Getting it right the first or second time? First and second revisions of metal-on-metal hip resurfacing prostheses compared to non-metal-on-metal total hip prostheses in young Dutch patients
A. Spekenbrink-Spooren1 L.N. van Steenbergen1 G.A.W. Denissen1 W.P. Zijlstra2 R.G.H.H. Nelissen1,3 B.W. Schreurs1,4 1 Dutch Arthroplasty Register (Landelijke Registratie Orthopedische Implantaten), ’s-Hertogenbosch, The Netherlands 2 Department of Orthopaedics, Medical Center Leeuwarden, Leeuwarden, The Netherlands 3 Department of Orthopaedics, Leiden University Medical Center, Leiden, The Netherlands 4 Department of Orthopaedics, Radboudumc, Nijmegen, The Netherlands
Corresponding author: A. Spekenbrink-Spooren, aspekenbrink@orthopeden.org
Background: hip arthroplasties are increasing in number, also in young patients. This will inevitably lead to not only more first, but also more second revision procedures. Metal-on-metal (MoM) resurfacing hip arthroplasties (RHA) showed initial promising results, but in some series disadvantageous mid-term results. In addition to the results after the primary RHA, outcome after first revisions of these RHAs in this young patient group were studied. We compared the outcome of these RHA with non-MoM total hip arthroplasty in young patients.
Methods: we selected all patients <55 years with a primary non-MoM total hip arthroplasty (THA) (n=10,328) or MoM RHA (n=1,467) in the period 2007-2013 from the nationwide Dutch Arthroplasty Register (LROI). Cumulative incidence of first and second revision was calculated using competing risk analyses. Cox regression analyses were performed, adjusted for age at surgery, ASA-score, and diagnosis. Results: the mean age at the primary procedure was 47.5 (SD: 5.8) years for MoM RHAs, compared to 47.0 (SD: 7.5) years for non-MoM THAs. The ten-year cumulative first revision rate for RHA was 13.9% (CI:12.2-15.8) and 6.2% (CI:5.6-6.8) for non-MoM THA. The adjusted hazard ratio for first revision of RHA compared to non-MoM THA was 1.6 (CI:1.2-2.0) for males and 4.0 (CI:3.2-4.9) for females. The five-year cumulative second revision rate for revised RHA was 9.0% (CI:5.7-14.2), compared to 21.7% (CI:18.3-25.7) for revised non-MoM THA. The adjusted hazard ratio for second revision of revised RHA compared to revised non-MoM THA was 0.5 (CI:0.3-0.8). Sub analyses excluding first revisions for infection showed similar results. Conclusion: MoM RHA is not a better option than non-MoM THA for young patients in terms of primary prosthesis survival, especially not for females. However, the second revision rate of MoM RHA is lower compared to non-MoM THA in both male and female patients.
Introduction Survival of total hip arthroplasties (THA) in younger patients is well below the survival rate in patients >65 years1. This might be related to existing preoperative deformities due to childhood diseases or trauma resulting in an (acetabular) bone stock loss. On the other hand, these younger patients often have a more active lifestyle, leading to higher wear rates and these particles can generate aseptic loosening. Therefore, hip prostheses with low wear characteristics were designed, including metal-on-metal (MoM) resurfacing hip arthroplasties (RHA)2. However, after the introduction of these large head MoM RHAs on a larger scale, a variety of unexpected complications were seen, like early aseptic loosening or osteolysis3, raised serum metal ion levels4–6 and adverse local periprosthetic tissue reactions7. In contrast, some studies still claim very good results in specific patient groups benefitting from RHA (i.e., young males)8–11. A recent study of Van Der Straeten et al. (2020) showed good results of RHAs performed in expert centres11. Nevertheless, overall, data from national registry studies show higher revision rates of a large number of these MoM RHAs in young patients12–15. Given the often young age of these patients at first revision, the outcome of these revisions is also important. Literature shows that second revision rates are high after revision of a failed RHA, although this depends on the reason for revision (e.g., adverse reaction to metal debris)16,17. These papers, however, do not focus on young patients. Studying the outcome of a second revision can also help to choose the type of first revision, improving care in these younger patients in need for revision surgery18. The aim of the current study was twofold; to compare first revision rates of resurfacing MoM RHAs versus non-MoM THAs in male and female patients <55 years, using data of the Dutch Arthroplasty Register. Second, to evaluate the rerevision rate of these first revisions (i.e., second revision rate) in these two populations.
Materials and methods Dutch Arthroplasty Register The Dutch Arthroplasty Register (LROI) is a nationwide population-based register in which information about joint arthroplasties in the Netherlands is collected since 2007. The LROI is initiated by the Netherlands Orthopaedic Association (NOV) and has a coverage of 100% of Dutch hospitals and a completeness up to 99% for primary THAs and 98% for hip revision arthroplasties19.
Data collection The LROI database contains information on patient, procedure, and prosthesis characteristics19. For each component the product number is registered to document prosthesis characteristics. Primary arthroplasty is defined as the first implantation of a prosthesis to replace the joint. Revision arthroplasty is defined as any exchange (insertion, replacement, removal) of one or more components, including head or liner exchanges19. Major revision was defined as a revision of at least the femoral or acetabular component. Minor revision was defined as a revision of only the acetabular liner and/or femoral head component. Vital status of the patient was cross-checked by linkage with the Vektis database: the national insurance database on health care in the Netherlands19. The LROI uses the opt-out system to require informed consent of patients. For the present study, all cases aged <55 years who received a primary MoM RHA or non-MoM THA in the period 2007-2013 were included (n=11,795). The study period was limited to the period in which MoM RHAs were used in the Netherlands19. MoM bearings in combination with standard THAs were excluded. In this study we focussed on the outcome of RTHA versus non-MoM standard total hips. Non-MoM THAs were selected based on the bearing surface of the head, the type of acetabular inlay or the (cemented) monoblock polyethylene (PE) cup and included ceramic-on-PE, metal-on-PE and ceramic-on-ceramic articulation. Patient records with missing data on gender or age were excluded (n=136). Age was divided into three groups: <40, 40-50, and 50-55 years. The overall physical condition was scored using the ASA-score (I-IV). Diagnosis was categorized as osteoarthritis (OA) or non-osteoarthritis (e.g., hip fractures, osteonecrosis, dysplasia, late posttraumatic and tumour). Periods of evaluation were divided into 2007-2009, 2010-2011, and 2012-2013: three periods of different MoM problem awareness20.
Statistics Patient characteristics were described for RHA and non-MoM THA and compared using a Chi-square-test to test differences between RHA and non-MoM THA stratified by gender. Survival time of the primary procedure (outcome: first revision) was calculated as the time from primary hip arthroplasty to first revision arthroplasty for any reason, death of the patient, or the end of the study follow-up (1-1-2019). Survival time of the first revision procedure (outcome: second revision) was calculated as the time from first to second revision arthroplasty. Survival analyses with outcome first revision were stratified by gender. Cumulative crude incidence of revision was calculated using competing risk analysis, where death was considered to be a competing risk21. Multivariable Cox regression analyses were performed to compare adjusted revision rates between RHA and non-MoM THA. Adjustments were made for age at surgery (first revision), ASA-score, diagnosis (first and second revision) and gender (second revision). For all covariates added to the model, the proportional hazards assumption was met after inspecting log-minus-log curves. Reasons for revision were described according to type of hip arthroplasty (RHA and non-MoM THA) and compared using a Chi-square test to test differences between types of hip arthroplasty (SAS 7.1 and SPSS 24). P-values below 0.05 were considered statistically significant; 95% confidence intervals were referred to as CI.
Results Descriptives In the period 2007-2013 1,467 MoM primary RHAs (910 male patients; 557 female patients; mean age: 47.0 (SD: 7.5)) and 10,328 non-MoM primary THAs (mean age: 47.5 (SD: 5.8)) were performed in patients aged <55 years and registered in the LROI. The overall median follow-up after the primary procedure was 7.6 years (range: 0-12 years), for RHA the median follow-up was 9.3 (range: 0-12 years); for non-MoM THA: 7.2 (range: 0-12 years)) with a maximum of 12 years. The median follow-up after first revision of all included cases was 5.2 years (range: 0-12 years); this was for RHA 6.8 years (range: 0-11) years and for non-MoM THA: 4.7 years (range: 0-12 years)). Young patients with a RHA at the primary procedure were more often male (62%, compared to 45.3% in non-MoM THA). Young female patients more often underwent a first revision procedure of a RHA as well as of a non-MoM THA. Young patients who underwent a first revision of a RHA generally had a lower ASA-score, and more often had osteoarthritis compared to young patients who underwent a revision to a non-MoM THA. Furthermore, the period in which first revisions were performed is evenly distributed among non-MoM THAs, whereas the period of first revision of most RHAs was 2007-2009 (table 1).
Outcome: first revision In total, 209 (14.2%) RHAs (83 males, 126 females) and 578 (5.6%) non-MoM THAs (251 males, 327 females) were revised during the study period. First revisions of RHAs were in general (97%) major revisions (i.e., revision of at least the femoral or acetabular component), while first revisions of non-MoM THAs were major revisions in 82% (p<0.001). In major revisions of RHAs, generally both the acetabular and femoral component were revised (88%; only femur: 8%; only acetabulum: 4%), compared to 28% total revisions of the major revisions (only femur: 40%; only acetabulum: 32%) of major revisions of non-MoM THAs (p<0.001). Furthermore, in 39% of major revisions of non-MoM THAs only the acetabular component was revised and in 32% only the femoral component. The ten-year cumulative first revision rate for MoM RHA was 13.9% (CI:12.2-15.38) and 6.2% (CI:5.6-6.8) for non-MoM THA. Among young males, RHA had a crude cumulative ten-year revision rate of 8.8% (CI:7.1-10.9) and non-MoM THA had a crude cumulative ten-year revision rate of 5.6% (CI:4.9-6.4) (figure 1a; table 2). In young men, RHAs were 1.6 (CI:1.2-2.0) times more likely to be revised, compared with conventional non-MoM THAs (table 3). Among young females, RHA had a crude cumulative ten-year revision rate of 22.1% (CI:18.9-25.9), while non-MoM THA had a crude cumulative ten-year revision rate of 6.6% (CI:5.8-7.5) (figure 1b; table 2). RHAs were 4.0 (CI:3.2-4.9) times more likely to be revised, compared with non-MoM THAs (table 3). A sub analysis among only the OA patients in both types of implants showed similar results (overall HR: 2.5 (CI:2.1-3.0); males HR: 1.6 (1.2-2.2); females: 4.0 (3.1-5.1)).
Figure 1a and 1b. Cumulative incidence of first revision for patients aged <55 years according to type of hip arthroplasty in the Netherlands in the period 2007-2013, males (n=5,590) (A) and females (n=6,205) (B). RHA: Resurfacing Hip Arthroplasty; MoM: metal-on-metal; THA: Total Hip Arthroplasty.
A symptomatic MoM bearing was registered as the most frequent reason for a first revision in patients with a RHA, followed by loosening of the acetabular component and loosening of the femoral component. Dislocation, loosening of the femoral component and loosening of the acetabular component were the most frequent reasons for revision for non-MoM THA. Furthermore, dislocation, infection and loosening of the femoral component were significantly more often a reason for revision in patients who underwent a non-MoM THA, compared to patients with a RHA (table 4).
Outcome: second revision In total, 27 (12.9%) patients with a RHA (14 males, 13 females) and 126 (21.8%) patients with a non-MoM THA (64 males, 62 females) had a second revision procedure during the study period. Minor second revisions (e.g., revision of only the inlay acetabular liner and/or femoral head component) were done in 3 (11%) revised RHAs and 23 (18%) revised non-MoM THAs. Patients with a RHA appear to be half as likely (HRadj: 0.5 (CI:0.3-0.8) to undergo a second revision procedure within five years of the first revision procedure, compared to patients with a non-MoM THA (table 3). A sub analysis excluding first revisions for infection showed comparable results (HRadj: 0.6 (CI:0.4-0.9)). The crude cumulative five-year second revision rate of RHAs was 9.0% (CI:5.7-14.2; n=136 at risk at five years), while the crude cumulative five-year second revision rate of non-MoM THAs was 21.7% (CI:18.3-25.7; n=266 at risk at five years) (figure 2; table 5). A sub analysis for major first revisions showed that the crude cumulative five-year second revision rate of RHAs was 9.2% (CI:5.8-14.4%; n=135 at risk at five years), while the crude cumulative five-year second revision rate of non-MoM THAs was 19.4% (CI:15.9-23.7; n=259 at risk at five years).
The most frequently registered reason for second revision of RHAs was symptomatic MoM bearing (n=10, 37%). These were revisions of acetabular components and femoral heads during the second revision procedure, after the femoral component was revised towards a large head MoM THA during the first revision. Other reasons for second revision were infection (n=4, 15%), loosening of the acetabular component (n=4, 15%) and dislocation (n=3, 11%). The most frequently registered reason for second revision of non-MoM THAs was infection (n=43, 34%), followed by Girdlestone situation (n=33, 26%), loosening of the acetabular component (n=23, 18%) and dislocation (n=22, 18%).
Discussion The likelihood of a first revision in young male patients with a RHA was 1.6 times higher compared to a non-MoM THA, while for female patients this was 4.0 times higher. After this first revision, a lower second revision rate was found for young patients following a primary RHA, compared to patients with a non-MoM THA. However, numbers at risk for second revision of a RHA were low but certainly merits further research.
As for the first revision of hip arthroplasty in patients <55 years, RHA showed higher revision rates in both young male and in female patients. The latter is in line with results from the NARA (i.e. patients aged <50 years)14, and the Australian registry12. Clearly, there are case-series and small cohort studies reporting very acceptable survivals of RHA in young patients9,10. Of course, expert centres will produce better outcomes11, as seen in many other fields in orthopaedics. These seemingly contradictory results between population-based registry studies and these small studies may however also indicate a selection bias or methodological flaws (e.g., lost to follow-up). Since the LROI has a high completeness rate of 98% for revision19, lost to follow-up is very low, although not revising an implant may also cause bias in national registry studies. Non registry data suggest that specific patient groups may possibly benefit from these RHAs, but needs rigorous follow-up and analysis15. Also in these highly selective populations, a sometimes observed extensive pseudotumor growth can lead to a dramatic disappointing end result. The risk of revision within five years after a RHA was reported to be 5.3% for men and 13.5% for female patients in the Netherlands in 201422. However, in 2012, the Netherlands Orthopaedic Association (NOV) advised against the use of all large head (≥36mm) MoM hip implants, including RHA, until safety and long-term effectiveness for patients had been established conclusively23,24. This advice included an active follow-up of all patients with MoM hip prostheses. The awareness of orthopaedic surgeons and patients and the active follow-up by orthopaedic clinics as advised by the NOV might have resulted in a higher revision rate of RHAs without the occurrence of any serious clinical problems. This attention certainly may have increased the number of revisions in RHA. This is reflected in the indication for revision: symptomatic metal on metal bearing was registered as (one of) the reason(s) in about 50% of the revised RHAs. This is a poorly defined indication that includes both extensive pseudotumors as well as patients who were anxious about all the negative press around these implants and sometimes wanted to exchange a well-functioning implant. Our finding that young patients with RHAs were less likely to undergo a second revision procedure compared to young patients with non-MoM THAs seems to be a new finding, but since the second revision group of RHA was small a type II error may be present. However, further research on this interesting topic is needed. A recent study with LROI-data on second revisions in patients <55 years, including all hip revision procedures, but excluding MoM primary hip arthroplasties, found a five-year second revision rate of 22% and a ten-year second revision rate of 28%25. Furthermore, the rate of infections was 9% in first, while this was 29% in second revision procedures, concluding that prevention of infection in primary THA in young patients is of uttermost importance. Also, the treatment of a once infected primary total hip implants is of critical importance to prevent re-revisions for infection. We also found a 22% (CI:18-26%) five-year second revision rate of non-MoM THAs, but only a 9% (CI:6-14%) five-year second revision rate of RHAs. Our study also showed higher infection rates in first revisions of non-MoM THAs, compared to RHAs. Even when excluding first revisions for infection, we still found significantly higher second revision rates for young patients with non-MoM THAs compared to RHA. A likely explanation for our findings may be found in the baseline characteristics of patients receiving a RHA, compared to patients with a non-MoM THA. Patient characteristics influence revision rates of THA, with ASA-score being one of the strongest predictors for short-term first revision based on LROI data26. In our study, patients receiving a primary RHA were more likely to be healthier (i.e., lower ASA-scores). Another remarkable finding was that patients with RHAs were less likely to be aged <40 years, compared to non-MoM THAs. Patients under 40 were more likely to receive the primary RHA for osteoarthritis. These are probably young patients who have secondary hip osteoarthritis mainly after juvenile childhood hip disease with bone stock loss, less easy to address with a RHA. Thus, our results concerning second revision rate may be contributable to patient factors as well as the prosthesis. To take these factors into account, Cox regression analyses adjusted for age at primary surgery, ASA-score and diagnosis were performed, finding similar results.
Strengths and limitations The Dutch Arthroplasty Register (LROI) has a coverage of 100% of all hospitals and >95% completeness27, reducing inclusion bias. The relatively large sample size and diversity of surgeons, patients, and prostheses guarantee good generalizability and external validity of data. Registry data confirm that patient demographics and disease profiles are comparable between our registry and registries from other countries28,29. An average of 30,000 primary hip arthroplasties were performed annually since 2010, which increased to 39,000 in 201919. In the younger population (<55 years of age) about 1,950 THA procedures were performed in 2010 which increased to 2,950 procedures in 2019. However, the number of revised hip arthroplasties (3,000 in 2010, increasing to 3,800 in 201919) is still relatively small due to the excellent overall long-term survival of hip prostheses. A limitation of registry studies is confounding by indication as well as limited collection of data on patient characteristics and surgical procedure. Residual confounders may be present, such as bone quality, physical activity, and patient and surgeon preferences. Surgical experience cannot be taken into account, nor can correct placement of the prosthesis. Causality cannot be proven due to the observational nature. RHAs were more often performed on otherwise healthy patients and in patients with a relatively normal hip anatomy. If bone defects were more extensive during the primary procedure, which is more likely at younger age, a RHA was not an option. These non-MoM THAs would have a more extensive first revision, increasing the risk for a second revision procedure. This discrepancy may have been amplified as a result of the NOV advice to actively follow-up on patients with a MoM RHA. The NOV advice may have lowered the threshold for first revision of MoM RHAs, risking overtreatment, resulting in a better condition of the hip for second revision. Another limitation is the limited follow-up period for second revision procedures in this study. This is therefore a subject for further research. Furthermore, the first years (2007-2009) of the LROI registration was the run-in phase and therefore, the completeness was suboptimal19. Lastly, we acknowledge that we have no data on quality of life or clinical hip function, since these data were only added to the registry from 2014 onwards.
Conclusion In conclusion, based on LROI data, MoM RHA is not a better option than non-MoM THA for young patients in terms of primary prosthesis survival, especially not for females (22% at ten years; males: 9%). However, the second revision rate of revised MoM RHA is lower compared to non-MoM THA (9% vs. 22% at five years) which needs further analysis and a longer follow-up. These results contribute to improved patient information for patients undergoing a RHA revision procedure. Disclosure statement Authors have no conflict of interest to declare.
References 1. Bayliss LE, Culliford D, Monk AP, et al. The effect of patient age at intervention on risk of implant revision after total replacement of the hip or knee: a population-based cohort study. Lancet 2017;389:1424–1430. 2. McMinn D, Daniel J. History and modern concepts in surface replacement. PIME H 2006;220:239–251. 3. Zijlstra WP, Bulstra SK, Raay JJ van, et al. Cobalt and chromium ions reduce human osteoblast-like cell activity in vitro, reduce the OPG to RANKL ratio, and induce oxidative stress. JOR 2012;30:740–747. 4. Engh CA, MacDonald SJ, Sritulanondha S, et al. Metal ion levels after metal-on-metal total hip arthroplasty: a five-year, prospective randomized trial. JBJS Am 2014;96:448–455. 5. Keegan GM, Learmonth ID, Case CP. Orthopaedic metals and their potential toxicity in the arthroplasty patient: A review of current knowledge and future strategies. JBJS Br 2007;89:567–573. 6. 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. JBJS Br 2010;92:469–476. 7. Mahendra G, Pandit H, Kliskey K, et al. Necrotic and inflammatory changes in metal-on-metal resurfacing hip arthroplasties. Acta Orthop 2009;80:653–659. 8. Azam MQ, McMahon S, Hawdon G, et al. Survivorship and clinical outcome of Birmingham hip resurfacing: a minimum ten years’ follow-up. Int Orthop 2016;40:1–7. 9. Gaillard MD, Gross TP. Metal-on-metal hip resurfacing in patients younger than 50 years: a retrospective analysis : 1285 cases, 12-year survivorship. JOSR 2017;12:79. 10. Lons A, Arnould A, Pommepuy T, et al. Excellent short-term results of hip resurfacing in a selected population of young patients. OTSR 2015;101:661–665. 11. Van Der Straeten C. Hip resurfacing arthroplasty in young patients: international high-volume centres’ report on the outcome of 11,382 metal-on-metal hip resurfacing arthroplasties in patients ⩽50 years at surgery. HIP Int 2020;112070002095735. 12. AOA. AOANJRR Annual Report 2008. 2008. 13. Smith AJ, Dieppe P, Howard PW, et al. Failure rates of metal-on-metal hip resurfacings: analysis of data from the National Joint Registry for England and Wales. Lancet 2012;380:1759–1766. 14. Johanson PE, Fenstad AM, Furnes O, et al. Inferior outcome after hip resurfacing arthroplasty than after conventional arthroplasty. Evidence from the Nordic Arthroplasty Register Association (NARA) database, 1995 to 2007. Acta Orthop 2010;81:535–541. 15. Stoney J, Graves SE, Steiger RN de, et al. Is the Survivorship of Birmingham Hip Resurfacing Better Than Selected Conventional Hip Arthroplasties in Men Younger Than 65 Years of Age? A Study from the Australian Orthopaedic Association National Joint Replacement Registry. Clin Orthop Relat Res 2020;478:2625–2636. 16. Wong JML, Liu YL, Graves S, et al. What Is the Rerevision Rate After Revising a Hip Resurfacing Arthroplasty? Analysis From the AOANJRR. CORR 2015;473:3458–3464. 17. Matharu GS, Pandit HG, Murray DW. Poor Survivorship and Frequent Complications at a Median of 10 Years After Metal-on-Metal Hip Resurfacing Revision. CORR 2017;475:304–314. 18. Matharu GS, Judge A, Murray DW, et al. Outcomes after Metal-on-metal Hip Revision Surgery Depend on the Reason for Failure: A Propensity Score-matched Study. CORR 2018;476:245–258. 19. LROI. LROI-Report 2019. 2019. 20. NOV. Metaal-op-Metaal (MoM) heupprothese, https://www.orthopeden.org/kwaliteit/kwaliteitsbeleid/mom-heupprothese (accessed 28 February 2020). 21. Lacny S, Wilson T, Clement F, et al. Kaplan-Meier Survival Analysis Overestimates the Risk of Revision Arthroplasty: A Meta-analysis. CORR 2015;473:3431–3442. 22. LROI. LROI-Report 2014: Arthroplasty in the Picture. 2015. 23. Netherlands Orthopaedic Association. Uitwerking NOV advies Metal op Metaal articulerende implantaten (MoM), 3 juni 2012 (in Dutch). 2012. 24. Verheyen CC, Verhaar JA. Failure rates of stemmed metal-on-metal hip replacements. Lancet 2012;380:105; author reply 106. 25. Kuijpers MFL, Hannink G, Steenbergen LN van, et al. Outcome of revision hip arthroplasty in patients younger than 55 years: an analysis of 1,037 revisions in the Dutch Arthroplasty Register. Acta Orthop 2020;1–6. 26. Peters RM, Steenbergen LN van, Stewart RE, et al. Patient Characteristics Influence Revision Rate of Total Hip Arthroplasty: American Society of Anesthesiologists Score and Body Mass Index Were the Strongest Predictors for Short-Term Revision After Primary Total Hip Arthroplasty. J Arthroplasty 2020;35:188-192.e2. 27. Steenbergen LN van, Denissen GA, Spooren A, et al. More than 95% completeness of reported procedures in the population-based Dutch Arthroplasty Register. Acta Orthop 2015;86:498–505. 28. AOA. AOANJRR Annual Report 2013. 2013. 29. NJR. NJR 10th Annual Report 2013. 2013.