OF METAL-ON-METAL HIP REPLACEMENT 1309 defined as patients awaiting revision with an unexplained, failed MoM hip following assessment including history and examination, serial pelvic radiographs and blood infection screening. Cross-sectional imaging to investigate for pseudotumour formation was not performed. Controls were defined as patients who were satisfied with their hip replacement and did not volunteer pain as a symptom. We recruited the first 88 cases of failure that presented to our institution between February 2007 and February 2011, and an equal number of matching controls from the author s (AJH) routine arthroplasty follow-up clinic from the same period. Matching criteria were gender and age (> 55 years or not). Patients were recruited at a minimum of 12 months after primary surgery. Institutional review board approval was obtained. All symptomatic patients have undergone revision surgery. Intra-operative findings and post-operative microbiological cultures were analysed to verify the cause of failure. Cases without an established cause of failure after these investigations were defined as true unexplained failures. We performed statistical analyses for pre-operative unexplained failures and true unexplained failures. All patients underwent whole blood analysis of cobalt and chromium following sampling from the antecubital vein using a 21-gauge needle connected to a Vacutainer system (Becton, Dickinson and Company, Franklin Lakes, New Jersey) and trace element blood tubes containing sodium ethylenediaminetetraacetic acid (EDTA). Sampling was performed at least one-year after initial MoM replacement. The first 10 ml of blood were used for C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) measurement. The second 10 ml were used for cobalt and chromium analyses. All patients had a normal estimated glomerular filtration rate from extrapolation of the blood creatinine measurement. Standard procedures were established for cobalt and chromium measurement using dynamic reaction cell inductively coupled plasma mass spectrometry (DRC-ICPMS) (PerkinElmer Elan DRCII, PerkinElmer, Waltham, Massachusetts). This method was validated with previously published methods following a blinded, inter-laboratory study. 8 The failed MoM hip group had blood metal ion levels taken immediately prior to revision surgery. Statistical analysis. The median and ranges of cobalt and chromium for well-functioning and failed groups were determined and the difference in medians was compared using the Wilcoxon rank sum test. A p-value of < 0.05 was considered to be statistically significant. Receiver operating characteristic (ROC) curves were constructed to show the relationship between true-positive (sensitivity) and falsepositive (1-specificity) cases for cobalt, chromium and the maximum value of either cobalt or chromium. The area under the curve (AUC), which depicts the accuracy of the test, was calculated for each of the above variables. The AUC was calculated using the trapezoidal method. 9 The AUC can be interpreted as the probability that the test result from a randomly chosen diseased individual is more indicative of disease than that from a randomly chosen non-diseased individual. If the two distributions are clearly separated, the probability will be close to 1; but if they are centred at the same value, the probability will be 0.5. An AUC of 1 demonstrates an ideal test with a 100% sensitivity and specificity, while an AUC of less than 0.5 indicates that the diagnostic test is less useful. We used the Wilcoxon non-parametric estimate to determine 95% confidence intervals (CI) for the AUC. The ROC curve correlates the true positive and false positive rates for a series of data points. The diagnostic cut-off values of cobalt and chromium were chosen as those values that corresponded to the points on the ROC curves nearest the upper left hand corner of the graph. This is to reflect the optimal mathematical balance between sensitivity and specificity, but may not be optimal in the clinical sense. Contingency tables were constructed to present the sensitivity, specificity, negative predictive value (NPV) and positive predictive value (PPV) for different cut-off levels for blood metal ions (including the MHRA defined level of 7 parts per billion (ppb) for either cobalt or chromium). Results Analysis of pre-operative unexplained failures. Demographic data. Details of the patients in the two groups are shown in Table I. There were 88 patients (44 male, 44 female) in each group. The most common type of hip in both groups was the Birmingham Hip Resurfacing (BHR; Smith & Nephew, Warwick, United Kingdom). The other hip types were: Cormet (Corin, Cirencester, United Kingdom); ASR (DePuy, Leeds, United Kingdom); Durom (Zimmer, Winterthur, Switzerland); Magnum M2a (Biomet, Bridgend, United Kingdom); Adept (Finsbury, Leatherhead, United Kingdom); Mitch (Stryker, Newbury, United Kingdom). The well-functioning patients had a median Harris hip score (HHS) 10 of 95 out of 100 (73 to 100) and a median Oxford hip score 11 of 45 out of 48 (30 to 48). In the patients in the unexplained failure group, the causes of failure remained unexplained after revision in 54 patients (61%), were due to aseptic loosening in 17 (19%), infection in six (7%), malalignment in four (5%), fracture in four (4%) and size-mismatch in three (3%). Patients were recruited to the well-functioning group at a median of 42 months (18 to 64) post-operation and to the failed group at a median of 39 months (13 to 131) postoperation. There was no significant difference between groups for this post-operative interval (Wilcoxon rank sum test, p = 0.646). Comparison of blood metal ion levels. Patients with a failed MoM replacement had a median cobalt level of 6.9 ppb (0.5 to 162.3) and a median chromium level of 5.0 ppb (0.0 to 116.0). Patients with a well-functioning MoM replacement had a median cobalt level of 1.7 ppb (0.4 to 80.7) and a median chromium level 2.3 ppb (0.4 to VOL. 93-B, No. 10, OCTOBER 2011

3 1310 A. J. HART, S. A. SABAH, A. S. BANDI, P. MAGGIORE, P. TARASSOLI, B. SAMPSON, J. A. SKINNER Table I. Summary of the clinical, implant and blood metal ion data from the two groups Well-functioning group Failed group Number of components Male:female 44:44 44:44 Median age (yrs) (range) 55.5 (33 to 71) 55.5 (22 to 83) Median duration implanted (mths) (range) 42 (18 to 64) 39 (13 to 131) Type of implant * 82 BHR 30 BHR 6 Cormet 15 Cormet 26 ASR 6 Durom 11 Other Median whole blood cobalt (ppb) (range) 1.7 (0.4 to 80.7) 6.9 (0.5 to 162.3) Median whole blood chromium (ppb) (range) 2.3 (0.4 to 27.7) 5.0 (0.0 to 116.0) Median whole blood maximum of cobalt and chromium (ppb) (range) 2.4 (0.4 to 80.7) 8.4 (0.5 to 162.3) * BHR, Birmingham Hip Resurfacing (Smith & Nephew); Cormet (Corin); ASR, Articular Surface Replacement (DePuy); Durom (Zimmer) Table II. Contingency table depicting the relationship between blood cobalt ion level and hip failure using the MHRA cut-off value of 7 ppb Well-functioning Failed Total Cobalt 7 ppb Cobalt > 7 ppb Total Table V. Area under the curve (AUC) with 95% confidence intervals (CI) of receiver operating characteristics curves. Non-parametric CIs are presented Cobalt Chromium Maximum of cobalt or chromium AUC 74.2% 68.6% 74.2% Lower 95% CI 66.6% 60.5% 66.6% Upper 95% CI 81.9% 76.7% 81.9% Table III. Contingency table to depict the relationship between blood chromium ion level and hip failure using the MHRA cut-off value of 7 ppb Well-functioning Failed Total Chromium 7 ppb Chromium > 7 ppb Total Table VI. Summary of diagnostic test characteristics for cobalt and chromium using the MHRA cut-off value of 7 ppb and confidence intervals (CI) Test Sensitivity (95% CI) Specificity (95% CI) Cobalt > 7 ppb 0.49 (0.38 to 0.60) 0.90 (0.81 to 095) Chromium > 7 ppb 0.38 (0.28 to 0.49) 0.92 (0.84 to 0.96) Cobalt or chromium > 7 ppb 0.52 (0.41 to 0.63) 0.89 (0.80 to 0.94) Table IV. Contingency table to depict the relationship between the maximum of cobalt and chromium and hip failure using the MHRA cutoff value of 7 ppb Well-functioning Failed Total Maximum of cobalt or chromium ppb Maximum of cobalt or chromium > 7 ppb Total Table VII. Summary of diagnostic test characteristics for cobalt and chromium using cut-off values (parts per billion (ppb)) determined from receiver operating characteristic curves (CI, confidence interval) Test Sensitivity (95% CI) Specificity (95% CI) Cobalt > 2.74 ppb 0.69 (0.58 to 0.78) 0.80 (0.69 to 0.87) Chromium > (0.58 to 0.78) 0.68 (0.57 to 0.77) ppb Maximum cobalt or chromium > 4.97 ppb 0.63 (0.51 to 0.72) 0.86 (0.77 to 0.92) 27.7). The metal ion levels were significantly different for cobalt, chromium and the maximum of cobalt and chromium (Wilcoxon rank sum test, all p < 0.001) in the two groups. Contingency tables to depict the relationship between failed and well-functioning hips for the MHRA cut-off value of 7 ppb are shown in Tables II to IV. ROC curves showed the relationship between true-positive (sensitivity) and false-positive (1-specificity) patients using different cut-off values of cobalt, chromium and maximum value of cobalt or chromium (Fig. 1). The values of AUC for cobalt, chromium and maximum of cobalt or chromium are presented in Table V. These show that cobalt or the maximum of cobalt or chromium are the most useful tests since they have the greatest AUCs but there is little difference between all three AUCs. The ROC curves enabled calculation of the following cut-off levels to provide the optimum mathematical balance between sensitivity and specificity: 2.74 ppb for cobalt, 2.69 ppb for chromium, and 4.97 ppb for maximum of cobalt or chromium. THE JOURNAL OF BONE AND JOINT SURGERY

4 SENSITIVITY AND SPECIFICITY OF BLOOD METAL IONS FOR PREDICTING FAILURE OF METAL-ON-METAL HIP REPLACEMENT 1311 Sensitivity Specificity Diagonal segments are produced by ties Fig. 1 Co Cr Max CoCr Receiver operating characteristic (ROC) curves showing the relationship between true-positive (sensitivity) and false-positive (1 - specificity) cases using different cut-off values of cobalt (Co), chromium (Cr) and maximum value of cobalt or chromium (MaxCoCr). A summary of the diagnostic features of all tests used with both MHRA and ROC curve derived cut-off levels is shown in Tables VI and VII. Analysis of true unexplained failures. Following intra- and post-operative analysis of patients in the unexplained failure group, 54 patients (61%) were determined to have true unexplained failure. The analyses performed above for pre-operative unexplained failures were repeated for this group and their paired controls. The effect on the data was small: for example the AUC for the maximum of cobalt and chromium was 74.2% and 75.4% for pre-operative unexplained and true unexplained subsets respectively. Discussion Blood metal ion analysis permits investigation of MoM hip replacement and has been widely used in the research setting. 5,12-16 However, the ability of metal ion levels to discriminate between well and poorly functioning hips has not previously been reported. In accordance with a previously published study, 5 we have shown that patients with a failed MoM hip replacement have significantly higher blood levels of both cobalt and chromium compared with patients with a well-functioning hip. We have also presented the diagnostic test characteristics for a number of different cutoff levels for cobalt, chromium and the maximum of either cobalt or chromium which might help surgeons to apply blood metal ion levels when investigating an unexplained, painful MoM hip replacement. We demonstrated that the MHRA cut-off level of 7 ppb for each metal ion and the maximum of the two had a high specificity (approximately 90%) but a low sensitivity (approximately 50%) for detecting a failure. As would be expected with any condition where a high level of a parameter is associated with disease, we showed that lowering the cut-off level improved sensitivity but reduced specificity. In the same manner, analyzing cobalt and chromium levels together using the maximum value of the two, increased sensitivity at the cost of specificity for a given cut-off level. The clinical advantage for grouping cobalt and chromium together is that only one cut-off level need be referenced. Since ROC curves and other test characteristics for cobalt and chromium are qualitatively similar, it would appear valid to do this. We recognise the limitations of our study. First, our hospital acts as a tertiary centre for assessment and revision of patients with unexplained, painful MoM hip arthroplasties, which accounts for the high proportion of prerevision unexplained failed MOM hips in our study population. Secondly, the diagnosis of failed MoM hip replacements was limited pre-operatively by the absence of crosssectional imaging to investigate for the presence of a periprosthetic pseudotumour. We did not exclude patients with raised inflammatory markers from our pre-operative unexplained failed population unless there was overt sepsis (such as discharging sinus or a fever with positive culture from a hip aspirate or positive blood culture). Establishing a pre-operative diagnosis of infection is a well-recognised problem for revision surgeons and is made more difficult in MoM hip replacements where sterile soft-tissue inflammation can also raise inflammatory markers. 17 We have minimised the inclusion of infected hips in our true unexplained failure population by categorising hips on the basis of the post-operative cultures. Additionally several MoM designs were included; however, the BHR was the most frequent in both groups and the MHRA safety alert referred to all types of MoM hip replacement, so that analytical evaluation of blood metal ions should probably involve all types. Furthermore, the different designs are thought to share a similar mechanism of failure of high wear rates and high metal ion release, 18 so that raised blood metal ions is a possible common association with failure of all types. One should be careful about generalising our results. For example, only patients with a unilateral MoM replacement were investigated making it unlikely that these results can be extrapolated to patients with bilateral MoM replacement. Applying the cut-off level of 7 ppb suggested in the MHRA safety alert, the maximum of cobalt or chromium gave a sensitivity of 52% and a specificity of 89%. A lower cut-off level increased the sensitivity (ability to identify failures) but decreased the specificity because more well-functioning replacements will have levels above this cut-off; this is a familiar situation for most clinical tests. The ROC analysis identified the best trade-off between sensitivity and specificity, which in our study showed a cut-off level of 4.97 ppb for the maximum of cobalt or chromium and gave a sensitivity of 63% and a specificity of 86%. Table VIII shows the effect of the cut-off level on sensitivity and VOL. 93-B, No. 10, OCTOBER 2011

5 1312 A. J. HART, S. A. SABAH, A. S. BANDI, P. MAGGIORE, P. TARASSOLI, B. SAMPSON, J. A. SKINNER Table VIII. The effect of changing the cut-off level (parts per billions) for the maximum of either cobalt or chromium on sensitivity and specificity (CI, confidence interval) Cut-off (ppb) Sensitivity (95% CI) Specificity (95% CI) % (51 to 72) 85% (76 to 92) % (41 to 63) 91% (82 to 96) % (34 to 55) 95% (88 to 99) specificity for predicting failure. However, a lower cut-off value risks classifying a large proportion of well-functioning MoM hips as failures. This highlights the fact that blood metal ion levels, as with other analyses and investigations, are only one part of the evaluation of a problematical MoM hip replacement, on which a clinical recommendation might be to undertake revision. The validity of blood cobalt and chromium testing relies on several assumptions: (i) that these metal ions originate from the hip of interest, which might not be so in bilateral cases or those with other metal implants or occupational exposure; (ii) that metal ion levels do not vary significantly on a day to day basis so that high levels can be detected by random testing; (iii) that the test used is sufficiently sensitive to predict future failure; (iv) that detecting high metal ion levels will be reasonably specific for predicting future failure. Whilst our study aids the assessment of blood metal ion levels as a screening test, the epidemiology and natural history of the condition are not fully understood and there has been no validation of the laboratory limits for blood metal ion levels. Accordingly, we cannot recommend screening asymptomatic patients from the data in this study alone. A high acetabular component inclination angle measured on plain radiographs is linked to failure 19 with angles > 50 associated with high blood metal ion levels. 6,20,21 However, many failures have occurred where the inclination angle has been satisfactory. The mean acetabular component abduction angle in 25 cases of failure with associated pseudotumour was Difficulties using plain radiological inclination angle as a predictor of failure include the uncertainty in measurement 22 and the influence of the implant design and component size on the coverage arc. 4,23 Overall, blood metal ion levels may be more predictive of failure than plain radiological findings. Further research, perhaps combining these tests is needed. Blood cobalt and chromium levels are a useful diagnostic test for discrimination of poor from well-functioning metal-on-metal hip replacements with levels of sensitivity and specificity comparable to other frequently used tests in clinical medicine. There was little difference between the cobalt, chromium or maximum of either cobalt or chromium for the sensitivity and specificity for each cut-off level. However, for simplicity we have chosen the maximum value for our approach to the clinical situation. The MHRA cut-off level of 7 ppb enables high specificity but a relatively low sensitivity. Further research may suggest other cut-off levels but blood metal ion levels should not be completely relied upon; they are likely to remain an important adjunct to the clinical assessment and further investigation of patients with MoM hip replacements. Supplementary material Tables and figures describing further statistical analyses by gender, true unexplained failures and unmatched well-functioning Birmingham Hip Resurfacing replacements, are available with the electronic version of this article on our website at The authors wish to acknowledge the help of Sister G. Lloyd (coordinator of the London Implant Retrieval Centre (LIRC)) for assistance in collection of data, and all the surgeons that referred patients with problematic metal-on-metal hips. This work was funded by the British Orthopaedic Association (BOA) through an industry consortium of nine manufacturers: DePuy, Zimmer, Smith & Nephew, Biomet, JRI, Finsbury, Corin, Mathys and Stryker. The contract allows for freedom to publish all the results. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. References 1. Bozic KJ, Kurtz S, Lau E, et al. The epidemiology of bearing surface usage in total hip arthroplasty in the United States. J Bone Joint Surg [Am] 2009;91-A: No authors listed. National Joint Registry for England and Wales: 5th annual report. 2008:78. (date last accessed 11 January 2011). 3. Toms AP, Marshall TJ, Cahir J, et al. MRI of early symptomatic metal-on-metal total hip arthroplasty: a retrospective review of radiological findings in 20 hips. Clin Radiol 2008;63: No authors listed. Medicines and Healthcare products Regulatory Agency (MHRA). Medical Device Alert: all metal-on-metal (MOM) hip replacements, 2010 MDA/2010/ CON (date last accessed 22 June 2011). 5. Hart AJ, Sabah S, Henckel J, et al. The painful metal-on-metal hip resurfacing. J Bone Joint Surg [Br] 2009;91-B: De Haan R, Pattyn C, Gill HS, et al. Correlation between inclination of the acetabular component and metal ion levels in metal-on-metal hip resurfacing replacement. J Bone Joint Surg [Br] 2008;90-B: Glyn-Jones S, Pandit H, Kwon YM, et al. Risk factors for inflammatory pseudotumour formation following hip resurfacing. J Bone Joint Surg [Br] 2009;91-B: Winship PD, Faria N, Skinner JA, Hart AJ, Powell JJ. The determination of cobalt and chromium in whole blood sampled from metal-metal and metal-polyethylene hip articulation patients by DRC-ICP-MS. In: Collery P, Maymard I, Theophanides T, Khassanova L, Collery T, eds. Metal ions in biology and medicine. Vol. 10. Paris: John Libbey Eurotext, 2008: Atkinson KE. An introduction to numerical analysis. Second ed. New York: John Wiley & Sons, Harris WH. Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty: an end-result study using a new method of result evaluation. J Bone Joint Surg [Am] 1969;51-A: Dawson J, Fitzpatrick R, Carr A, Murray D. Questionnaire on the perceptions of patients about total hip replacement. J Bone Joint Surg [Br] 1996;78-B: Antoniou J, Zukor DJ, Mwale F, et al. Metal ion levels in the blood of patients after hip resurfacing: a comparison between twenty-eight and thirty-six-millimeterhead metal-on-metal prostheses. J Bone Joint Surg [Am] 2008;90-A(Suppl): Bitsch RG, Zamorano M, Loidolt T, et al. Ion production and excretion in a patient with a metal-on-metal bearing hip prosthesis: a case report. J Bone Joint Surg [Am] 2007;89-A: De Smet K, De Haan R, Calistri A, et al. Metal ion measurement as a diagnostic tool to identify problems with metal-on-metal hip resurfacing. J Bone Joint Surg [Am] 2008;90-A(Suppl): Jacobs JJ, Gilbert JL, Urban RM. Corrosion of metal orthopaedic implants. J Bone Joint Surg [Am] 1998;80-A: Langton DJ, Sprowson AP, Joyce TJ, et al. Blood metal ion concentrations after hip resurfacing arthroplasty: a comparative study of articular surface replacement and Birmingham Hip Resurfacing arthroplasties. J Bone Joint Surg [Br] 2009;91-B: Mikhael MM, Hanssen AD, Sierra RJ. Failure of metal-on-metal total hip arthroplasty mimicking hip infection: a report of two cases. J Bone Joint Surg [Am] 2009;91- A: THE JOURNAL OF BONE AND JOINT SURGERY

6 SENSITIVITY AND SPECIFICITY OF BLOOD METAL IONS FOR PREDICTING FAILURE OF METAL-ON-METAL HIP REPLACEMENT Langton DJ, Jameson SS, Joyce TJ, et al. Early failure of metal-on-metal bearings in hip resurfacing and large-diameter total hip replacement: a consequence of excess wear. J Bone Joint Surg [Br] 2010;92-B: Morlock MM, Bishop N, Zustin J, et al. Modes of implant failure after hip resurfacing: morphological and wear analysis of 267 retrieval specimens. J Bone Joint Surg [Am] 2008;90-A(Suppl): Hart AJ, Buddhdev P, Winship P, et al. Cup inclination angle of greater than 50 degrees increases whole blood concentrations of cobalt and chromium ions after metal-on-metal hip resurfacing. Hip Int 2008;18: Langton DJ, Jameson SS, Joyce TJ, Webb J, Nargol AV. The effect of component size and orientation on the concentrations of metal ions after resurfacing arthroplasty of the hip. J Bone Joint Surg [Br] 2008;90-B: Kalteis T, Handel M, Herold T, et al. Position of the acetabular cup: accuracy of radiographic calculation compared to CT-based measurement. Eur J Radiol 2006;58: Shimmin AJ, Walter WL, Esposito C. The influence of the size of the component on the outcome of resurfacing arthroplasty of the hip: a review of the literature. J Bone Joint Surg [Br] 2010;92-B: VOL. 93-B, No. 10, OCTOBER 2011