Abstract

Objectives

This study investigated the biomechanical performance of decellularized porcine superflexor tendon (pSFT) grafts of varying diameters when utilized in conjunction with contemporary ACL graft fixation systems. This aimed to produce a range of ‘off-the-shelf’ products with predictable mechanical performance, depending on the individual requirements of the patient.

Decellularised extracellular matrix scaffolds show great promise for the regeneration of damaged musculoskeletal tissues (cartilage, ligament, meniscus), however, adequate fixation into the joint remains a challenge. Here, we assess the osseo-integration of decellularised porcine bone in a sheep model. This proof-of-concept study supports the overall objective to create composite decellularised tissue scaffolds with bony attachment sites to enable superior fixation and regeneration.

Porcine trabecular bone plugs (6mm diameter, 10mm long) were decellularised using a novel bioprocess incorporating low-concentration sodium dodecyl sulphate with protease inhibitors. Decellularised bone scaffolds (n=6) and ovine allograft controls (n=6) were implanted into the condyle of skeletally mature sheep for 4 and 12 weeks. New bone growth was visualised by oxytetracycline fluorescence and standard resin semi-quantitative histopathology.

Scaffolds were found to be fully decellularised and maintained the native microarchitecture. Following 4-week implantation in sheep, both scaffold and allograft appeared well integrated. The trabecular spaces of the scaffold were filled with a fibro-mesenchymal infiltrate, but some areas showed a marked focal lymphocytic response, associated with reduced bone deposition. A lesser lymphocytic response was observed in the allograft control. After 12-weeks the lymphocytic reaction was minimised in the scaffold and absent in allografts. The scaffold showed a higher density of new mineralized bone deposition compared to allograft. New marrow had formed in both the scaffold and allografts.

Following the demonstration of osteointegration this bioprocess can now be transferred to develop decellularised composite musculoskeletal tissue scaffolds and decellularised bone scaffolds for clinical regeneration of musculoskeletal tissues.

The concept of decellularised xenografts as a basis for anterior cruciate ligament (ACL) reconstruction was introduced to overcome limitations in alternative graft sources such as substantial remodelling delaying recovery and donor site morbidity. This study aimed to measure the biomechanical properties of decellularised porcine super flexor tendon (pSFT) processed to create ACL grafts of varying diameters, with a view to facilitating production of stratified ‘off the shelf’ products with specified functional properties for use in ACL reconstructive surgery.

Decellularisation was carried out using a previously established procedure, including antibiotic washes, low concentration detergent (0.1% sodium dodecyl sulphate) washes and nuclease treatments. Decellularised pSFTs were prepared to create double-bundle grafts of 7, 8 and 9mm diameter (n=6 in each group). Femoral and tibial fixations were simulated utilising Arthrex suspension devices (Tightrope®) and interference screws in bovine bone respectively.

Dynamic stiffness and creep were measured under cyclic loading between 50–250N for 1000 cycles at 1Hz. This was followed by ramp to failure at 200mm/min from which linear stiffness and load at failure were measured. Data were analysed using either 1- or 2-way ANOVA as appropriate with Tukey post-hoc analysis (p<0.05).

Significant differences were found between all groups for dynamic stiffness and between 7 & 9mm and 8 & 9mm groups for dynamic creep. Significant differences were also found between 7, 8 & 9mm groups for linear stiffness (167.8±4.9, 186.9±16.6 & 216.3±12.4N/mm respectively), but no significant differences were found between groups for load at failure (531.5±58.9, 604.1±183.3 & 627.9±72.4N respectively).

This study demonstrated that decellularised pSFTs possess comparable biomechanical properties to other ACL graft options (autografts and allografts). Furthermore, grafts can be stratified by their diameter to provide varying biomechanical profiles depending on the anatomy and individual needs of the recipient.

We have developed a decellularised porcine superflexor tendon (pSFT), which has shown promising regenerative capacity in an ovine model of anterior cruciate ligament (ACL) repair. This study investigated the strain rate dependent and dynamic mechanical properties of native and decellularised pSFTs.

Decellularisation was carried out using a previously established procedure, including antibiotic washes, low concentration detergent (0.1% sodium dodecyl sulphate) washes and nuclease treatments.

Three different strain rates were employed: 1, 10 & 100%s-1 (n=6 for all groups). Toe-region modulus (E0), linear-region modulus (E1), transition coordinates (εT, σT), tensile strength (UTS) and failure strain were calculated. For DMA, specimens were loaded between 1 & 5MPa with increasing frequency up to 2Hz. Dynamic (E*), storage (E') and loss (E'') moduli, and tan delta were calculated for native and decellularised groups (n=6). Data was analysed by 2-way ANOVA and Tukey post-hoc test (p<0.05).

For decellularised tendons, altering the strain rate did not affect any of the static tensile properties. For native pSFTs, the UTS, failure strain and E1 were not affected by changing the strain rate. Increasing the strain rate significantly increased E0 (1% vs 10% and 1% vs 100%) and σT (1% vs 100%) and decreased εT (1% vs 10% and 1% vs 100%) for native pSFT. E*, E' and E'' were all significantly reduced in decellularised specimens compared to native controls across all frequencies investigated. No significant differences were found for tan delta.

Evidence of strain rate dependency was witnessed in the native pSFTs by increase of the toe region modulus and displacements of the transition point coordinates. This response was not seen in the tissue following decellularisation. DMA demonstrated a reduction in dynamic, storage and loss moduli. Tan delta (E''/E') remained unchanged, indicating reductions in solid and fluid components are interlinked.

A pre-clinical experimental simulation model has been previously successfully developed, and was shown to have the potential for investigation of the biomechanical and tribological performance of early stage knee therapies. In order to investigate interventions that may necessitate sacrifice of the natural ligaments, it is necessary to replicate their function. This study investigated the most effective spring constraint conditions for the porcine knee model with the aim of replicating the natural ligament function.

The replication of natural ligament function was achieved through the use of physical springs in the anterior-posterior (AP) axis. Spring-9 (9 N/mm) and spring-20 (20 N/mm) were set at different free lengths in a natural knee simulator. The A/P displacement and shear force outputs from porcine knee samples (N=6) were measured and the most appropriate spring setting was determined by comparing the outputs at different spring settings with intact knee.

The A/P displacement of both spring-9 and spring-20 showed similar shapes to the all ligament control. Spring-9 with a free length of 4 mm and spring-20 with a free length of 5 mm showed minimal differences in A/P displacement output compared to the all ligament controls. There was no statistical difference between the two minimal differences either in A/P displacement or in shear force (paired t-test, p>0.05), which indicated that both conditions were appropriate spring constraint settings for the natural porcine knee model.

A porcine knee simulation model with refined spring constraint conditions was successfully developed in this study. Human knee model is currently under investigation using the methodology developed in porcine knee model, which will be more appropriate to investigate the effect of early stage knee therapies on the tribological function of the natural knee.

The ability to pre-clinically evaluate new cartilage substitution therapies in viable physiological biotribological models, such as the femoral-tibial joint would be advantageous. Methods for osteochondral (OC) plug culture have been developed and the aim of this study was to extend these methods to organ culture of whole femoral condylar and tibial osteochondral tissues.

Porcine femoral condyles and tibial plateau were aseptically dissected. The majority of cancellous bone was removed leaving intact cartilage and a layer of cortical bone. OC plugs were from porcine knee condyles. “Whole joint” tissues and OC plugs were cultured in defined medium and the viability of the cartilage at day 0, 8 or 14 days of culture assessed by XTT assay and LIVE/DEAD staining. Histological analysis (H&E; alcian blue staining) was used to determine cell number and visualise glycosominoglycans (GAGs). GAG levels were quantified in the cartilage using the dimethylene blue assay.

XTT conversion by OC plug cartilage reduced significantly between day 0 and day 8 with no further change between day 8 and 14. GAG levels did not change. “Whole joint” tissue behaved similarly with reduced XTT conversion between days 0 and 8 (femoral only) and days 0 and 14 (femoral and tibial). LIVE/DEAD staining showed the majority of cells remained alive in the mid and deep cartilage zones. There was a band of mainly dead cells in the surface zone, from day 0. There was no change in the GAG levels over the 14 day culture period.

In conclusion, large cuts of femoral and tibial osteochondral tissues were maintained in organ culture for extended periods. Surface zone chondrocytes rapidly lost membrane integrity ex-vivo whereas mid- and deep zone chondrocytes remained viable. It is hypothesised that physiological loading in a novel physically interactive bioreactor will improve the viability and will be the focus of future studies.

Acellular porcine super flexor tendon (pSFT) offers a promising solution to replacement of damaged anterior cruciate ligament [1]. It is desirable to package and terminally sterilise the acellular grafts to eliminate any possible harmful pathogens. However, irradiation techniques can damage the collagen ultra-structure and consequently reduce the mechanical properties [2]. The aims of this study were to investigate the effects of irradiation sterilisation of varying dosages on the biomechanical properties of the acellular pSFT.

Tendons were decellularised using a previously established protocol [1] and subjected to irradiation sterilisation using either 30 kGy gamma, 55 kGy gamma, 34 kGy E-beam, 15 kGy gamma, 15 kGy E-beam and (15+15) kGy E-beam (fractionated dose). Specimens then underwent stress relaxation and strength testing at 0 and 12 months post sterilisation to determine whether any effect on these properties was progressive. For stress relaxation testing, specimens were analysed using a Maxwell-Wiechert model. For strength testing, the ultimate tensile strength, Young's modulus and failure strain were assessed.

Significant differences were found which demonstrated that all irradiation treatments had an effect on the time-independent and time-dependent viscoelastic properties of irradiated tendons compared to per-acetic acid only treated controls. Interestingly, no significant differences were found between the irradiated groups. Similar trends were found for the strength testing properties. No significant differences were found between groups at 0 and 12 months.

Tendons retained sufficient biomechanical properties following sterilisation, however it was notable that there were no significant differences between the irradiated groups, as it was believed higher dosages would lead to a greater reduction in the mechanical properties. The changes observed were not altered further after 12 months storage, indicating the acellular pSFT graft has a stable shelf-life.

Introduction

Increased wear rates [1, 2] and acetabular rim fracture [3] of hip replacement bearings reported clinically have been associated with edge loading, which could occur due to rotational and/or translational mal-positioning [4]. Surgical mal-positioning can lead to dynamic microseparation mechanisms resulting in edge loading conditions. In vitro microseparation conditions have replicated stripe wear and the bi-modal wear debris distribution observed clinically [5, 6]. The aim of this study was to investigate the effect of steep cup inclination, representing rotational mal-positioning, on the magnitude of dynamic microseparation, severity of edge loading, and the resulting wear rate of a ceramic-on-ceramic bearing, under surgical translational mal-positioning conditions.

Tribology and wear of articular cartilage is associated with the mechanical properties, which are governed by the extracellular matrix (ECM). The ECM adapts to resist the loads and motions applied to the tissue. Most investigations take cartilage samples from quadrupeds, where the loading and motions are different to human. However, very few studies have investigated the differences between human and animal femoral head geometry and the mechanical properties of cartilage.

This study assessed the differences between human, porcine, ovine and bovine cartilage from the femoral head; in terms of anatomical geometry, thickness, equilibrium elastic modulus and permeability.

Diameter of porcine (3-6 months old), bovine (18-24 months old), ovine (4 years old) and human femoral heads were measured (n=6). Plugs taken out of the superior region of each femoral head and creep indentation was performed. The human femoral heads were obtained from surgery due to femoral neck fracture. Cartilage thickness was measured by monitoring the resistive force change as a needle traversed the cartilage and bone at a constant feed rate using a mechanical testing machine. The percentage deformation over time was determined by dividing deformation by thickness. A biphasic finite element model was used to obtain the intrinsic material properties of each plug. Data is presented as the mean ± 95% confidence limits. One-way ANOVA was used to test for significant differences (p < or = 0.05).

Significant differences in average femoral head diameter were observed between all animals, where bovine showed the largest femoral head. Human cartilage was found to be significantly thicker than cartilage from all quadrupedal hips. Human cartilage had a significantly larger equilibrium elastic modulus compared to porcine and bovine cartilage. Porcine articular cartilage was measured to be the most permeable which was significantly larger than all the other species. No significant difference in permeability was observed between human and the other two animals: bovine and ovine (Table 1).

The current study has shown that articular cartilage mechanical properties, thickness and geometry of the femoral heads differ significantly between different species. Therefore, it is necessary to consider these variations when choosing animal tissue to represent human.

Introduction

Articular hyaline cartilage has a unique structural composition that allows it to endure high load, distribute load to bone and enables low friction movement in joints. A novel acellular xenogenic graft is proposed as a biological cartilage replacement, for repair of osteochondral defects. Acellular porcine cartilage has been produced using repeated freeze thaw cycles and washing using hypotonic buffers and sodium dodecyl sulphate solution (SDS; Keir, 2008). DNA content of the acellular matrix was reduced by 93.3% compared to native cartilage as measured by nanodrop spectrophotometry of extracted DNA, with a corresponding reduction in glycosaminoglycan (GAG) content.

Introduction

The biological response to UHMWPE particles generated by total joint replacements is one of the key causes of osteolysis, which leads to late failure of implants. Particles ranging from 0.1-1.0μm have been shown to be the most biologically active, in terms of osteolytic cytokine release from macrophages [1]. Current designs of lumbar total disc replacements (TDR) contain UHMWPE as a bearing surface and the first reports of osteolysis around TDR in vivo have appeared recently in the literature [2]. The current wear testing standard (ISO18192-1) for TDR specifies only four degrees of freedom (4DOF), i.e. axial load, flexion-extension, lateral bend and axial rotation. However, Callaghan et al. [3] described a fifth DOF, anterior-posterior (AP) shear. The aim of this study was to investigate the effect that this additional AP shear load component had on the size and morphology of the wear particles generated by ProDisc-L TDR devices over five million cycles in a spine simulator.

Introduction: The aim of this study was to develop a technique to decellularise a porcine cartilagebone construct with a view to using this as a biological scaffold for transplantation into human osteochondral defect as a cartilage substitute.

Methods: Decellularisation was based on a modification of the technique of Booth et al (2002). Cartilage bone matrix (n=9) were decellularised by exposing the tissue to 2 cycles of dry freeze-thaw followed 2 more cycles with the addition of hypotonic (10mM tris-HCl, pH8.0) buffer. Samples were then cycled through hypotonic buffer, followed by ionic detergent (0.1% [w/v] sodium dodecyl sulphate [SDS]) in the presence of protease inhibitors (aprotinin 10 KIU/ml) and 0.1% (w/v) ethylene diamine tetraacetic acid (EDTA). This was followed by washes in PBS with aprotinin and incubation in nuclease solution containing DNase (50U/ml) and RNase (1U/ml). Decontamination using 0.1% (v/v) peracetic acid in PBS was then incorporated to achieve disinfection of the tissue samples. Finally, samples were washed in PBS. Three decellularisation protocols were used depending on the number of hypotonic/SDS cycles: this was either done once, three or six times referred to as DC1, DC3 and DC6 respectively. Fresh & decellularised cartilage were compared histologically using haematoxylin and eosin staining, to visualize cellular content, sirius red, to visualise collagen fibres & alcian blue, to visualise glycosaminoglycans (GAG). Immunohistochemistry staining for galactose-α-1,3-galactose (α-gal), collagen I, II & VI was performed for fresh and decellularised samples. DNA assay: Genomic DNA was extracted using a DNA isolation kit for tissues (Roche Applied Sciences). Collagen and DMB sulphated sugar assay, as described by Stapleton et al. (2008), were performed to measure collagen and GAG content. The biphasic property of fresh and decellularised cartilage was determined using a pin on plate indentation test.

Results: H& E staining revealed the absence of visible whole cells. Sirius red stain gave evidence of the retention of collagen following decellularisation. In contrast, the acellular matrix showed evidence of loss of GAGs. There was no evidence of the expression of α-gal in the acellular scaffold. DNA analysis revealed the absence of genomic DNA in comparison to fresh tissues (ANOVA, p< 0.05). The decellularisation process had minimal effect on the collagen content of the cartilage. Nevertheless there was a significant difference in the sulphated sugar content of the fresh tissue when compared to the decellularised tissue (ANOVA, p< 0.05), indicating loss of 92% GAG. Biomechanical testing of decellularised tissues showed a significant change (ANOVA, p< 0.05) in comparison to the fresh cartilage.

Discussion: In conclusion this study has generated data on the production of an acellular cartilage bone matrix scaffold for use in osteochondral defect repair. To our knowledge, this is the first study that has successfully removed whole cells and α-gal from xenogeneic cartilage and bone tissue. Future studies are required to investigate methods to recellularise the acellular matrix using an appropriate cell type and mechanical conditioning and to investigate replenishing GAG loss following decellularisation.

Introduction: Nanometre sized UHMWPE particles have recently been isolated from periprosthetic tissues and hip simulator lubricants [1,2]. The biological response to UHMWPE particles of 0.1 μm and above has been well characterised, with particles in the 0.1–1.0 μm size range having the highest biological activity [3]. The purpose of the study was to determine the biological activity of nanometre-sized particles in terms of osteolytic cytokine release from primary human monocytes.

Methods: Monocytes were isolated from peripheral blood from 5 healthy donors by density gradient centrifugation over Lymphoprep. Cells were cultured using the agarose gel technique [3] at particle volume (μm3):cell number ratios of 10:1 and 100:1. The particles used were:

1 Polystyrene FITC-conjugated FluoSpheres (FS; Invitrogen) in 20 nm, 40 nm, 0.2 μm and 1.0 μm sizes.

All particles were tested for the presence of endotoxin prior to culture with cells. Cells without particles were used as a negative control and 200 ng/ml LPS was used as a positive control. Cell viability was assessed using the ATP Lite assay (Perkin Elmer) and ELISA was used to determine TNF-alpha, IL-1beta, IL-6 and IL-8 release at 3, 6, 12 and 24 h.

Results: FluoSpheres and CD had no effect on cell viability at 10 or 100:1. Clinically relevant UHMWPE particles had no effect on cell viability at 10:1, however, at 100:1 significant differences (P< 0.05) were seen at 3, 12 and 24 h for Donors 1 and 3. The 40 nm, 0.2μm and 1.0 μm FS caused significant elevation of TNF-α release at the 12 and 24 h time points at 100:1. There was no significant increase in TNF-α release for the 20 nm FS (3/5 donors). Particle volume and particle size showed correlation with cellular response, with the 20 nm FS showing the lowest biological activity. Clinically relevant UHMWPE particles and nanometre sized CD produced significantly higher quantities of TNF-alpha at 100:1. Release of interleukins IL-1beta, IL-6 and IL-8 followed a similar trend to TNF-alpha release.

Discussion: This study found that all nanometre-sized particles had the potential to provoke inflammatory cytokine release from macrophages. Particle volume and particle size played critical roles in initiating cellular responses. There was a lower particle size limit, with the 20 nm FS showing the lowest activity. Nanometre-sized polyethylene particles (CD) caused elevated TNF-α release, and since it has been shown that nanometre-sized UHMWPE particles are produced in large numbers in vivo [2], the relative contribution of these particles to osteolysis should be considered. The biological response to nanometre-sized clinically relevant UHMWPE particles is currently under investigation.

Total meniscectomy has been shown to induce osteoarthritic changes in the underlying articular cartilage(AC) and bone in the natural knee (Fairbank 1948; McDermott 2006). This indicates the meniscus plays an important protective role, providing joint congruity and distributing contact forces, hence reducing contact stress. However, no friction and wear studies have been performed on meniscectomy. The aim of this study was to study the tribological response of the medial compartmental natural knee with and without the intact meniscus, under physiological dynamic loading and motion. The effect of normal and reduced loading was investigated.

Eighteen month old bovine medial compartmental knees were used. A pendulum friction simulator (Simulation Solutions, UK) was used to apply a dynamic axial loads with peak loads of 1000N (normal) and 260N (reduced). Flexion-extension of amplitude 23degrees was applied and the experiments ran for 3600 cycles at 1Hz. Lubricant was 25% bovine serum in saline. A 9.4 Tesla MRI (Bruker) scanner and Analyze software (Mayo Clinic, US) were used to calculate wear volumes. A surface profilometer (Talysurf, Taylor-Hobson, UK) was used to measure the surface roughness of the specimen before and after the test.

Coefficient of friction was found to increase with increased loading, with and without meniscus. With meniscus intact, no wear was found on AC and contact stresses were 4.9MPa and 2.8MPa, for normal and reduced loading respectively. On removal of meniscus, friction was higher at both loading conditions and surface fibrillation found on some of the AC surfaces. Contact stresses rose to 17.2MPa and 8.6MPa for normal and reduced loading.

This study has shown for the first time, the direct elevation of the coefficient of friction, immediate surface fibrillation and biomechanical wear of AC upon removal of the meniscus. On removal of meniscus, peak stresses rose and surface damage occurred on AC surfaces. The removal of the meniscus means forces act across smaller areas and contact stresses are increased. Wear is increased due to the subsequent increase in direct solid-solid contact and loss of fluid support due to the unique biphasic nature of AC. This further supports retaining meniscus whenever possible in knee joint surgery.

Nanometre-sized particles of ultra-high molecular weight polyethylene have been identified in the lubricants retrieved from hip simulators. Tissue samples were taken from seven failed Charnley total hip replacements, digested using strong alkali and analysed using high-resolution field emission gun-scanning electron microscopy to determine whether nanometre-sized particles of polyethylene debris were generated in vivo. A randomised method of analysis was used to quantify and characterise all the polyethylene particles isolated.

We isolated nanometre-sized particles from the retrieved tissue samples. The smallest identified was 30 nm and the majority were in the 0.1 μm to 0.99 μm size range. Particles in the 1.0 μm to 9.99 μm size range represented the highest proportion of the wear volume of the tissue samples, with 35% to 98% of the total wear volume comprised of particles of this size. The number of nanometre-sized particles isolated from the tissues accounted for only a small proportion of the total wear volume. Further work is required to assess the biological response to nanometre-sized polyethylene particles.

Considerable differences in kinematics between different designs of knee prostheses and compared to the natural knee have been seen in vivo. Most noticeably, lift off of the femoral condyles from the tibial insert has been observed in many patients. The aim of this study was to simulate lateral femoral condylar lift off in vitro and to compare the wear of fixed bearing knee prostheses with and without lift off.

Twelve PFC Sigma cruciate retaining fixed bearing knees (DePuy, Leeds, UK) were tested using six station simulators (Prosim, Manchester, UK). The kinematic input conditions were femoral axis loading (maximum 2.6 kN), flexion-extension (0–58°), internal/external rotation (±5°) and anterior/posterior displacement (0–5 mm). Six knees were tested under these standard conditions for 4 million cycles. Six knees were tested under these conditions with the addition of lateral femoral condylar lift off, for 5 million cycles. The lubricant used was 25% newborn calf serum. Wear of the inserts was determined gravimetrically.

Under the standard kinematic conditions the mean wear rate with 95% confidence limits was 8.8 ± 4.8 mm 3/million cycles. When femoral condylar lift off was simulated the mean wear rate increased to 16.4 ± 2.9mm 3/million cycles, which was statistically significantly higher (p < 0.01, Students t-test). The wear patterns on the femoral articulating surface of all the inserts showed more burnishing wear on the medial condyle than the lateral. However, in the simulation of lift off the medial condyle was more aggressively worn with evidence of adhesion and surface defects.

The presence of lateral femoral condylar lift off accelerated the wear of PFC Sigma cruciate retaining fixed bearing knees. The lateral lift off produced uneven loading of the bearing, resulting in elevated contact stresses and hence more wear damage to the medial side of the insert. The implications of condylar lift off include increased wear of the polyethylene and possible osteolysis.

Different wear rates have been reported for ceramic-on-ceramic (COC) and metal-on-metal (MOM) hip replacements tested in simulators with different loading conditions and lubricants. We postulate that differences in wear rates may be associated with changes in lubrication and friction in the joint. This study aimed to compare the friction of COC and MOM bearings under different lubrication regimes, simulated by varying swing-phase loads and lubricants.

Alumina COC and CoCr MOM 28mm-diameter bearings were studied in a pendulum friction simulator. Flexion-extension of +/−25 degrees was applied to the head, a peak load of 2kN and swing-phase loads of 25N,100N, 300N used. Lubricants used included water, 25% and 100%-bovine serum.

COC and MOM bearings showed increased friction as the swing-phase load increased. COC bearings produced higher friction in 100%-serum compared to 25%-serum. In contrast, friction was lower when MOM bearings were tested in 100%-serum compared to 25%-serum. When COC bearings were tested in water, the friction decreased in comparison to testing in serum, however, MOM friction was higher in water.

Increasing the swing-phase load reduced the thickness of the fluid-film in the stance-phase and this increased friction. The increase in friction when COC bearings were tested in 100%-serum (compared to 25%) may be due to the increased forces required to shear the increased concentration of proteins, similarly friction is reduced in water. MOM bearing friction was reduced in 100%-serum, in this instance increased proteins may be acting as solid-phase lubricants, and similarly MOM friction increased in water.

Introduction: Cobalt-chrome particles from metal hip implants can accumulate in the liver, spleen, lymph nodes and bone marrow of patients. This is a concern as studies have reported neoplastic changes in cells of patients with metal implants. The aims of this study were to determine the effect of wear particles generated by metal-on-metal and ceramic-on-metal implants from hip simulations upon the viability of L929 cells and to determine their genotoxic potential when cultured with primary human fibroblasts.

Methods: Particles were generated in a 10 station Prosim hip simulator run with water as lubricant under microseparation and standard conditions. Bearings comprised medical grade HIPed ‘BIOLOX Forte’ alumina ceramic femoral heads against Ultima metal CoCr acetabular cups (CoM) and wrought CoCr alloy ASTM F1537 femoral heads and acetabular cups (MoM). Particles were sterilised at 1800C for 4 hours and cultured with L929 fibroblasts at particle volume(μm3):cell number ratios of 500:1, 100:1, 50:1, 5:1, 0.5:1, 0.05:1, 0.005:1 and 0.0005:1. Camptothecin (1 and 2μg.ml-1) and latex beads (100μm3 per cell) were used as positive and negative controls. Cultures were for 0, 1, 2, 3, 4 and 5 days at 37oC in 5%(v/v) CO2 in air. Cell viability was assessed using the ATPlite assay. Sterile particles were cultured with primary human fibroblasts at particle volume (μm3):cell number ratios of 50:1, 5:1 and 0.5:1. Cells were exposed to 30%(v/v) H2O2 (positive control) and latex beads (50μm3 per cell; negative control). Cells were cultured for 24 hours and 5 days at 37oC in 5%(v/v) CO2 in air. Genotoxicity was assessed using the comet assay. Statistical analysis between the cell-only negative controls and the cells with the particles at various concentrations, were determined by ANOVA and calculating the minimum significant difference (MSD;p< 0.05) using the T-method.

Results: Particle volume(μm3):cell ratios of 500:1, 100:1 and 50:1 caused a significant decrease in cell viability over 5 days. Wear particles from MoM implants under microseparation wear conditions were also significantly reduced viability at particle volume(μm3):cell ratios of 5:1 over 5 days. Particles from MoM implants under standard wear conditions and CoM implants under both wear conditions resulted in increases in tail length and tail moment relative to the cells only negative control for all treatment groups after 24 hours. These decreased by day 5. Tail length and tail moment were increased at 24 hours relative to day 5 for each of the three particle types. Particles generated by MoM implants under microseparation conditions had different effects upon cells. Tail lengths increased between days 1and 5 for all particle concentrations. A significant increase in tail moments between days 1 and 5 was recorded.

Discussion: This study has shown that metal particles can cause cytotoxic effects and immediate DNA damage to fibroblasts in vitro. Particles were found to reduce cell viability over 5 days and this may account for the decreases in tail length and moments between 1 and 5 days for three particle types. This is of concern as MoM and CoM implants are designed to be implanted into young patients and, despite their low wear rates generate circa 10¹³ particles per mm3 of wear.

Introduction: Joint replacement surgery is one of the most common operations that take place in United Kingdom. The major problem in total hip arthroplasty is the generation of particulate wear debris and the subsequent biological responses. Wear debris induces osteolysis and a subsequent failure of the implant that lead to the liberation of greater quantities of particulate and soluble debris to bone marrow, blood, lymph nodes, liver and spleen. Recently, it has been suggested that these adverse effects depend not only on the chemical composition but also on the particulate nature of the material (size and shape). Particle size has been shown to influence the inflammatory response of macrophages to wear debris. This study evaluated whether particle size also influences the viability and mutagenic damage.

Methods: Cobalt chrome alloy particles of two sizes (large 2.9±1.1μm, small 0.07±0.04 μm) were generated and characterised by Scanning Electron Microscopy. Different concentrations of particles were added to primary human fibroblasts in tissue culture. The release of cytokines in the medium was assayed by Enzyme-Linked ImunnoSorbent Assay (ELISA). Cell viability was determined by MTT conversion and the degree of DNA damage was quantitatively analysed by the Alkaline Single Cell Gel Electrophoresis (COMET) assay with image analysis.

Results: Small particles initialise DNA damage at much lower volumetric concentrations (0.05 and 0.5 μm³/cell) than larger particles (500 μm³/cell). The difference in the doses was approximately related to the difference in surface area of the particles. DNA damage was related to a delayed decrease in cell viability, which was noted after three days of exposure.

In contrast, the release of the inflammatory cytokine TNF-α and the multifunctional growth factor TGF-β-2 occurred at lower doses (0.0005 to 5 μm³/cell for TNF-α and 0.5 to 50 μm³/cell for TGF-β-2). No release of IL-6 was detected at any dose. Only growth factor FGF-23 was increased in similar pattern to the DNA damage.

Conclusions: This study has demonstrated important differences between the mutagenicity, toxicity and inflammatory potential of small (nanometre sized) and large (micrometer sized) chrome particles.

Introduction: In vivo fluoroscopic studies have shown considerable differences in kinematics between different designs of knee prostheses and compared to the natural knee. Most noticeably, lift off of the femoral condyles from the tibial insert has been observed in many patients (Dennis et al, 2003). The aim of this study was to simulate lateral femoral condylar lift off in vitro and to compare the wear of fixed bearing knee prostheses with and without lift off.

Materials and Methods: 12 PFC Sigma cruciate retaining fixed bearing knees (DePuy, Leeds, UK) were tested. The 10 mm thick inserts were manufactured from GUR1020 UHMWPE and gamma irradiated in a vacuum. The inserts snap fitted into titanium alloy tibial trays, and articulated against Co-Cr-Mo alloy femoral components. The testing was carried out on six station simulators (Prosim, Manchester, UK). Femoral axis loading (maximum 2.6 kN) and the flex-ion-extension profile (0–58°) were adopted from ISO 14243 (1999). The internal/external rotation was ± 5° and anterior/ posterior displacement 0–5 mm. Six of the knees were tested under these standard conditions for 4 million cycles. A further six knees were tested under these conditions with the addition of lateral femoral condylar lift off, for 5 million cycles. The lift off was achieved by introducing an adduction moment to the tibial carriage, producing a separation of approximately 1 mm during the swing phase of the simulator cycle. The simulator was run at 1 Hz and the lubricant used was 25% newborn calf serum. Wear was determined gravimetrically, using unloaded soak controls to adjust for moisture uptake. Statistical analysis was performed using Students t-test (p < 0.05).

Results: Under the standard kinematic conditions the mean wear rate with 95% confidence limits was 8.8 ± 4.8 mm3/million cycles. When femoral condylar lift off was simulated the mean wear rate increased to 16.2 ± 2.9 mm3/million cycles, which was statistically significantly higher (p < 0.01). The wear patterns on the femoral articulating surface of all the inserts showed more burnishing wear on the medial condyle than the lateral. However, in the simulation of lift off the medial condyle was even more aggressively worn with evidence of adhesion and surface defects.

Discussion: The presence of lateral femoral condylar lift off resulted in a higher wear rate on the medial compartment of the PFC Sigma fixed bearing knee. This could be due to elevated contact stresses as the lateral lift off produced uneven loading of the bearing. Further, additional medial/lateral sliding of the medial condyle whilst it remained in contact may have accelerated the wear by cross shearing of the polyethylene in the medial/lateral direction. This direction is weakened when the polyethylene is preferentially molecularly orientated by sliding in the flexion-extension axis. The implications of condylar lift off include premature wear of the polyethylene and possible component loosening.

Introduction: Crosslinking has been extensively introduced to reduce the wear of UHMWPE. Zero wear of highly crosslinked UHMWPE has been reported by some groups (1) in hip simulators, clinical studies have reported finite wear rates (2). The aim of this study was to compare the wear rates produced by UHMWPE with different levels of crosslinking.

Materials and Methods: Studies were carried out using 28mm diameter cobalt chrome femoral heads. These were articulated against UHMWPE in the Leeds ProSim hip joint simulator. The acetabular cups were manufactured from UHMWPE GUR 1050. The GUR 1050 was highly crosslinked with 10MRad or 7.5MRad of gamma irradiation in nitrogen followed by re-melting at a temperature above 150°C. Slightly crosslinked GUR 1050 was also tested (gamma irradiated with 2.5MRad in air). Non-crosslinked GUR 1050 UHMWPE was used as a control. Five cups of the materials were tested with one station from each set of five being used for creep data. Wear measurements were taken every million cycles using a coordinate measuring machine and tests were run to 5 million cycles. The tests were carried out in low serum concentrations of 25% (v/v) bovine serum diluted with 0.1% (w/v) sodium azide in water. At each million cycles a 3D measurement was taken of the contact region of the acetabular cups using a Form Talysurf profilometer.

Results and Discussion: The wear rate decreased as crosslinking levels increased. The non-crosslinked material had an overall average wear (mm3/million cycles) determined by volume change of 45.6+/−1.35, the 2.5MRad material 46.9+/−9.4, the 7.5MRad 15.04+/−4.28 and the 10MRad material 8.7+/−3.11. The intentionally cross-linked materials showed a significantly lower volume change than the other two materials, with the 10MRad polyethylene having a slightly lower volume change than the 7.5MRad polyethylene. All four polyethylenes showed greater volume change in the first million cycles than the subsequent four and this was associated with initial creep deformation in the first million cycles. The individual creep deformation cups confirmed this with volume changes in the first million cycles followed by stability. Creep volumes of between 10 and 25 mm3 total were measured with the lowest value being for the 10MRad polyethylene. The steady state wear rates for the PE’s between one and five million cycles were 0MRad 36.9+/−1.92 mm3/million cycles, 2.5MRad 44.12+/−10.09, 7.5MRad 7.89+/−2.32 and 10MRad 4.62+/−2.73. The results of the surface topography of the acetabular cups showed that the highly crosslinked materials became smoother than the other materials as the test progressed. This would benefit the crosslinked materials in aiding lubrication and could have contributed to the lower wear rate seen with these materials.

Conclusion: The highly crosslinked UHMWPE gave lower wear volumes than the noncrosslinked materials. This could have been due to the smoother surfaces of the cups as the study progressed which resulted in better lubrication of the components. Finite wear rates have been recorded for the first time with highly cross-linked polyethylene, that compare with clinical observations.

Viscosupplementation is the current treatment modality for early stage arthritis and in some cases for delaying joint replacement procedures. Rheological properties similar to that of synovial fluid and high molecular weight have been recognized as the determining factors in hyaluronic acids (HA) therapeutic and analgesic value (1). In this study, the self assembly of peptides into beta-sheet structures in solution (2–4) is explored to develop novel biocompatible injectable joint lubricants. These peptides can be delivered into the joint easily in their low viscosity monomer form, while they are designed to self-assemble in situ under physiological conditions. Four different peptides P11-4, P11-8, P11-9, and P11-12 were designed based on the chemical motif of hyaluronic acid and were found to self-assemble into nematic fluids and gels under physiological conditions. Friction characteristics of these peptides as lubricants were evaluated in a bovine cartilage on cartilage model using a simple pin on plate geometry and under various sliding conditions. Friction tests were carried out using both healthy and damaged bovine cartilage samples, to study the therapeutic effect of these peptides as lubricants. Further, a rheometer with cone-on-plate configuration was used to study these peptides in shear viscosity and oscillatory shear modes to determine their viscoelastic properties. Both the friction properties and rheological behaviour of the peptides were compared to that of a commercially available hyaluronic acid preparation that was tested along with the peptides. Peptide P11-9 was found to have very similar viscoelastic properties to that of HA, and was also the most effective in friction level reduction among the four peptides tested. When compared to HA, P11-9 showed slightly better friction characteristics in all the healthy cartilage models, while HA was the best lubricant in damaged cartilage models when compared to P11-9 and other peptides. The results indicate that these novel self assembling peptides can be developed as a new generation of synthetic viscosupplements for the treatment of early stage arthritis.

There is currently much interest in the wear of metal-on-metal THRs and potential concerns about elevated metal ion levels. Generally, wear of metal-on-metal THR’s has been low in simulator studies. Slightly higher and more variable wear has been found clinically. Variations in surgical approach, technique and fixation method may influence the level of force applied across the prosthesis during gait. It is hypothesised that increased joint tensioning may increase loading of THR’s during the swing-phase; leading to elevated wear and friction due to depleted fluid film lubrication. This study aimed to assess the effect of swing-phase load on the friction, lubrication and wear of metal-on-metal THR’s.

Cobalt-chrome 28mm metal-on-metal THR’s were tested in a physiological hip simulator, loading was modified to provide; (1) ISO swing-phase load (280N, as per ISO 14242-1) and (2) low swing-phase load (< 100N). Friction testing was conducted using a pendulum friction simulator, with 280N and 100N swing-phase loads. Theoretical lubrication modelling was carried out using elastohydrodynamic lubrication theory.

The overall mean volumetric wear rates was 10-times greater in THR’s tested with an ISO swing-phase load in comparison to THR’s tested with low swing-phase loads (0.58±0.49 compared to 0.06±0.039mm³/million cycles). The friction factors were 0.129 and 0.173 respectively under low and ISO swing-phase conditions. A decrease in the predicted lubricant film thickness when the swing-phase load was increased was observed; at the start of stance phase this was 0.12microns and 0.07microns under low and ISO swing-phase conditions respectively.

The results demonstrate that the performance of metal-on-metal THR’s is highly dependent on swing-phase load conditions. It is postulated that fixation method and surgical technique can affect the swing-phase load. This study has demonstrated that over-tensioning of the tissues may also accelerate wear. These observations may explain some of the variations reported clinically.

Introduction: Following hip replacement surgery the tension of the soft tissues and the laxity of the joint may vary. Variations in surgical approach, technique and fixation method may influence the effective joint laxity and the level of force applied across the prostheses during the swing phase of gait. The aim of this study was to investigate the effect of different swing phase load conditions on the wear metal-on-metal hip prostheses using a hip simulator.

Methods: Cobalt chrome metal-on-metal bearings, 28mm in diameter were tested for five million cycles in a Prosim hip simulator with flexion-extension and internal-external rotation kinematic inputs. A Paul-type twin peak loading curve was applied, which was modified to provide three different swing phase load conditions;

Low positive swing phase load (< 100N)

All tests were carried out in 25% (v/v) new-born bovine serum, with gravimetric wear measurements completed every million cycles.

Results: The wear rates for the different swing phase conditions are shown in Figure I. Elevating the swing phase load from 100N to 300N (ISO load) increased the overall wear rate by 10-fold. Introducing microseparation into the gait cycle increased wear by a further 3-fold. These results indicate the sensitivity of metal-on-metal bearing wear to swing phase load conditions and joint laxity.

Discussion: Little attention to date has been paid to the importance of joint laxity and swing phase load on the wear rate of hip replacements. Elevation of wear rates with increased swing phase load was probably due to the depletion of fluid film lubrication. This was consistent with the findings under stop-start motion [Medley et al., 2002] and demonstrates the dependency of metal-on-metal hip replacements on fluid film lubrication conditions. Testing with a negative swing phase load elevated wear due to microseparation of the components, the head contacted the insert rim at heel strike which caused a stress concentration and damage to the insert rim. The results demonstrate that the wear performance of metal-on-metal hip replacements is highly dependent on swing phase load conditions. It is postulated that the fixation method and surgical technique can effect the swing phase load; over tensioning of the soft tissue may increase the swing phase load, whereas joint laxity will cause a negative swing phase load and possibly microseparation.

Ultra high molecular weight polyethylene (UHMWPE) wear debris induced osteolysis is a major cause of long term failure of total hip replacements. Particles in the 0.1–1.0_m size range are believed to have greater osteolytic potential than larger wear debris. Crosslinked polyethylenes have been shown to have improved wear resistance compared to non-crosslinked materials on smooth counterfaces, however wear debris from cross-linked UHMWPE has been shown to be smaller than that produced from non-crosslinked materials. The aim of this study was to compare the wear, wear debris and biological activity of non-crosslinked and crosslinked polyethylenes when worn against smooth and scratched counterfaces.

Materials and Methods: Test pins were machined from non-crosslinked GUR1050 and GUR1050 crosslinked with either 5 or 10Mrad of gamma irradiation. Sterile endotoxin free clinically relevant wear debris was generated using a bi-directional pin-on-plate test rig. Tests were performed on scratched (Rp=1.0mm) or smooth (Ra=0.02mm) counterfaces. Particles were cultured with murine macrophages at particle volume (mm³): cell number ratios of 50:1,10:1,1:1 and 0.1:1. The levels of TNF-a produced were determined by ELISA following 0,2,4,6,8,24 and 48 hours of culture.

Results: On both smooth and scratched counterfaces crosslinked polyethylene had lower wear than non-crosslinked polyethylene. Determination of the volume distribution of the wear debris demonstrated a greater percentage of wear debris in the submicrometre size range from crosslinked material when worn on scratched counterfaces. Analysis of the debris when worn on smooth counterfaces showed a further reduction in size of debris with particles observed below 100nm in size which reduced the percentage of debris in the sub-micrometre size range for both materials. Crosslinked material worn against scratched counterfaces generated wear debris that was able to stimulate macrophages to produce significant levels of TNF-a after just six hours of co-culture at the highest volumetric concentration and after 24 hours at lower volumetric concentrations. The non-crosslinked material was able to stimulate macrophages only after 24 hours at the highest volumetric concentration. There were no differences between the biological activity of the particles from the 3 materials articulating on the smooth counterfaces they were only able to stimulate significant TNF-a release following 24 hour with the highest volumetric concentration.

Discussion: Although wear resistance is increased by cross-linking on both smooth and scratched counterfaces, when worn against a scratched counterface crosslinked polyethylene generated a greater percentage of debris in the 0.1–1.0mm size range than non-crosslinked polyethylene and this led to an increase in biological activity. However when worn against smooth counterfaces the production of nanometre size wear particles by both materials reduced the volume of debris in the 0.1–1.0mm size range which in turn lead to a lower biological activity.

Following total hip replacement surgery, fluroscopy studies have shown that a mean separation of 2 mm can occur between CoCr femoral heads and UHMWPE acetabular cups during the swing phase of gait [1]. In vivo and in vitro studies [2, 3] of alumina ceramic on ceramic hip replacements have demonstrated that swing phase microseparation followed by the impact of the femoral head on the acetabular insert rim can lead to accelerated wear. However, wear remains low. A similar trend was observed when metal on metal hip replacements were tested under microseparation conditions [4]. The purpose of the current study was to examine the wear of ceramic on polyethylene bearings under standard and microseparation conditions.

A physiological hip simulator was used, loads and motions were applied to approximate in vivo conditions. The alumina ceramic heads and polyethylene cups were 28 mm in diameter and were tested for 5 million cycles in 25% new born calf serum at 1 Hertz. Microseparation was achieved by displacing the femoral head inferiorly during swing phase, where the head contacted the inferior cup rim and was laterally displaced. On heel strike the head contacted the superior cup rim prior to relocation.

The volumetric wear of the polyethylene inserts was approximately four times less under microseparation conditions (5.6 ± 5.3 mm3 per million cycles), in comparison to standard conditions (25.6 ± 5.3 mm3 per million cycles). Deformation of the cup rim was observed, but some of this was attributed to creep. It is postulated that this reduction in wear was due to the separation of the components in swing phase improving the entrapment of lubricant, hence reducing wear via a squeeze film lubrication mechanism. In conclusion, surgical procedures that produce a small and controlled amount of joint laxity and microseparation may lead to a reduction in wear of the polyethylene acetabular cups.

Proponents of the biological theory of aseptic loosening have in recent years tended to concentrate on the production and distribution of particulate ultra-high-molecular-weight polyethylene (UHMWPE) debris around the potential joint space. However, mechanical loading of cemented implants with the differing elastic moduli of metal stems, polymethylmethacrylate (PMMA) cement and bone can result in relative micromotion, implying the potential for production of metal and PMMA particles from the stem-cement interface by fretting wear.

In order to investigate the production and biological reactivity of debris from this interface, PMMA and metal particulate debris was produced by sliding wear of PMMA pins containing barium sulphate and zirconium dioxide against a Vaquasheened stainless steel counterface. This debris was characterised by SEM, energy-dispersive analysis by X-ray (EDAX) and image analysis, then added to cell cultures of a human monocytic cell line, U937, and stimulation of pro-osteolytic cytokines measured by ELISA.

Large quantities of PMMA cement debris were generated by the sliding wear of PMMA pins against Vaquasheened stainless steel plates in the method developed for this study. Both cements stimulated the release of pro-osteolytic TNFα from the U937 monocytic cell line, in a dose-dependent fashion. There was a trend towards greater TNFα release with Palacos cement than CMW cement at the same dose. Palacos particles also caused significant release of IL-6, another pro-osteolytic cytokine, while CMW did not. The particulate cement debris produced did not stimulate the release of GM-CSF or IL1β from the U937 cells. These results may explain the cytokine pathway responsible for bone resorption caused by particulate PMMA debris.

Radio-opaque additives are of value in surgical practice and clinical studies to quantify the relevance of these in vitro findings are required before the use of cement containing radio-opacifier is constrained.

In vivo and in vitro studies of ceramic on ceramic (COC) bearings have demonstrated that swing phase microseparation followed by the impact of the femoral head on the superior acetabular insert rim leads to accelerated wear. However, resultant wear remained low. The wear of ceramic on polyethylene (COP) and metal on metal (MOM) couples under swing phase microseparation is unknown, this study aimed to compare the wear of these total hip replacements under standard and microseparation conditions.

A physiological hip simulator was used, loads and motions were applied to approximate in vivo conditions. Microseparation was achieved by displacing the femoral head inferiorly during swing phase, the head contacted the inferior cup rim and was laterally displaced. On heel strike the head contacted the superior cup rim prior to relocation. Components (as shown in table 1) were tested for 5 million cycles, at a frequency of 1 hertz in 25% (v/v) new born calf serum. Under standard conditions, wear of COC and MOM bearings was significantly lower than wear of COP couples. Under microseparation conditions the COC and MOM wear increased by 4 and 25 times respectively. Microseparation conditions reduced wear of COP couples by a factor of 4. Creep deformation and damage to the UHMWPE cup rim was observed, however, wear remained low. It is postulated that this reduction in polyethylene wear is due to the separation of the components in swing phase improving the entrapment of lubricant, hence wear is reduced via a squeeze film lubrication mechanism.

Introduction: UHMWPE wear debris is known to be a major cause of periprosthetic osteolysis and the long-term failure of total joint replacements by a macrophage-mediated mechanism. The aim of this study was to compare the in vitro response of mononuclear phagocytes from patients undergoing total hip arthroplasty to challenge with polyethylene particles or stimulation with lipopolysaccharide (LPS).

Methods: Peripheral blood was taken from 2 healthy donors and 16 patients admitted to hospital to undergo total hip arthroplasty. Human mononuclear phagocytes were isolated by density centrifugation. Polyethylene particles were sequentially filtered to obtain biologically active particles (0.1–0.6 μm diameter). Cells plus particles, cells plus LPS and cells only were co-cultured in supplemented RPMI-1640 culture medium. Culture supernatants were harvested and the concentration of TNFα quantified by ELISA. Mean specific activity was calculated.

Results:.

When considering all the subjects, no correlation was found between the response of their cells to polyethylene particles and LPS stimulation. However the cells of four subjects gave a much higher response to LPS than the rest and when these where excluded the correlation between the response to LPS and PE particles was significant with an R² value of 0.9076.

Discussion: Despite the different mechanisms by which PE particles and LPS activate macrophages, the patient group with ‘normal’ or low response to LPS had a significant correlation in their response to LPS and particle stimulation. Why a small number of subjects had a much higher response to LPS without a proportional response to PE particles is not known, but it could be due to an increased expression of LPS receptors or genetic polymorphism. A greater than ten-fold difference in the patient response to particles may be of clinical importance in their potential susceptibility to loosening through osteolysis.

We have examined 26 retrieved, failed titanium-alloy femoral stems. The clinical details, radiological appearances and the histology of the surrounding soft tissues in each patient were also investigated.

The stems were predominantly of the flanged design and had a characteristic pattern of wear. A review of the radiographs showed a series of changes, progressive with time. The first was lateral debonding with subsidence of the stem. This was followed by calcar resorption and fragmentation or fracture of the cement. Finally, osteolysis was seen, starting with a radiolucency at the cement-bone interface and progressing to endosteal cavitation.

Three histological appearances were noted: granulomatous, necrobiotic and necrotic. We suggest that an unknown factor, possibly related to the design of the stem, caused it to move early. After this, micromovement at the cement-stem interface led to the generation of particulate debris and fracture of the cement. A soft-tissue reaction to the debris resulted in osteolysis and failure of fixation of the prostheses.

We compared and quantified the modes of failure and patterns of wear of 11 Mittelmeier and 11 Ceraver-Ostal retrieved alumina-alumina hip prostheses with reference to the corresponding clinical and radiological histories.

Macroscopic wear was assessed using a three-dimensional co-ordinate measuring machine. Talysurf contacting profilometry was used to measure surface roughness on a microscopic scale and SEM to determine mechanisms of wear at the submicron level.

The components were classified into one of three categories of wear: low (no visible/measurable wear), stripe (elliptical wear stripe on the heads and larger worn areas on the cups) and severe (macroscopic wear, large volumes of material lost). Overall, the volumetric wear of the alumina-alumina prostheses was substantially less than the widely used metal and ceramic-on-polyethylene combinations. By identifying and eliminating the factors which accelerate wear, it is expected that the lifetime of these devices can be further increased.

We examined stainless-steel, cobalt-chrome, titanium and alumina and zirconia ceramic femoral heads retrieved at revision surgery. All the heads had articulated against ultra-high-molecular-weight-polyethylene (UHMWPE) acetabular cups. We studied the simulation of third-body damage and the wear of UHMWPE against the various materials used for the heads. The surfaces of the retrieved heads were analysed using a two-dimensional contacting profilometer. Third-body damage was characterised by the mean height of the scratches above the mean line (R_pm).

The alumina ceramic and zirconia ceramic retrieved heads were found to have significantly less damage. In laboratory studies the ceramics were also more resistant to simulated third-body damage than the metal alloys. We studied the wear of UHMWPE against the damaged counterfaces in simple configuration tests. The damaged ceramics produced less polyethylene wear than the damaged metal counterfaces. The wear factor of UHMWPE against the damaged materials was dependent on the amount of damage to the counterface (R_p). Our study has shown the benefit of using the harder and more damage-resistant ceramic materials for femoral heads.

Ultra-high-molecular-weight polyethylene (UHMWPE) components for total joint replacement generate wear particles which cause adverse biological tissue reactions leading to osteolysis and loosening. Sterilisation of UHMWPE components by gamma irradiation in air causes chain scissions which initiate a long-term oxidative process that degrades the chemical and mechanical properties of the polyethylene. Using a tri-pin-on-disc tribometer we studied the effect of ageing for ten years after gamma irradiation in air on the volumetric wear, particle size distribution and the number of particles produced by UHMWPE when sliding against a stainless-steel counterface.

The aged and irradiated material produced six times more volumetric wear and 34 times more wear particles per unit load per unit sliding distance than non-sterilised UHMWPE. Our findings indicate that oxidative degradation of polyethylene after gamma irradiation in air with ageing produces more wear.