OBJECTIVE Currently, there is a lack of evidence on the medium-term follow-up of
cementless total hip arthroplasty (THA) using metal on highly cross-linked
polyethylene (MoP), ceramic-on-ceramic (CoC), and metal-on-metal (MoM)
bearings. Our aim was to calculate the 5- to 10-year wear rate and the
incidence rate of osteolysis for 3 types of bearings.
METHODS A total of 77 patients underwent MoP, 105 underwent ceramic CoC, and 55 underwent MoM THAs. The average patient age at the time of surgery was 64.7, 55.9, and 59.9 years old in the MoP, CoC, and MoM bearings, respectively. Clinical and radiologic measurements at a mean follow-up of 7.6 years were analyzed.
RESULTS The mean postoperative Harris hip scores showed no difference among the groups. The mean annual liner wear rates were 0.0160, 0.0040 and 0.0054 mm/year in MoP, CoC bearings, and MoM bearings, with that of CoC bearings being significantly lower than the others. Osteolysis (14.5%) among MoM bearings was significantly more frequently observed compared with the others. Kaplan-Meier survival at 10 years with implant loosening, or revision THA as the endpoint, was 96.1% (95% confidence interval [CI]: 88.4–98.7) for MoP, 98.6% (95% CI: 90.3–98.6) for CoC bearing, and 98.2% (95% CI: 88.0–99.7) for MoM bearings (p = 0.360).
CONCLUSION Excellent clinical and radiological outcomes were obtained for MoP and CoC bearings.
Keywords: arthroplasty; ceramics; metals; polyethylenes; prosthesis design.
OBJETIVO Hoje, não há evidências sobre o acompanhamento em médio prazo da artroplastia
total do quadril (ATQ) não cimentada com componentes de polietileno de alta
ligação cruzada (highly cross-linked) sobre metal (MoP, na
sigla em inglês), de cerâmica sobre cerâmica (CoC, na sigla em inglês), e de
metal sobre metal (MoM, na sigla em inglês). Nosso objetivo foi calcular a
taxa de desgaste entre 5 a 10 anos e a taxa de incidência de osteólise nos 3
tipos de componentes.
MÉTODOS Um total de 77 pacientes foram submetidos a ATQ com componentes de MoP; 105 foram submetidos ao mesmo procedimento com componentes de CoC, e 55 foram submetidos à ATQ com componentes de MoM. A média de idade dos pacientes no momento da cirurgia foi de 64,7, 55,9 e 59,9 anos nos grupos de componentes MoP, de CoC, e de MoM, respectivamente. As medidas clínicas e radiológicas em um período médio de acompanhamento de 7,6 anos foram analisadas.
RESULTADOS As pontuações médias de Harris no período pós-operatório não mostraram diferença entre os grupos. As taxas médias anuais de desgaste do revestimento foram de 0,0160, 0,0040 e 0,0054 mm/ano em componentes de MoP, de CoC, e de MoP, respectivamente; a taxa de desgaste dos componentes de CoC foi significativamente menor do que as dos demais. A osteólise (14,5%) nos componentes de MoM foi significativamente mais frequente em comparação aos demais. A sobrevida de Kaplan-Meier aos 10 anos com afrouxamento do implante ou revisão da ATQ como desfecho foi de 96,1% (intervalo de confiança [IC] de 95%: 88,4–98,7) nos componentes de MoP, de 98,6% (IC95%: 90,3–98,6) nos componentes de CoC, e de 98,2% (IC95%: 88,0–99,7) nos componentes de MoM (p = 0,360).
CONCLUSÃO Os resultados clínicos e radiológicos de componentes de MoP e de CoC foram excelentes.
Palavras-chave: artroplastia; cerâmicas; metais; polietilenos; desenho de próteses.
|Citation: Higuchi Y, Seki T, Morita D, Komatsu D, Takegami Y, Ishiguro N. Comparison of Wear Rate between Ceramic-on-Ceramic, Metal on Highly Cross-linked Polyethylene, and Metal-on-Metal Bearings*. 54(3):295. doi:10.1055/s-0039-1691762|
|Note: * The present study was conducted at the Department of Orthopedic Surgery of the Nagoya University Graduate School of Medicine, Tsurumai-cho, Showa-ku Nagoya, Japan.|
|Received: January 19 2018; Accepted: July 10 2018|
Total hip arthroplasty (THA) for patients with end-stage osteoarthritis of the hip aims at ensuring long-lasting pain relief.1,2 However, despite improved implant designs and surgical techniques, aseptic loosening and osteolysis due to particulate debris, generated from conventional polyethylene, have become major limitations to long-term survival of prosthetics.3-6 Osteolysis has been reported in up to 60% of young, active patients with conventional polyethylene.6 In an attempt to avoid problems caused by wear-related debris, various types of bearing surfaces have been designed, such as metal on highly cross-linked polyethylene (MoP), a bearing-surface interaction with reduced linear and volumetric wear in comparison with conventional polyethylene.3,4,6,7 Similarly, hard bearing surfaces, which ensure even lower levels of volumetric wear, such as ceramic-on-ceramic (CoC) and metal-on-metal (MoM), have also been developed to address the problem of osteolysis.
However, they have their own inherent limitations, such as ceramic fractures and squeaking, or adverse local tissue reaction (ALTR).1,2,6,8 Over the last 2 decades, several systematic reviews and many clinical randomized controlled trials comparing the survivorship of THA implants with different bearing surfaces have been published. However, the results of these studies are inconsistent and continue to be debated.2,6,8-11
Therefore, the aim of our study was to perform a retrospective evaluation of the clinical and radiographic outcomes between MoP, CoC, and MoM cementless THAs over a 5- to 10-year follow-up period to compare the wear rates and the incidence of osteolysis.
Between August 2007 and December 2011, we performed consecutive primary CoC, MoP, or MoM THAs, with the type selected based on the recommendation of the senior surgeon performing the procedure. In total, 220 patients (237 THAs), who completed a minimum follow-up period of 5 years, were eligible for the present study. The exclusion criteria were: death from causes unrelated to surgery (one patient; one joint), patient inaccessibility (five patients moved abroad; five joints) and Crowe group III/IV hips (three patients; three joints). No significant differences were noted between the groups with regard to gender, body mass index (BMI), or diagnosis (► Table 1). However, the age of the patients at the time of the surgery for the group who received CoC THAs was significantly lower than that of the other groups. Informed consent was obtained from all of the patients, and the present study was approved by the ethics committee of our hospital.
|Parameters||MoP(n = 77)||CoC(n = 105)||MoM(n = 55)||p-value|
|Age (years old) (range)||64.7 ± 9.7 (29-76)||55.9 ± 8.7 (28-76)||59.9 ± 10.8 (33-74)||< 0.001*†|
|Body mass index (kg/m2) (range)||23.1 ± 3.2 (15.6-31.2)||23.0 ± 3.5 (16.7-31.9)||23.3 ± 3.4 (18.3-35.1)||0.849|
|Follow up (years) (range)||7.0 ± 1.0 (5-10)||7.0 ± 1.1 (5-10)||7.6 ± 1.1 (5-10)||0.466|
|Diagnosis (n, %)||0.769|
|Osteoarthritis||73 (94.8%)||93 (88.6%)||50 (90.9%)|
|Avascular necrosis||4 (5.2%)||11 (10.5%)||5 (5.1%)|
|Post-traumatic osteoarthritis||0||1 (0.9%)||0|
|Trident PSL||24 (31.2%)||22 (21.0%)||NA|
|TriAD HA||53 (68.8%)||74 (70.5%)||NA|
|Secur Fit AD||NA||9 (8.5%)||NA|
|Super Securfit HA||77 (100%)||105 (100%)||NA|
|G2 stem||NA||NA||55 (100%)|
|Fomoral head diameter|
|22 mm||7 (9.1%)||NA||0|
|26 mm||70 (91.9%)||NA||0|
|28 mm||0||67 (63.8)||28 (50.9%)|
|32 mm||0||38 (36.2)||NA|
|36 mm||0||0||27 (49.1%)|
A posterolateral approach was used for all procedures, with patients in the lateral decubitus position. The socket was fixed in the acetabulum using an acetabular alignment guide, with a target acetabular position of 20º of anteversion and 45º of inclination.12
Acetabular and Femoral Components
Acetabular and femoral components used for MoP and CoC THAs were manufactured by Stryker Orthopaedics (Mahwah, NJ, USA), while those for MoM were from Depuy Orthopaedics (Warsaw, IN, USA) (► Fig. 1 and ► Table 1). The Crossfire Polyethylene Insert (Stryker Orthopaedics) was used for all cases in which a highly cross-linked polyethylene liner was implanted, with the BIOLOX forte (Ceramtec, Plochingen, Germany) used as a ceramic liner and head. The diameter of the cobalt chrome head and of the ceramic head used is provided in ► Table 1.
The clinical data of the patients, including the Harris hip score (HHS), were prospectively recorded by a senior surgeon 1 month prior to the THA, at 6 months and 1 year post-THA, and then annually, up to the endpoint of the follow-up. These data were then retrospectively extracted from the medical records of the patients for analysis.
Radiographs of the hips were obtained in the standard anteroposterior (AP) view, with both hips in neutral rotation and 0º abduction, and in the Lauenstein (frog position) view, with the patient in the supine position and the hips in 45º abduction.7,13 Radiographs were obtained at 3 months postoperatively and then annually until the last follow-up.
Definite loosening of the femoral component was defined as a progressive axial subsidence of > 3 mm or a varus or valgus shift.14 Definite loosening of the acetabular component was defined as a change in the vertical and/or medial position of > 2 mm or as the presence of a continuous radiolucent line > 2 mm in width on both AP- and Lauenstein-view radiographs.15 Osteolysis was defined as areas of endosteal, intracortical or cancellous bone destruction of > 2 mm, which were non-linear and progressive.16 The acetabular inclination was measured using the transischial line as a reference, and anteversion was measured from digitized AP radiographs, using the method of Lewinnek et al.17 Using the methods of Dorr et al apud Barrack et al,18 the penetration of the femoral head into the liner was measured from digitized AP- and Lauenstein-view radiographs (► Fig. 2) by two surgeons using the Roman software, Version 1.70 (Institute of Orthopaedics, Oswestry, UK).
The femoral head penetration was measured at annual intervals to calculate the rate of wear, true wear and creep. The true wear represented the steady-state linear wear rate, while creep represented a measure of early head penetration, a feature attributable to the so-called “running-in” wear, in cases involving hard-on-hard bearings, or polyethylene creep and deformation in cases involving MoP.5 The wear rate was calculated by dividing total femoral head penetration at the end-point of observation by the number of years of follow-up. Linear regression models of femoral head penetration over time were constructed, with the y-intercept representing the creep rate or “running-in,” and the slope of the line representing the true wear rate.11,19 Intraclass correlation coefficients (ICCs) were calculated and used to determine interobserver reliability regarding the measurement of femoral head penetration. In the MoM group, patients underwent magnetic resonance imaging (MRI) when indicated, based on clinical symptoms, including hip pain or discomfort of the hip and concern for ALTR.
Statistical analyses were performed using SPSS Statistics for Windows, Version 21 (IBM Corp., Armonk, NY, USA) and EZR (Saitama Medical Center, Jichi Medical University, Tochigi, Japan). The differences between the groups were evaluated using analysis of variance (ANOVA), the Tukey test, the Pearson chi-squared test, the Holm-Bonferroni methods, and the log-rank test, as appropriate. A p-value < 0.05 was considered statistically significant.
The mean postoperative HHS was 85.3, 89.9, and 90.3 among cases involving MoP, CoC, and MoM THAs, respectively (p = 0.063) (► Table 2). Neither the incidence of dislocation nor the heterotopic bone formation differed among the three groups (► Table 3). All of the five dislocations were successfully treated conservatively via single closed reduction, without recurrence. All of the joints were classified as Brooker class 120 with the patients showing no symptoms. There was no ceramic fracture among cases involving CoC THAs (► Table 3).
|Parameters||MoP(n = 77)||CoC(n = 105)||MoM(n = 55)||p-value|
|Preoperative||57.0 ± 9.7||59.9 ± 9.2||61.2 ± 10.5||0.067|
|Last follow-up||87.0 ± 8.0||89.9 ± 7.3||90.3 ± 7.1||0.054|
|Acetabular component position|
|Anteversion (º)||20.7 ± 5.5||17.4 ± 6.9||19.6 ± 3.1||0.059|
|Inclination (º)||46.9 ± 5.5||45.9 ± 5.5||45.1 ± 3.1||0.158|
|Overall||2 (2.6%)||2 (1.9%)||8 (14.5%)||< 0.001*†|
|Acetabular||2 (2.6%)||1 (0.9%)||5 (9.1%)||0.016‡|
|Femoral||1 (1.3%)||2 (1.9%)||4 (7.3%)||0.143|
|Overall||2 (2.6%)||0||1 (1.8%)||1|
|Cup||1 (1.3%)||0||1 (1.8%)||1|
|Parameters||MoP(n = 77)||CoC(n = 105)||MoM(n = 55)||p-value|
|Revision THA||1 (1.3%)||1 (0.9%)||0||0.598|
|Deep joint infection||0||0||0||1|
|Dislocation||2 (2.6%)||2 (1.9%)||1 (1.8%)||0.797|
|DVT, PE||1 (1.3%)/0||0/0||0/0||0.556|
|Heterotopic bone||4 (5.2%)||2 (1.9%)||2 (3.6%)||0.138|
|Audible Squeaking||1 (0.9%)|
Abbreviation: CoC: ceramic-on-ceramic; DVT, Deep vein thrombosis; MoM: metal-on-metal; MoP: metal on highly cross-linked polyethylene; PE, Pulmonary embolism; THA, Total hip arthroplasty.
One joint among those having undergone MoP THAs (1.3%) required a revision THA owing to aseptic loosening of the cup. One joint (0.9%) among those having undergone CoC THAs required a revision THA owing to progressive femoral osteolysis caused by femoral neck impingement on the elevated acetabular metal liner. No cases of revision THA was observed among those involving MoM THAs.
One joint (0.9%) among those involving CoC THAs exhibited audible squeaking; however, a revision THA was not required because the condition was painless. No incidence of deep infection or of pulmonary embolism was identified in either group. One incidence of deep vein thrombosis (DVT) was identified among the cases involving MoP THAs; however, no DVT was observed in the CoC and MoM THA groups.
Based on the Kaplan-Meier survival curve analysis, considering implant loosening or revision THA as the endpoint, at 10 years, prosthetic survival was 96.1% (95% confidence interval [CI]: 90.0–99.3) for the MoP group, 98.6% (95% CI: 90.3–98.6) for the CoC group, and 98.2% (95% CI: 88.0–99.7) for the MoM THA group. These values were not significantly different among the groups (p = 0.360)
Neither the anteversion nor the inclination of the implanted acetabular component differed between the three THA groups (► Table 2). The incidence rate of osteolysis was greater for MoM THAs than either MoP (p < 0.05) or CoC (p < 0.001) THAs (► Table 3). Although no joints among CoC THAs showed aseptic loosening, 2 joints (2.6%) among MoP THAs showed loosening, with 1 of these cases (1.3%) requiring revision because of progression of loosening. In contrast, 1 joint (1.8%) among MoM THAs showed aseptic cup loosening (► Table 3). Magnetic resonance imaging was performed on a total of 10 hips (18.2%). A pseudocystic tumor was identified in four hips; as the symptoms were minimal, a revision THA was not required.
The Wear Rate
Femoral head penetration could not be measured in eight joints among patients with CoC THAs, and in two joints among patients with MoM THAs because the margin of the femoral head could not be identified clearly in these joints. Thus, measures of femoral head penetration were obtained for 97 and 53 joints among patients with CoC and MoM THAs, respectively. The ICC of measurement was 0.737 (95% CI: 0.16–0.982; p = 0.0043) for CoC, 0.876 (95% CI: 0.490–0.977; p = 0.002) for MoP, and 0.761 (95% CI: 0.203–0.958; p = 0.001) for MoM THAs.
The wear rates of CoC and MoM THAs were significantly lower than that of MoP THAs, irrespective of the radiographic view (AP and Lauenstein, p < 0.0001). In addition, the wear rate was significantly lower for CoC than for MoM THAs on AP-view radiographs, while the wear rate in the Lauenstein-view showed no difference between CoC and MoM THAs (► Table 4).
|Parameters||MoP(n = 77)||CoC(n = 105)||MoM(n = 55)||p-value|
|Wear rate (mm/year)|
|AP view||0.0161 ± 0.0051||0.0041 ± 0.0022||0.0054 ± 0.0019||< 0.001‡§¶|
|Lauenstein view||0.0158 ± 0.0061||0.0043 ± 0.0026||0.0053 ± 0.0020||< 0.001**††|
Abbreviation: AP, anteroposterior; CoC: ceramic-on-ceramic; MoM: metal-on-metal; MoP: metal on highly cross-linked polyethylene
Femoral head penetration could not be measured for 8 joints in the CoC group and for 2 joints in the MoM group, because the margin of the femoral head could not be identified clearly in these joints. The wear rates were measured for 97 joints in the CoC group and for 53 joints in the MoM group.
All of the values are given as mean and standard deviation (SD).
‡ CoC THAs vs MoM THAs: p < 0.05 (95% confidence interval [CI]: 0.00071-0.00292).
§ CoC THAs vs MoP THAs: p < 0.001 (95% CI: 0.01115-0.00137).
¶ MoM THAs vs MoP THAs: p < 0.001 (95% CI: 0.00934-0.01253).
** CoC THAs vs MoP THAs: p < 0.001 (95% CI: 0.00997-0.01268).
†† MoM THAs vs MoP THAs: p < 0.001 (95% CI: 0.00900-0.01221).
The running-in rate for CoC THAs was of 0.0036 mm/year based on AP radiographs, and of 0.0028 mm/year based on Lauenstein-view radiographs. The running-in rate for MoM THAs was of 0.0013 mm/year based on AP radiographs and of 0.0018 mm/year based on Lauenstein-view radiographs. Conversely, the creep rate for MoP THAs was of 0.0426 mm on AP-view radiographs and of 0.0440 mm on Lauenstein-view radiographs (► Fig. 3). The true wear rate for MoP THAs was of 0.0096 mm/year on AP-view radiographs and of 0.0093 mm/year on Lauenstein-view radiographs, while the rate for CoC was of 0.0037 mm/year on AP-view radiographs and of 0.0038 mm/year on Lauenstein-view radiographs. Conversely, the true wear rate for MoM THAs was of 0.0051 mm/year on AP-view radiographs and of 0.0050 mm/year on Lauenstein-view radiographs.
The main reason for failure of THA with conventional polyethylene is wear of the bearing surface and the resultant wear-induced osteolysis, which can result in implant loosening and failure.8,9,21 Therefore, modern materials with improved wear characteristics, such as metal, ceramics, and highly cross-linked polyethylene, have been implanted worldwide for use in THA.9,21-23
Previous studies have shown that survival with aseptic loosening at the midterm follow-up does not differ significantly between CoC and MoP THAs.6,10 However, few studies have compared midterm outcomes between MoP, CoC, and MoM THA, without a meta-analysis being available.24 In our study, we have found that the survival rates at the 10-year follow-up with regard to implant loosening or revision THA did not differ significantly between the three types of THAs. However, the risk of osteolysis was significantly higher with MoM than with CoC and MoP THAs. Therefore, the revision rate of MoM THA in long-term follow-up may be higher than that of the other THAs, because of gradually increasing osteolysis with the passage of time.25
Previous studies reported a mean liner wear rate of between 0.01 and 0.059 mm/year for MoP THAs, measured using AP radiographs.5,6,26,27 In contrast, the mean liner wear rate for CoC THAs was of between 0.00183 and 0.0067 mm/year based on AP radiographs.6,28 The in vitro wear rate of CoC and MoM components were of < 0.1 mm3/million cycles and of < 1.0 mm3/million cycles, respectively.29 Thus, our results are fairly consistent with those of previous reports in that the wear rate for CoC THAs was roughly 25% of the rate for MoP THAs. The wear rate for MoM THAs was ~ one-third of that for MoP THAs.
The true wear rate for MoP THAs was previously reported to be of 0.006 mm/year.19 However, to the best of our knowledge, the true wear rate for MoM THAs has not been reported thus far. The true wear rate of CoC THAs, as per our previous study, was ~ 50% of that of MoP THAs.
The overall THA survivorship was high for all of the groups, with no significant differences between the groups. The wear rates of MoM and CoC THAs were significantly lower than that of MoP THAs. However, MoM THAs exhibited more frequent osteolysis and demonstrated a significantly higher risk of revision compared with CoC and MoP THAs.24 Thus, excellent clinical and radiological outcomes were obtained for CoC and MoP THAs. Longer-term studies are needed to determine the effect of the wear rate and true wear rate on implant longevity.
The present study has some limitations. First, our study is retrospective in nature, with a relatively small number of patients. Also, the analysis was based on consecutive cases, with no randomization and no power analysis performed to determine if sufficient patients were included to detect clinically meaningful differences between groups. The patients who received CoC THAs were significantly younger than the patients in the other groups. This might have been caused by selection bias and might have affected the clinical and radiographic results. Second, we have measured femoral head penetration using the technique described by Dorr et al along with the computer-digitizer facilities of the Roman software, Version 1.70. The validity of this method for measuring femoral head penetration following CoC THAs has been previously reported.6,11,19 However, validation of MoM THA has never been reported. This method, which is typically used to examine femoral head penetration following MoP THA, is not easily applicable for CoC and MoM THAs. Third, we could not analyze the metal artifact reduction sequence MRI and serum metal ion levels for all of the patients with MoM THAs and could not identify an objective causative factor of osteolysis. Further studies are needed to clarify the association between bearing type and osteolysis. Finally, there were major differences in relation to the diameter of the prosthetic heads used in the three groups. In a hip simulator study, the wear rate of 36-mm ceramic bearings was significantly higher than that of 28-mm bearings, owing to their larger contact area, and paucity of lubrication when subjected to edge loading conditions.29 This suggests the possibility of increased wear with the use of large heads in certain circumstances.
Although CoC and MoM THAs had a lower wear rate than MoP THAs, the three types of THAs provided good implant survivorship. However, osteolysis among MoM THAs was more frequently observed compared with MoP and CoC THAs. Studies that span a longer follow-up period are needed to better understand the effects of wear rate and material bioactivity on implant longevity.