Background: Oxidized zirconium (Oxinium) was introduced as an alternative bearing surface to cobalt-chromium (CoCr) in an attempt to reduce polyethylene wear and decrease aseptic mechanical failure of total knee replacements. While noncomparative reports have been described as promising, we were aware of no short or long-term clinical studies showing the superiority of Oxinium on polyethylene as a bearing surface. Using data from a comprehensive national joint replacement registry, we compared the long-term outcomes after cruciate-retaining total knee arthroplasty (TKA) with an Oxinium femoral component and those with the same prosthetic design but with a CoCr femoral component.
Methods: The cohorts consisted of 17,577 cemented Genesis-II cruciate-retaining total knee replacements using non-cross-linked polyethylene, which included 11,608 with CoCr femoral components and 5,969 with Oxinium femoral components. The cumulative percent revision and hazard ratio (HR) for revision risk were estimated for the cemented Genesis-II Oxinium and CoCr cruciate-retaining TKAs performed in Australia from September 1, 1999, to December 31, 2013. In addition, the revision diagnoses and the effects of age and patellar resurfacing were examined.
Results: No difference in the HR for revision risk was found between the Oxinium and CoCr cohorts for any age category for all causes of revision (HR = 0.92 [95% confidence interval (CI), 0.92 to 1.29]; p = 0.329), loosening or lysis, or aseptic causes, except for loosening or lysis in the group of patients who were ≥75 years old (p = 0.033). In these patients, TKA with Oxinium femoral components had a higher rate of revision. Younger patients preferentially received Oxinium femoral components. The revision risk was not affected by patellar resurfacing or nonresurfacing. At 12 years, the cumulative percent revision was 4.8% (95% CI, 4.2% to 5.4%) for the CoCr Genesis-II prosthesis compared with 7.7% (95% CI, 6.2% to 9.5%) for the Oxinium Genesis-II prosthesis.
Conclusions: In this cohort study involving the same prosthetic design, Oxinium femoral components did not reduce revision rates for all causes, loosening or lysis, or when infection as a cause of revision was removed compared with the same CoCr femoral component across all age groups including patients who were <55 years old. The cumulative percent revision was greater for the Oxinium components than for the CoCr components.
Level of Evidence: Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.
As the rates of revision total knee arthroplasty (TKA) are growing faster than the rates of primary TKA in some countries, such as the United States1, increasing the survivorship of TKA components is vital to improve outcomes and reduce costs. The most common early and intermediate causes of revision TKA are infection, instability, and malalignment, while the dominant cause over the long term is aseptic mechanical failure due to polyethylene wear and lysis or loosening2,3.
The bearing surface articulation in a TKA prosthesis results in polyethylene particle generation4. These submicrometer-sized wear particles are absorbed by macrophages, which respond by activating osteoclasts and depressing osteoblastic function at the prosthesis-bone interface. This altered macrophage-induced bone activity leads to osteolysis and component instability5.
Alternative bearing surfaces to cobalt-chromium (CoCr), such as oxidized zirconium (Oxinium; Smith & Nephew) have been introduced6-8 in an attempt to reduce polyethylene wear and hence decrease long-term TKA failure rates. While noncomparative reports have been described as promising9-15, we were aware of no short or long-term clinical studies that have shown the superiority of Oxinium on a polyethylene bearing surface. Published literature on comparative outcomes is sparse and limited to medium-term results7,16,17.
We investigated the long-term outcomes of Oxinium and CoCr total knee replacement in a cohort analysis of the same prosthetic design (Genesis II; Smith & Nephew). Using data from a large national joint replacement registry, we selected 2 cohorts of patients who received the same cruciate-retaining design, with the same method of fixation and polyethylene type, differing only in the femoral component bearing surfaces.
As younger patients may have preferentially received Oxinium femoral components, we also sought to remove the confounding variable of age by stratifying the cohorts into age groups and by different causes of revision. In addition, we removed the possible confounder of patellar resurfacing by comparing revision rates for the patients in each cohort who had and had not undergone patellar resurfacing.
The aim of the study was to compare the rates of revision for a single prosthetic design that had both Oxinium and CoCr options, using data from the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR). The rates of revision for loosening or lysis and for aseptic causes were also compared.
Materials and Methods
Data were obtained from the AOANJRR, collected from September 1, 1999, until December 31, 2013, for 2 cohorts: those who received a cemented Genesis-II cruciate-retaining TKA with a CoCr femoral component and those who received the same design of total knee replacement but with an Oxinium femoral component. The cohorts were otherwise identical with respect to fixation method, prosthetic design, constraint, and polyethylene type. All TKA components in both cohorts were implanted with non-cross-linked polyethylene, and patients managed with and without patellar surfacing were included in both cohorts. The study design with cohorts that had the same prosthesis aimed to eliminate confounders due to these prosthesis-related variables.
Data prospectively obtained at the time of surgery included patient demographics, side, diagnosis, hospital, type of procedure, and prosthesis-related details by catalog and lot number. The Registry represents a nearly complete national data set relating to knee arthroplasty in Australia3,18,19. Registry data are validated against patient-level data provided by each state and territory health departments with a sequential, multilevel matching process. A monthly matching program searches for all primary and revision arthroplasty procedures recorded in the Registry that involved the same side and joint in the same patient, linking each revision to the primary procedure. The National Death Index Data of the Department of Health and Aging are also matched biannually to obtain information on a patient’s date of death. Revision of a joint replacement is defined by the AOANJRR as any subsequent procedure that involves the insertion, removal, and/or replacement of a prosthesis or implant. The Registry records the reasons for revision and the type of TKA revision.
The outcomes for the cohorts were obtained and stratified into 4 age groups: <55 years, 55 to 64 years, 65 to 74 years, and ≥75 years. As younger patients may also have a higher risk of infection20, we investigated the outcome for all causes and then analyzed the outcome for loosening or lysis, as well as for all causes of revision excluding infection (aseptic revision). TKAs done without resurfacing the patella are known to have a higher risk of revision21,22; hence, analysis of the revision rates of both cohorts who had, and those who had not had, patellar resurfacing was also undertaken.
Kaplan-Meier estimates of survivorship were used to describe the time to first revision, with right-censoring for death or closure of the database at the time of analysis. The unadjusted cumulative percent revision after the primary arthroplasty, along with 95% confidence intervals (CIs), was calculated using unadjusted pointwise Greenwood estimates. The hazard ratio (HR), calculated using Cox proportional hazard models and adjusted for age and sex, was used to make statistical comparisons of the rate of revision between the cohorts. All tests were 2-tailed at the 5% level of significance, and p values were not adjusted for multiple testing. The analysis was performed using SAS software (version 9.3; SAS Institute).
The study was approved as a Declaration of Quality Assurance Activity by the Commonwealth of Australia under the National Health Insurance Act, 1973. All investigations were conducted in accordance with ethical principles of research (the Helsinki Declaration II).
During the study period, 17,577 Genesis-II cruciate-retaining TKAs with non-cross-linked polyethylene were performed with cement for all diagnoses, with 11,608 that had CoCr femoral components and 5,969 with Oxinium femoral components.
At 12 years, the cumulative percent revision was 4.8% (95% CI, 4.2% to 5.4%) for the CoCr Genesis-II group and 7.7% (95% CI, 6.2% to 9.5%) for the Oxinium Genesis-II group. There was no significant difference between Oxinium and CoCr with respect to the rate of revision (HR = 0.92 [95% CI, 0.92 to 1.29]; p = 0.329) (Fig. 1). No difference between Oxinium and CoCr components was found with respect to the rate of revision for patients with patellar resurfacing (Fig. 2) (HR = 0.98; 95% CI, 0.75 to 1.27) or without patellar resurfacing (HR = 0.88; 95% CI, 0.70 to 1.11).
As a percentage of primary procedures revised, loosening or lysis was the reason for revision in 1.1% of the knees in the Oxinium femoral component cohort (26.6% of the total number of revisions) and in 0.6% of the knees in the CoCr femoral component cohort (18.2% of the total number of revisions). Infection, loosening or lysis, patellofemoral pain, pain, and instability were the 5 most common causes of cumulative incidence of revision in both cohorts (Fig. 3). No difference was found between the cohorts with respect to the revision rate for loosening or lysis (Fig. 4) (HR = 0.87 [95% CI, 0.61 to 1.26]; p = 0.461), the revision rate for all causes of revision excluding infection (Fig. 5) (HR = 0.87 [95% CI, 0.71 to 1.06]; p = 0.155), and the revision rate for patients who were <55 years old (Fig. 6) (HR = 1.06 [95% CI, 0.68 to 1.66]; p = 0.798).
Oxinium was preferentially used in younger patients (Table I), with 63.8% of patients who were <55 years old receiving an Oxinium femoral component. An analysis stratified by age revealed that there was no difference between the Oxinium and CoCr TKA groups with respect to the rate of revision for any age category for all causes of revision, loosening or lysis, or aseptic causes (Table II), except for loosening or lysis in the group that was ≥75 years old (p = 0.033). In the patients who were ≥75 years old, Oxinium TKA components had a higher revision rate.
Oxinium implants are alloys of zirconium and niobium with the surface oxidized by thermal diffusion to transform into a layer of zirconium oxide ceramic approximately 5 μm thick12. The transformed Oxinium surface layer has a reported advantage over CoCr in resisting roughening, frictional behavior, and biocompatibility6. It has also been reported to have lower wear than CoCr in TKA simulated wear studies6,23,24 and femoral component retrieval studies25,26.
In this age-stratified matched-cohort registry study, Oxinium femoral components did not reduce revision rates for all causes, loosening or lysis, or when infection as a cause of revision was removed, compared with the same CoCr femoral component, across all age groups, including patients who were <55 years old. There was also no difference between cohorts with respect to revision rates when patellar resurfacing or nonresurfacing was examined.
By focusing on the cemented Genesis-II cruciate-retaining TKA prosthesis (which has one of the lowest cumulative percent revision rates for a cemented cruciate-retaining knee prosthesis in the AOANJRR at 13 years3 and high survivorship rates in other studies27), the possible confounding prosthesis-related variables, such as femoral component design, fixation method, polyethylene, and constraint, have been removed.
Patient age is a well-recognized risk factor for revision TKA, with higher revision rates in younger patients3,20,28,29, not only because of aseptic mechanical failure, such as loosening or lysis, but also because of infection and pain. It has been suggested that Oxinium be utilized in these younger patients because of proposed improved wear characteristics10. In this matched cohort, patients who were <55 years old received almost double the proportion of Oxinium, possibly in the belief that Oxinium may provide younger patients with a reduced rate of revision. Stratification by age removed this as a possible confounder7, and young patients with Oxinium and CoCr prostheses had similar outcomes.
Measures to reduce the rate of revision of TKA implants because of wear-related issues are important to examine, particularly as the burden of revision falls disproportionally more on younger patients30. Improved survivorship of TKA components has been linked to a number of factors, including cross-linked polyethylene3,18, improved polyethylene sterilization31, surgeon volume3,29, surgeon training29, computer navigation32, patient age20, patellar resurfacing22,33, and prosthesis design27,29,34.
Hui et al.17 reported no clinical or radiographic differences at 5 years in 80 patients with bilateral Genesis-II cruciate-retaining knee prostheses who had received both an Oxinium and a CoCr femoral component. Kim et al.16 found no clinical or radiographic difference, or difference in characteristics of the polyethylene wear particles, in 331 patients at 7.5 years after bilateral Oxinium and CoCr TKAs in a randomized controlled trial. Minoda et al.35 also showed no difference in polyethylene debris in knee aspirations from a small number of knees at 12 months after implantation of Oxinium and CoCr TKA prostheses. Inacio et al.7 recently reported that the short to intermediate-term survivorship of Oxinium on conventional polyethylene was not different from that for CoCr in a large registry-based series comparing a variety of alternative bearing surfaces with conventional surfaces in a heterogeneous cohort of implants. The results of these clinical trials support the findings of our registry study.
The clinical assessment of Oxinium as a potential means of reducing wear and associated clinical failure was justified by a number of preclinical studies. White et al.6 reported that Oxinium femoral components in a knee simulator had less polyethylene wear and no delamination in a 2-million-cycle test compared with the same CoCr total knee replacement. Ezzet et al.23 simulated polyethylene wear in athletically active patients with modestly malaligned Oxinium and CoCr TKA implants of the same design and reported that the mean gravimetric and volumetric wear rate after >5 million cycles was 55% lower in the oxidized zirconium group. The results of these simulator studies do not support the findings of the present study, and it remains uncertain why the improved wear characteristics of Oxinium do not result in a reduced revision rate. Essner et al.36 compared wear in a simulated stair-climbing load between the Genesis-II Oxinium TKA prosthesis and CoCr TKA components with a different femoral design. They reported a lower wear rate in the alternative prosthetic design and suggested that design geometry had the predominant effect on long-term wear rather than the materials chosen for the bearing. The results of that study support the hypothesis that prosthetic design is the first-order determinant of prosthesis survivorship, reinforcing the importance of arthroplasty registries.
Younger patients are at a reportedly higher risk of infection20 after TKA. Given that a higher proportion of younger patients received Oxinium implants, an age-stratified analysis of revision rates excluding infection as a cause of revision was also undertaken to remove infection as a revision confounder. The analysis revealed that there was no difference between the aseptic revision rates for the Oxinium and CoCr cohorts.
The only difference between the cohorts was the revision rate due to loosening or lysis in patients ≥75 years old, which was lower in the CoCr cohort. It remains unexplained why these older patients had a higher revision rate because of loosening or lysis with an Oxinium femoral component when the cumulative percent revision for all causes and for aseptic causes were the same; however, it should be noted that our significance testing was not corrected for multiple comparisons.
This study has numerous strengths, including a large number of procedures and the use of >12 years of population-based data, making it, to our knowledge, the largest study with the longest follow-up to date to describe the outcomes after Oxinium TKA. Different rates of patellar resurfacing between the cohorts may have biased the data; however, by comparing revision rates in patients with and without patellar resurfacing, this potential confounder was removed. It is also important to note that we excluded cross-linked polyethylene from the analysis, as its follow-up was only short term. To include cross-linked polyethylene would have resulted in a heterogeneous group of bearing surfaces. Whether cross-linked polyethylene with Oxinium components would result in lower wear rates than cross-linked polyethylene with CoCr components remains uncertain.
There are limitations to our study. While this age-stratified, matched prospective cohort study contains a large number of patients, a limitation is its observational design. Unrecognized selection bias may have occurred because of surgeons choosing the Oxinium implant for patients at higher risk, such as those who were obese, highly active, or with self-described metal allergies. The Registry collects Type-I and II data37, and information on comorbidities and prosthetic alignment is currently not collected. Surgeons with less surgical experience or lower volume may have preferentially chosen the Oxinium implant, resulting in a higher proportion of component malalignment and a higher wear rate, despite the possibly improved wear characteristics. In addition, we cannot comment on the comparative revision rates for Oxinium compared with CoCr components at >12 years.
In conclusion, at 12 years, there was no advantage to the use of Oxinium compared with CoCr in TKA by overall revision rate, revision for loosening or lysis, or for revision for non-infection-related causes.
Investigation performed at the Australian Orthopaedic Association National Joint Replacement Registry
A commentary by Michael S. Reich, MD, and Richard H. Walker, MD, is linked to the online version of this article at jbjs.org.
Disclosure: No funding was received specific to this study. On the Disclosure of Potential Conflicts of Interest forms, which are provided with the online version of the article, one or more of the authors checked “yes” to indicate that the author had a relevant financial relationship in the biomedical arena outside the submitted work.
- Copyright © 2017 by The Journal of Bone and Joint Surgery, Incorporated