Introduction

Total hip arthroplasty (THA) is one of the most successful surgical procedures, with satisfactory clinical outcomes and survival rates of up to 90% over 15 to 20 years1,2. Despite the technical progress in THA, complications and failures still occur, including periprosthetic joint infection, instability, and aseptic loosening caused by polyethylene wear3,4.

Over the last decades, novel advances, such as the development of new implant designs and materials, have further decreased revision rates in THA5. In THA, different types of bearing surfaces can be divided into two major categories: hard-on-hard bearings (metal-on-metal, ceramic-on-metal, and ceramic-on-ceramic) and hard-on-soft bearings (metal-on-polyethylene, ceramic-on-polyethylene) are currently in use.

Metal-on-metal (MOM) surfaces have been employed in physically active young patients. However, MOM leads to a significant increase in metal ion concentrations in the body6. Furthermore, MOM may result in metal hypersensitivity and aseptic lymphocyte-dominated vasculitis6. Although no significant differences in functional and imaging outcomes were found for ceramic-on-metal bearings (COM) compared to MOM, chromium ion levels were significantly lower in the COM group after 3 years but increased after 5 years6. Cceramic-on-ceramic bearings provide good wear resistance, but this bearing carries an increased risk of liner and head breakage7. While ceramic-on-polyethene (COP) provides good mechanical properties and high resistance to scratch and deformation, it was criticised for its high cost and susceptibility to prosthetic head fracture8. Metal-on-polyethylene (MOP) are the most widely used THA bearings because of their relative safety and cost-effectiveness8. Low friction polymer materials such as polytetrafluoroethylene, poly 2-methacryloyloxyethyl phosphorylcholine, polycarbonate-urethanethe, polyether-ether-ketone, ultra-high molecular weight polyethylene (UHMWPE) and cross-linked polyethylene have outstanding mechanical properties and wear resistance9. Polyethylene is produced by the polymerization of ethylene, and a molecular weight of at least 1 million g/mole is defined as the standard10.

As UHMWPE used in THA liners may induce an imunological reaction to wear particles which can cause osteolysis and aseptic loosening, PE inlays have been cross-linked by thermal treatment (cross-linked polyethylene (XLPE), with improved wear resistance5.

Crosslinking is achieved by irradiating polyethylene at a dose higher than required for sterilisation, which is approximately 25 kGy11. Depending on the level of crosslinking and manufacturing process, several different cross-linked polyethylene types have been described. First-generation XLPE were irradiated with a dose between 50 and 100 KGy12. After initial promising results of XLPE, the cross-linking affected the mechanical properties of UHMWPE, compromising its toughness, ultimate mechanical properties, stiffness, and hardness. This was explained by the formation of free radicals during the manufacturing process, leading to oxidative changes in the XLPE with potentially decreased resistance to wear in the long term8. The free radicals in XLPE can be removed during the crosslinking process by annealing and remelting10. This procedure has led to moderately/modified cross-linked polyethylene (MXLPE) and highly cross-linked polyethylene (HXLPE), which mainly differ in the degree of irradiation (MXLPE: 50–75 kGy, HXLPE: 90 kGy), and thus the level of crosslinking. While remelted HXLPE (90 kGy) has good oxidation and wear resistance but poor fatigue properties, annealed HXLPE (90 kGy) shows good wear and fatigue resistance but has poor oxidation resistance13. Moderately cross-linked (50–75 kGy) and remelted UHMWPE (MXLPE) exhibits good oxidation resistance, with moderate wear and fatigue resistance14.

Kim et al. showed promising results of HXPLE liners in combination with delta ceramic heads, with annual wear rates of 0.022 mm/year15. However, despite zero revisions for wear-related problems and clinically nonsignificant wear rates, osteolysis is still observed in 35% of HXLPE THA in young patients at 16-year follow-up16.

Reduction of free radical production in liners is obtained by blending the liner with vitamin E (α-tocopherol). The infusion of Vitamin E chemically stabilizes the polyethylene by interrupting the oxidation cycle by decreasing the reactivity of reactive species17. However, the amount of Vitamin E that can be added is limited because Vitamin E in higher doses may interfere with the cross-linking process and thereby increase the wear rates17. Galea et al. showed significantly decreased wear of femoral heads (metal and ceramic) with vitamin E-diffused HXLPE compared to a moderately cross-linked and mechanically annealed UHMWPE in the first 5 years after THA18. However, the wear rates for both liners were very low (0.00–0.07 mm/year) and have not led to osteolysis or implant failures caused by aseptic loosening18. Comparable results were shown when polyethylene wear of MXLPE and HXLPE was compared with HXLPE-VEPE at 5 years19.

Given the large number of different types of liners, this network meta-analysis was conducted to compare the types of polyethylene liners and determine which is associated with the lowest wear penetration rates (mm/year) and with the lowest rate of revision.

The type of liners compared were the crosslinked ultra-high molecular weight polyethylene (CPE/UHMWPE), vitamin E infused highly cross-linked polyethylene (HXLPE-VEPE), modified cross-linked polyethylene (MXLPE), highly cross-linked polyethylene (HXLPE), cross-linked polyethylene (XLPE).

Methods

Eligibility criteria

All investigations that compared two or more types of polyethylene liners for THA published from January 2000 to July 2024 were accessed. Only studies that clearly stated the type of the liner were included. According to the Oxford Centre of Evidence-Based Medicine20, only clinical studies with levels I to III of evidence were considered. Articles in English and German language were eligible. Only studies that reported quantitative data under the endpoints of interest were considered.

Search strategy

This study followed the PRISMA extension statement for reporting systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations21. The PICOD algorithm was established:

  • P (Problem): End stage hip OA;

  • I (Intervention): THA;

  • C (Comparison): CPE/UHMWPE, HXLPE, HXLPE-VEPE, MXLPE, XLPE;

  • (Outcomes): Rate of revision surgery, wear penetration (mm/year)

  • D (Design): Comparative clinical investigations.

In July 2024, PubMed, Scopus, Embase, Google Scholar, and Cochrane databases were accessed. A time constraint was set from January 2000. Medical subject headings (MeSH) used for the database search are reported in the appendix.

Selection and data collection

Three authors (FM, FH, LS) performed the database search. The resulting titles were screened by hand, and if suitable, the abstract and the full text were accessed. If the full text was not accessible or available, the article was not included. The bibliography of the included studies was also screened by hand to identify additional studies. A third senior author (NM) solved disagreements.

Data items

Three authors (FM, FH, LS) performed data extraction. The following data at baseline were extracted: first author and year of publication and journal, length of the follow-up, number of patients with related mean age and BMI (kg/m2), number of women, side of surgery, and Harris hip score (HHS). The following data were collected at the last follow-up: inlay wear penetration per year (mm/year) and revision rate. Data were collected in Microsoft Office Excel version 16 (Microsoft Corp, Redmond, USA).

Assessment of the risk of bias and quality of the recommendations

The risk of bias was evaluated in accordance with guidelines in the Cochrane Handbook for Systematic Reviews of Interventions22. Two reviewers (FM & LS) evaluated the risk of bias in the extracted studies. Randomised controlled trials (RCTs) were evaluated using the revised Risk of Bias assessment tool (RoB2)22,23 of the Cochrane tool for assessing the Risk of Bias in randomised trials (RoB)24. Non-RCTs were evaluated using the Risk of Bias in Nonrandomised Studies of Interventions (ROBINS-I) tool25. The figure of the ROBINS-I was elaborated using the Robvis Software (Risk-of-bias VISualization, Riskofbias.info, Bristol, UK)26.

Synthesis methods

The main author (FM) performed the statistical analyses following the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions27. Mean and standard deviation were used for descriptive statistics. For baseline comparability, the IBM SPSS software was used. The sum of squares and mean of squares were evaluated. Comparability was assessed through the analysis of variance (ANOVA), with PANOVA > 0.1 considered satisfactory. The network meta-analyses were made through the STATA/MP software (Stata Corporation, College Station, Texas, USA). Only studies which clearly stated the nature of the type of polyethylene of the liner were included in the analyses. The analyses were performed using the STATA routine for Bayesian hierarchical random-effects model analysis. Continuous variables were analysed using the inverse variance method, using the standardised mean difference (SMD) effect measure. Binary data were analysed through the Mantel–Haenszel method, with the Log Odd Ratio (LOR) effect measure. Edge, interval, and funnel plots were performed and analysed. The overall transitivity, consistency, heterogeneity, and the size of the treatment effect of interest within-study variance were evaluated. The overall inconsistency was evaluated through the equation for global linearity via the Wald test. In PWald values > 0.05, the null hypothesis could not be rejected, and the consistency assumption could be accepted at the overall level of each treatment. Confidence and percentile intervals (CI a d PrI, respectively) were each set at 95%. Edge, interval, and funnel plots were performed. Egger’s test assessed plot asymmetry, with values of PEgger < 0.05 indicating statistically significant asymmetry. The Egger test is a linear regression of the intervention effect estimates on their standard errors weighted by their inverse variance.

Results

Study selection

The systematic literature search resulted in 1541 articles. Of them, 1061 were identified as duplicates and therefore excluded. After reviewing the abstracts, a further 384 articles were discarded because they did not match the defined eligibility criteria: study design (N = 191), low level of evidence (N = 104), not comparing two or more types of polyethylene liners (N = 41), not clearly stated the nature of the liner (N = 31), and language limitations (N = 17). A further 36 studies were excluded as they missed quantitative data under the outcomes of interests. In conclusion, 60 comparative studies were included in the present investigation: 16 RCTs, 23 prospective, and 21 retrospective clinical trials. 52 (83.3%) of these investigations compared CPE/UHMWPE liner with HXLPE liner. The results of the literature research are shown in Fig. 1.

Fig. 1
figure 1

PRISMA flow chart of the literature search.

Full size image

Risk of bias assessment

To investigate the risk of bias for RCTs included in the present meta-analysis, the revised Risk of Bias assessment tool (RoB2) was performed. 27% (16 of 60) of the studies reviewed were RCTs. Most authors reported high-quality allocation concealments, resulting in comparable study groups at baseline, leading to an almost low risk of bias arising from the randomisation process. Some concerns about deviations from the intended intervention, missing outcome data, and selection of the reported outcome were detected in a few studies, leading to a low to moderate risk of bias. Given the lack of blinded assessors to intervention status, a high risk of bias was identified in the outcome measurement in two of the included investigations. In summary, the risk of bias graph indicates a low to moderate quality of methodological assessment of RCTs (Fig. 2).

Fig. 2
figure 2

Cochrane risk of bias 2.0 tool (RoB2 tool).

Full size image

The ROBINS-I was applied to investigate the risk of bias of non-RCTs. 73% (44 of 60) of the included investigation were NRSIs. 23% (10 of 44) studies were rated as having a serious risk of bias in at least one domain, but no critical risk of bias in any domain. One study was identified with a critical risk of bias in the domain of participant selection, but all other domains had a low to moderate risk of bias. Given the mainly good methodological quality of the included studies, the overall risk of bias was low to moderate (Fig. 3).

Fig. 3
figure 3

The ROBINS-I of non-RCTs.

Full size image

Study characteristics

Data from 37,352 THAs were collected. 56% of patients were women. The mean patient age was 60.0 ± 6.6 years, the mean BMI was 27.5 ± 2.0 kg/m2. The mean length of follow-up was 81.6 ± 44.4 months. The generalities and demographic data of the included studies are shown in Table 1.

Table 1 Generalities and patient baseline data of the included studies.
Full size table

Baseline comparability

Between groups, baseline comparability in mean age, mean BMI, women:men ratio, side, length of the follow-up, and HHS was evidenced (Table 2).

Table 2 Baseline comparability.
Full size table

Synthesis of results

XLPE, followed by HXLPE, demonstrated the lowest wear penetration (Fig. 4). The equation of global linearity found no statistically significant inconsistency in all comparisons (PWald = 0.2). The Egger test found no statistically significant asymmetry (PEgger = 0.9).

Fig. 4
figure 4

From left to right: edge, funnel, and interval plots of the comparison: wear penetration (mm/year).

Full size image

XLPE, followed by HXLPE, demonstrated the lowest rate of revision at the last follow-up (Fig. 5). The equation of global linearity evidenced no statistically significant inconsistency in all comparisons (PWald = 0.9). The Egger test found no statistically significant asymmetry (PEgger = 0.07).

Fig. 5
figure 5

From left to right: edge, funnel, and interval plots of the comparison: revision.

Full size image

Discussion

The present study shows that XLPE followed by HXLPE demonstrated the lowest wear penetration and the lowest rate of revision after THA at a mean follow-up of 81.6 ± 44.4 months compared to other polyethylene liners (CPE/UHMWPE), HXLPE-VEPE, MXLPE).

Polyethylene is a complex material, and its morphological and mechanical properties are temporal and dependent on functional loading and environmental conditions85. UHMWPE is a linear (non-branching) semi-crystalline polymer, which can be described as a two-phase composite of crystalline and amorphous phases, which both influence the mechanical properties of the polymer85. While the crystalline phase provides modulus or stiffness to the material, the amorphous phase provides ductility and toughness13. Given its excellent wear resistance, high strength and biological inertness, UHMWPE remains the most commonly used bearing material in THA since its introduction in 19622,86,87. However, particulate wear and the consequent osteolysis related to its wear debris and delamination wear from oxidation reduced the longevity of the UHMWPE implants, with subsequent aseptic loosening, with the need for revision surgery13,86. In light of these increasing demands in revisions, the necessity to develop longer-lasting, more resilient formulations of UHMWPE was obvious85. Osteolysis is an inflammatory process induced by exposure to wear particles of UHMWPE, which is in part consequent to the oxidation process and is even more evident when using gamma sterilisation in air. This has led to introducing gamma irradiation in an inert environment, using ethylene oxide and gas plasma88, significantly decreasing the wear rates in conventional UHMWPE.

Additionally, gamma radiation can break the C–C bonds of the polyethylene chain and induce cross-linking, which can potentially increase wear resistance13,88. Therefore, in the late 1990s, cross-linked—UHMWPEs (XPLE) were proposed for joint arthroplasties13. The initially used sterilisation dose of 25–40 kGy of gamma radiation was increased up to 100 kGy with a linear correlation to the degree cross-linking obtained12,89. With higher doses of radiation, the cross-link density did not increase further, and the mechanical properties of UHMWPE were affected, compromising toughness, ultimate mechanical properties, stiffness, and hardness, mainly caused by possible formation of free radicals during the manufacturing process, leading to oxidative changes in the XLPE8. For this reason, all first-generation XLPE were irradiated with a dose between 50 and 100 KGy12. Improvements in XLPE were achieved by increasing the irradiation dose and removing the free radicals trapped in the crystalline phase89. Two thermal treatments were used to remove free radicals in XLPE: remelting or annealing89. With remelting, it is possible to remove all free radicals. However, the process decreases its mechanical properties. Annealing does not alter the mechanical properties significantly, but it cannot remove all free radicals, and the oxidative process continues during storage and in vivo after implantation13.

In general, remelted HXLPE (90 kGy) has good oxidation and wear resistance but poor fatigue properties, while annealed HXLPE (90 kGy) shows good wear and fatigue resistance but has poor oxidation resistance13. Moderately cross-linked (50–75 kGy) and remelted UHMWPE (MXLPE) exhibits good oxidation resistance, with moderate wear and fatigue resistance14.

Supporting our results, several studies on the in vivo performance of acetabular components have shown significantly reduced wear rates for HXLPE compared with conventional polyethylenes36,39,41,42,51,60,64,75,90. However, the highly crosslinked annealed UHMWPE formulations may also undergo oxidation in vivo, as was shown in irradiation sterilised conventional UHMWPE13. Several studies reported that remelted HXLPE show little or no in vivo oxidation91,92, while annealed HXLPE do undergo oxidation in vivo91,92,93, with maximum oxidative degradation near the rims91,92. Complications are also reported in remelted HXLPE, such as rim fractures after short-term implantation (6 months to 3.8 years)94,95,96.

A second generation of HXLPE was developed to improve fracture resistance to maintain the wear resistance of the first-generation HXLPE while retaining the superior mechanical properties of conventional UHMWPEs13.

Several methods have been developed: mechanical deformation97, high-pressure crystallisation after melting HXLPE98, sequential annealing99,100, and incorporation of vitamin E101,102,103. Vitamin E-infused highly cross-linked polyethylene (HXLPE-VEPE) should reduce free radical production and stabilise polyethylene by interrupting the oxidation cycle, thereby decreasing the reactivity of radical species17.

Oral et al. showed promising early in vitro results compared with irradiated UHMWPE101,102. HXLPE-VEPE showed some oxidation on the surface, which stayed constant thereafter,

UHMWPE exhibited substantial oxidation in the subsurface region, which increased over time102. The hip simulator wear rate of HXLPE-VEPE showed a fourfold to tenfold decrease from that of conventional UHMWPE101. Studies comparing HXLPE-VEPE with UHMWPE or MXLPE showed similar results78,79. Six-year results in a recent RCT including 199 patients reported superior results of vitamin E blended HXLPE (0.028 mm/year) compared with UHMWPE (0.035 mm/year)65. Significantly decreased wear rates of vitamin E-diffused HXLPE compared to a moderately cross-linked and mechanically annealed UHMWPE coupled with metal and ceramic femoral heads were shown within the first 5 years after THA18. However, both liners showed very low rates of wear (0.00–0.07 mm/year), with osteolysis or implant failure from aseptic loosening18. Similar results were shown when polyethylene wear of MXLPE and HXLPE was compared with HXLPE-VEPE over 5 years19. A prospective, randomised, controlled, multicenter study also compared the mid-term results of HXLPE with HXLPE-VEPE, with no significant differences between the two cohorts regarding wear rate (HXLPE: 23.2 μm/year vs HXLPE-VEPE: 24.0 μm/year, p = 0.73) after 5-years follow up34. The antioxidative benefit of vitamin E is expected to become evident in long-term follow-up. However, the results of this present network meta-analysis indicate better performance of XLPE and HXLPE after 7 years of follow-up. Skoldenberg reported a significantly higher total migration and continuous proximal migration of the component in the VEPE group compared to conventional argon-gas gamma-sterilized PE, with a difference at 2 years of a mean of 0.21 mm80.

This network meta-analysis has several limitations. First, there is no consistency in the liner/head coupling, and different implants were used in the various studies. Metal-on-polyethylene (MOP) is the most widely used THA bearings. At the same time, ceramic-on-polyethene (COP) may provide better mechanical properties, but has higher costs and is susceptible to femoral head fractures8. There was also a high level of heterogeneity between the studies regarding surgical approaches, which is considered a source of bias. Most studies did not report data separately according to the femoral head size or did not report information on the size of the head. Therefore, it was not possible to conduct additional analyses based on head sizes. Several techniques exist to investigate wear penetration (e.g. radiostereometry, Martell method, Polyware104,105). However, given the between-studies heterogeneity in these techniques, the analyses were not conducted separately according to each method. The mean follow-up of the included studies was 81.6 ± 44.4 months, allowing only short to midterm conclusions about the wear rates and revision rates of each liner. Although aseptic loosening caused by polyethylene wear is frequent3,4, there is a lack of large prospective long-term clinical trials. In addition, different study types were analysed: 16 RCTs, 23 prospective, and 21 retrospective clinical trials. 83% (52 of 60) of these investigations compared CPE/UHMWPE liner with HXLPE liner. Second-generation HXLPE stabilised with vitamin E is underrepresented. Additional studies should be performed to overcome the current limitations, with long-term trials comparing more types of liner and including new-generation polyethylenes.

Conclusion

XLPE and HXLPE liners in THA are associated with the lowest wear penetration (mm/year) and the lowest revision rate at approximately 7 years of follow-up.