The poor outcome of large head metal on metal total hip replacements (LHMOMTHR) in the absence of abnormal articulating surface wear has focussed attention on the trunnion / taper interface. The RedLux ultra-precision 3D form profiler provides a novel indirect optical method to detect small changes in form and surface finish of the head taper as well as quantitative assessment of wear volume. This study aimed to assess and compare qualitatively tapers from small and large diameter MOMTHR's. Tapers from 3 retrieval groups were analysed. Group 1: 28mm CoCr heads from MOMTHRs (n=5); Group 2: Large diameter CoCr heads from LHMOMTHRs (n=5); Gp 3 (control): 28mm heads from metal on polyethylene (MOP) THRs; n=3). Clinical data on the retrievals was collated. RedLux profiling of tapers produced a taper angle and 3D surface maps. The taper angles were compared to those obtained using CMM measurements. There was no difference between groups in mean 12/14 taper angles or bearing surface volumetric and linear wear. Only LHMOMs showed transfer of pattern from stem trunnion to head taper, with clear demarcation of contact and damaged areas.3D surface mapping demonstrated wear patterns compatible with motion or deformations between taper and trunnion in the LHMOM group. These appearances were not seen in tapers from small diameter MOM and MOP THRs. Differences in appearance of the taper surface between poorly functioning LHMOMTHRs and well functioning MOP or MOM small diameter devices highlight an area of concern and potential contributor to the mode of early failure.

The poor outcome of large head metal on metal total hip replacements (LHMOMTHR) in the absence of abnormal wear at the articulating surfaces has focussed attention on the trunnion / taper interface. The RedLux ultra-precision 3D form profiler provides a novel indirect optical method to detect small changes in the form and surface finish of the head taper as well as a quantitative assessment of wear volume. This study aimed to assess and compare qualitatively the tapers from well functioning small diameter, with poorly functioning LHMOMTHR's using the above technique.

Introduction

Edge loading commonly occurs in all bearings in hip arthroplasty. Edge loading wear can occur in these bearings when the biomechanical loading axis reaches the edge and the femoral head loads the edge of the cup producing wear damage on both the head and cup edge. When the biomechanical loading axis passes through the polished articulating surface of the acetabular component and does not reach the edge, the center of the head and the center of the cup are concentric. The resulting wear known as concentric wear is low in metal-on-metal (MOM) bearings, and is negligible in ceramic-on-ceramic (COC) bearings. Edge loading is well defined in COC hip bearings. However, edge loading is difficult to identify in MOM bearings, since the metal bearing surfaces do not show wear patterns macroscopically. The aims of this study are to compare edge loading wear rates in COC and MOM bearings, and to relate edge loading to clinical complications.

Introduction

Two types of ceramic materials currently used in total hip replacements are third generation hot isostatic pressed (HIPed) alumina ceramic (commercially known as BIOLOX®forte, CeramTec) and fourth generation alumina matrix composite ceramic consisting of 75% alumina, 24% zirconia, and 1% mixed oxides (BIOLOX®delta, CeramTec). Delta ceramic hip components are being used worldwide, but very few studies have analyzed retrieved delta bearings. The aim of this study is to compare edge loading ‘stripe’ wear on retrieved femoral heads from delta-on-delta, delta-on-forte and forte-on-forte ceramic bearings revised within 2 years in vivo.

Two types of ceramic materials currently used in total hip replacements are third generation hot isostatic pressed (HIPed) alumina ceramic (commercially known as BIOLOX®forte, CeramTec) and an alumina matrix composite material consisting of 75% alumina, 24% zirconia, and 1% mixed oxides (BIOLOX®delta, CeramTec). The aim of this study is to compare BIOLOX delta femoral heads to BIOLOX forte femoral heads revised within 2 years in vivo.

Ceramic bearings revised at one center from 1998 to 2010 were collected (61 bearings). BIOLOX delta heads (n=11) revised between 1–33 months were compared to BIOLOX forte femoral heads with less than 24 months in vivo (n=20). The surface topography of the femoral heads was measured using a chromatically encoded confocal measurement machine (Artificial Hip Profiler, RedLux Ltd.).

The median time to revision for BIOLOX delta femoral heads was 12 months, compared to 13 months for BIOLOX forte femoral heads. Sixteen out of 20 BIOLOX forte femoral heads and 6 out of 11 BIOLOX delta femoral heads had edge loading wear. The average volumetric wear rate for BIOLOX forte was 0.96 mm3/yr (median 0.13 mm3/yr), and 0.06 mm3/yr (median 0.01 mm3/yr) for BIOLOX delta (p=0.03). There was no significant difference (p>0.05) in age, gender, time to revision or femoral head diameter between the two groups.

Early results suggest less volumetric wear with BIOLOX delta femoral heads in comparison to BIOLOX forte femoral heads.

Introduction: Metal on metal hip resurfacing arthroplasty-induced pseudotumours are a serious complication, which occur in 1.8% of patients who undergo this procedure in our institution. The aim of this study was to measure the 3D in vivo wear on the surface of resurfacing components revised for pseudotumour, compared to a control group.

Method: Thirty-six hip resurfacing implants were divided into two groups; 18 patients with a clinical and histopathological diagnosis of pseudotumour and 18 controls (revised for femoral neck fracture and infection). Three dimensional contactless metrology (Redlux^™ Ltd) was used to scan the surface of the femoral and acetabular components, to a resolution of 20 nanometers. The location, depth and area of the wear scar were determined for each component. A separate blinded analysis to determine the presence of absence of impingement was performed by one of the authors.

Results: The volumetric wear rate for femoral component of the pseudotumour group was 3.29 mm3/yr (SD5.7) and 0.79 mm3/yr (SD1.2) for the control group (p=0.005). In the pseudotumour group, the volumetric wear rate of the acetabular component was 2.5 mm3/yr (SD6.9) compared to 0.36 mm3/yr (SD0.80) for the control group (p=0.008). Edge-wear was detected in 89% of acetabular components in the pseudotumour group and 21% of those in the control group (p=0.01). Anterior or posterior edge-wear, consistent with impingement was present on the femoral components of 73% of patients in the pseudotumour group and 22% in the control group (p=0.01).

Discussion: This work demonstrates that implants revised for pseudotumour have significantly higher volumetric wear rates than controls. They also have a significantly higher incidence of edge-wear than controls. We suggest that a significant proportion of pseudotumours are associated with high concentrations of metal wear debris; however a minority may result from a hypersensitivity reaction to metal ions.

Introduction: Metal on metal hip resurfacing arthroplasty-induced pseudotumours are a serious complication, which occur in 4% of patients who undergo this procedure. The aim of this study was to measure the 3D in vivo wear on the surface of resurfacing components revised for pseudotumour, compared to a control group.

Method: Thirty-nine hip resurfacing implants were examined; these were sourced from our institutions prosthesis retrieval bank. They were divided into two groups; 22 patients with a clinical and histopathological diagnosis of pseudotumour and 17 controls. Patient demographics and time to revision were known. Three dimensional contactless metrology (Redlux™ Ltd) was used to scan the surface of the femoral and acetabular components, to a resolution of 20 nanometers. The location, depth and area of the wear scar was determined for each component. Volumetric wear was determined, along with the presence of absence of edge-loading. A separate blinded analysis to determine the presence of absence of impingement was performed by one of the authors. ANOVA was used to test for differences in wear and Fishers Exact test was used to compare the incidence of edge-loading between the groups.

Results: The volumetric wear rate for femoral component of the pseudotumour group was 4.7mm3/yr (SD3.5) and 1.7 mm3/yr (SD1.5) for the control group (p=0.03). In the pseudotumour group, the volumetric wear rate of the acetabular component was 3.5 mm3/yr (SD3.6) compared to 0.02 mm3/yr (SD0.07) for the control group (p=0.01). Edge-loading was detected in 74% of acetabular components in the pseudotumour group and 22% of those in the control group (p=0.01). Anterior or posterior edge-loading, consistent with impingement was present on the femoral components of 73% of patients in the pseudotumour group and 22% in the control group (p=0.01).

Discussion: This work demonstrates that implants revised for pseudotumour have significantly higher volumetric wear rates than controls. They also have a significantly higher incidence of edge-loading and impingement than controls. Edge-loading significantly increases wear. We suggest that pseudotumours are caused by high concentrations of metal wear debris, which have been shown to have a toxic effect on osteocytes and macrophages. This is the one of the first studies to demonstrate a clear link between pseudotumours and increased bearing surface wear. It is also the first to demonstrate that edge-loading, due to impingement, occurs in a significant number of patients who develop this condition. Improved implantation techniques and resurfacing designs may help avoid this serious complication of hip resurfacing.

Introduction: Traditionally Short arm plaster casts have been used to treat distal radius fractures. Judging adequacy of immobilisation has never been defined. A significant proportion of these fractures loose reduction due to inadequate immobilisation. A new non-invasive external fixator technology has been introduced to address the shortcomings of plaster casts. Aim: Is the new non invasive fixator better at reducing skin device interface movement, than conventional plasters.

Materials and Methods: A prospective healthy volunteer study involving application of Short arm plaster of Paris cast, fibreglass cast and a new device Cambfix non-invasive wrist fixator with 15 forearms in each group, was undertaken. IRB approval and informed consent obtained from the volunteers. Colle’s type cast configuration was used. Displacement at the skin-cast and skin-new device interface was measured at proximal and distal ends. Maximal displacements noted immediately after application and after a specified intervals. Casts were windowed at the end of experiment and Cast index and Gap index were measured as ratios at the time of removal of casts. Statistical analysis was done using T-test and SPSS.

Results: The non-invasive Cambfix fixator showed less mean displacement at both the proximal and distal parts compared to plaster and fibreglass casts (p< 0.01). The mean gap index for the Cambfix device was 0.09, which was statistically significantly less than 0.15 and 0.14 for Plaster of Paris and fibreglass casts respectively (p< 0.01). Casts with higher gap index showed increased displacement, however cast index was less predictive of skin-cast displacement.

Conclusions: Skin-device interface movement was significantly better reduced with the Cambfix non-invasive fixator as compared with Short arm plaster of Paris and fibreglass casts. Lesser gap index is known to provide less interface movement. The Cambfix non-invasive fixator appears to achieve a better gap index more consistently. Limitations include healthy volunteer group, and relatively small numbers.

Retrieval analysis offers a direct insight into in vivo wear mechanisms. However, the 3D measurement of wear patch characteristics on spherical highly reflective bearings has been difficult. An instrument based on an optical technique has been developed over the past 3 years. It is capable of scanning metallic head and cup in a single measurement, within minutes, at a resolution of 20 nm. From the cloud of 3D points obtained during scanning (typically 35,000 To 1,000,000), a 3D image of the measured part can be obtained. The associated computer program allows for sphericity, roughness, radius and local radius to be calculated, and surface maps of the 3D model can easily be plotted.

Both head and cup of two failed MoM resurfacing devices, a wear simulator test couple and intact components were analysed using the new technique. A successful McKee Farrar head (20 years in vivo) was also scanned. Results were compared with traces obtained on a Mitutoyo RA 300 roundness machine (resolution 0.01 microns).

3D maps of the bearing surfaces of MoM devices were obtained. The maximum linear wear values on heads were 2.5 microns, 99 microns 53.5 microns and 298 microns for the simulator sample, the McKee Farrar head and the two failed resurfacing devices respectively. The corresponding maximum linear wear values on cups were 11 microns, 529 microns and 645 microns for the simulator sample and the two failed resurfacing devices respectively. These results were in good agreement with results obtained on the Mitutoyo machine. Contrary to other worn samples, the two latter cups showed that the cup had worn on the edge of the bearing surface. This resulted in an oval shaped wear patch on the head. For the McKee Farrar device and the simulator device, the wear patch was away from the edge and the outline of the wear patch was circular in shape.

This novel technique has allowed for high resolution 3D mapping of the full bearing surfaces on successful McKee Farrar device and on more recent resurfacing devices. Further studies are required. However, the results suggest that component positioning is paramount to wear performance of metal on metal devices.

Aim: Lower friction in metal on metal (MOM) hip joints can reduce the wear, production of metal ions and loosening of acetabular cups. The effect of the fluid viscosity on the friction, especially in the relation to the physiological range, is still not fully investigated. The aim of this paper was to study the frictional behaviour of MOM hip joint within the human physiological fluid viscosities.

Materials and Methods: Friction measurement was carried out using a friction simulator on CoCrMo hip prostheses of 50 mm diameter with 100 and 200 micron diametral clearances. The samples were taken from a wear simulator test at 1.4, 2.3 and 3.2 million cycles. A dynamic loading of 100–2000N was applied to the femoral head with a movement of 1 Hz and +/−23 degree amplitude. Tests were performed using 25% new born calf serum which consisted of different ratios of serum and carboxymethyl cellulose with viscosities ranged from 0.0011 to 1.1 Pa s determined by a viscometer at a shear rate of 300/s.

Results: The friction of both clearances reduced with the progress of wear. In the lower range of viscosity, the friction of 100 micron clearance was lower than that of 200 micron clearance. However, when the viscosity reached the range of 0.01–0.06 Pa s, the friction of 100 micron clearance surpassed that of 200 micron clearance and this difference became wider with the increase of viscosity. During one measurement, the curves of 100 micron clearance started from low friction (≈0.05), but increased rapidly for all viscosities except the lowest of 0.001 Pa s. For the 200 micron clearance, the curves were stabilized when the viscosity was below 0.36 Pa s and low friction was observed as the viscosity increased. The friction started to increase only from 0.36 Pa s, but the gradient was less stiff compared to that of 100 micron clearance.

Discussion: Normal synovial fluid is non-Newtonian in nature with shear-rate dependent viscosity. Researchers have reported 300/s shear rate viscosities for normal, osteoarthritis, and inflammatory synovial fluids as 0.010.4, 0.0025–0.2 and 0.001–0.07 Pa s respectively. In this study, the 200 micron clearance had lower friction than that of 100 micron clearance in the majority of physiological viscosity range. Theoretical studies have suggested that smaller clearance and higher viscosity can benefit the lubrication in MOM bearings. However, this theory is valid if continuous and complete lubrication film is achieved. Small clearance and high viscosity may prevent the recovery of lubricant between cycles and cause depletion of lubricant, which can lead to direct contact of bearings and increase of friction. It is therefore concluded that the selection of clearance for MOM components should consider the human physiological fluid viscosities so that an optimal tribological performance can be achieved.

Studies have shown that the normal patella tracks laterally with flexion of the knee joint, consistent with the findings of Eckhoff et al. that the femoral sulcus is lateral to the mid-plane between the 2 femoral condyles. Patellar pain and instability is a known complication of Total Knee Arthroplasty (TKA). To date, several studies have identified the effect of femoral and tibial components on complication after TKA. However, there is very little work on how the design of the implant affects patellar tracking. Our study compares lateralization of the patella in two different AP stabilized knee implants.

A modified caliper was used to measure the width and position of the patella relative to the femur at different degrees of knee flexion. The relationship of the patella midpoint to that of the femur was subsequently assessed. Group 1 consisted of 25 native knees. Group 2 consisted of 25 patients with antero-posterior stabilized knee implant with a spherical medial condyle and a deep lateralized patellar groove (MRK, Finsbury Orthopaedics, UK). And Group 3 consisted of 25 patients with traditional cam-and-post posterior cruciate-substituting implant with a symmetrical patellar groove (PFC-Sigma, DePuy, UK). The mean follow-up for the 50 TKAs was 28 months.

Lateral tracking corresponded well in all groups, but the mean lateral displacement of the patella in group 2 correlated more closely to that of group 1. At 90 degrees of flexion, the patella was displaced a mean of 7mm laterally in both groups 1 and 2, but a mean of 4mm in group 3. Two-tailed Mann-Whitney U test (95% confidence interval) showed that the difference in lateral patellar displacement between groups 1 and 3, and that between groups 2 and 3 were statistically significant (p< 0.05). However, the patellar displacement between groups 1 and 2 was not statistically different.

Our results indicate that lateral patellar displacement in group 2 is similar to that of native knees (group 1). The effect of the underlying lateralized deep patellar groove of the femoral component in group 2 is more able to mimic that of the native femoral sulcus. This intrinsic implant design accommodates the natural tracking of the patella.

Current CoC hip bearing ranges are typically from 28mm to 36mm diameter, but to improve stability and range of motion, a novel large diameter hip bearing is introduced, with bearing diameter from 32mm to 48mm. To minimise acetabular bone loss, a low profile acetabular component is required and achieved with a ceramic wall thickness of 3.5mm and metal shell thickness of 1.5mm at the acetabular rim. This paper presents some of the testing required to develop this novel design.

Finite element (FE) modelling was performed to simulate the standard 46kN burst loading of the acetabular cup for 5 different geometries of ceramic liner and metal (titanium) shell. By smoothing the facets on the back of the ceramic and thinning the metal shell, the stresses in the ceramic were reduced by 20% and failure was not predicted for the burst test. Reducing the thickness of the metal shell increased the stresses in the metal, but these were kept below the yield strength of the material. When assembled, the hoop stresses in the titanium shell caused a greater volume of the ceramic to be in compression and the strength of the assembled cup was therefore increased.

To assess the effect of fatigue loading on the ceramic/titanium taper-lock, cups were loaded at 45° to the horizontal for 10,000 cycles in Ringer’s solution at 37°C. The load required to push the ceramic from the metal shell were recorded after the test and compared to the push out load of unloaded specimens. There was no significant decrease in the push-out load (mean 2kN) indicating that the taper lock retains its strength during fatigue loading.

The new CoC acetabular cup design is assembled under controlled conditions before packaging. To demonstrate the effectiveness of this, the new device was compared to a commercially available intraoperatively assembled Ti/ceramic device which had a metal shell thickness of 5mm. The cups were placed in reamed cavities of polyurethane foam and the rims impacted with increasing impact energy until the ceramic came loose from the metal shell. An average impact energy of 4J (1kg dropped from 400mm) was necessary to separate the ceramic from the metal liner of the new design compared to 2J for the commercially available design. The thicker titanium wall thickness and intraoperative assembly method of the commercially available design limited the amount of shell deformation/hoop stress generated, and therefore limited the ‘grip’ of the Ti/ceramic interface. The thinner titanium shell (1.5mm) and controlled assembly load of the new design allowed greater shell deformation/hoop stress which produced a two-fold improvement in interface strength. Further effects of assembly in vivo, in particular the effects of periprosthetic or lavage fluids, remain to be investigated. In any case, incomplete ceramic liner seating has been reported in 16% of procedures in vivo (Lang-down, JBJS Br 2007) and the preassembled design therefore represents a notable and necessary improvement to current technology.

Femoral component sizing can play a critical role in the clinical outcome and success of a TKR prosthesis. In particular, achieving the correct AP dimension for the femur is important to ensure correct balancing and to maintain flexion/extension spacing and the ML width dictates bone coverage which, if insufficient, can cause complications or affect long-term outcomes.

There has been some discussion in the literature about the optimal femoral component shape and size with reports of differences in anatomy between male and female patients or those of larger or smaller stature. The majority of these publications have been conducted on normal anatomy with un-cut bone, reporting on the epicondylar width of the femur which is difficult to relate back to the dimensions of a prosthesis. Some studies have measured resected bone, however, the prosthesis and instruments used to make the cuts dictate the amount of bone removed anteriorly and posteriorly which, in turn affect the footprint of exposed bone that is measured.

Data was gathered to assess whether a generic prosthesis with a standard AP/ML sizing ratio could be used to cover the range of femoral sizes dictated by a Caucasian population of 26 male and 26 female patients. MRI scans were obtained for these patients, all between 20 and 45 years of age and diagnosed with a meniscal tear. A theoretical size range for a prosthesis was determined from an analysis of literature data and a review of currently available devices. This consisted of 8 femoral sizes ranging from 50 – 74.5 mm in AP dimension with a constant AP/ML ratio of 0.9. Each MRI scan was viewed in the sagital plane and the maximum AP dimension was measured. This was sized to the closest available femoral component using the criteria of matching the existing articulating geometry as closely as possible. A ‘virtual’ distal condyle cut was made on the scan relating to the component size and the ML dimension of the resected bone was taken. The measured ML data was then compared to the implant dimension for each subject and component overhang/underhang was determined.

An appropriate femoral component match was found in all cases with a mean AP dimensional undersize of 1.71 mm across all patients (range: 0.16 – 3.77 mm). The mean ML femoral component overhang was 0.34 mm for the male population, 1.52 mm for the female population and 0.89 mm for all 52 patients. These values were all considered to be well within an acceptable range and not be significant in terms of clinical outcome. No patient was too large for the largest component, however no patient in the population that was assessed matched the smallest of the 8 components.

This simple dimensional assessment has shown that using a prosthesis with a standard AP/ML ratio, it is possible to accommodate a mixed gender population. The data reported here suggests that the anatomical differences between men and women femora is not hugely significant and can be covered with a common implant provided a sufficient size range is used.

Finsbury Orthopaedics would like to acknowledge Dr. Pinskerova for providing the MRI scans.

Variability in femoral head preparation and high cement pressures may be associated with failure to seat femoral components during hip resurfacing. Furthermore, excessive pressures may lead to over penetration of bone by cement with resulting necrosis of the underlying bone. We designed an experimental model to test the hypothesis that partial-length pressure-relief slots made longitudinally in the proximal bone of the femoral head, without extending to the head neck junction, would allow controlled leakage of cement during initial insertion of a femoral head resurfacing component, but would then become sealed during final insertion to prevent excessive loss of cement while still allowing accurate seating of the component.

Thirty-one resurfacing femoral components were cemented onto foam femoral head models. The clearance between foam model and implant was measured to determine the minimum space available for cement. Eleven components were inserted using hand pressure alone, 20 were hammered. Pressure relief slots were prepared in 10 femoral heads. The slots, 4mm deep grooves, were made in the proximal bone only, without extending to the head-neck junction. Cement pressure inside the component was measured during insertion. Implants were sectioned after implantation in order to determine whether they had been fully seated or not. The clinical relevance of the measures taken was tested by measuring the diameter of prepared femoral heads during 20 hip resurfacing operations in order to determine the extent of variability in intra-operative femoral head preparation.

Mean intraoperative clearance between bone and implant was −0.19mm (0.11 to −0.93mm). Mean clearance between foam model and implant was −0.30mm (0.35 to −0.94mm). Full seating was obtained in 22/31 components. Of those not fully seated, all had clearance less than −0.74mm. Full seating with a clearance of less than −0.35mm was only possible when pressure relief slots had been made in the femur. The use of a pressure relief slot longer than half the femoral head length allowed full seating in 9/9 cases, compared to 13/22 without. Cement pressure obtained with a hand pressure technique was less than half that observed with hammering (20.8vs56.0psi, p=0.0009) but was not associated with failure to seat the implant if a slot was used.

Variability of the actual diameter of the femoral head prepared may be associated with difficulty in fully seating resurfacing components. The same degree of variability in the space available for cement was observed in both intra-operative and test specimens. The use of a pressure-relief slot allows full seating of resurfacing implants with hand pressure alone, thereby halving cement pressure, in an experimental model, even when clearance between implant and bone is less than optimal.

Squeaking is a rare complication of hard-on-hard hip bearings. Occasionally the noise is troublesome enough to warrant revision surgery. The purpose of this study is to contribute to the understanding of the mechanism(s) underlying squeaking.

We analyzed 10 alumina ceramic-on-ceramic bearings from squeaking hips collected at revision surgery. The reason for revision was given as squeaking (6 cases) or squeaking and pain (4 cases). Six of the 10 patients were male, average patient age was 48. Bearings were retrieved after an average of 23 months in service (11 to 61 months). There were 4 different designs of acetabular component from 2 different manufacturers. Nine have an elevated metal rim which is proud of the ceramic and one does not. Two bearings were 36mm in diameter, 6 were 32mm and 2 were 28mm.

All 10 bearings showed evidence of edge loading wear. Mean dimensions of the wear patch were 37mm by 12mm on the acetabular component and 32mm by 13mm on the femoral heads. Wear dimension was not related to bearing diameter. Seven of the 10 implants also had evidence of impingement of the femoral neck against the elevated metallic rim or the ceramic insert or both. There was no chipping or fracture of any of the ceramic components.

Squeaking is a recently recognized complication of hard on hard bearing surface. This retrieval study is the first of its kind, to our knowledge attempting to unravel the mechanism of this undesirable complication. Although impingement seems to be present in majority of cases, the latter does not seem to be necessary. Edge loading wear was the common factor in all cases and this may prove to be a critical mechanism.

The clearance between the femoral head and the acetabular cup can significantly affect the lubrication, the wear and the lifetime of metal on metal (MOM) hip joints. The objective of this study was to compare the frictional behaviour of MOM joints with different clearance.

Two CoCrMo MOM 50mm diameter hip joints, with a small diametral clearance of 17 microns and a big diametral clearance of 212 microns, were used in this study. The friction measurement was carried on the wear patches of MOM bearings during a long-term wear simulator test. A dynamic trapezoidal-form loading cycle was applied to the femoral head with a minimum load of 100N during the swing phase and a maximum load of 2000N throughout the stance phase. A simple harmonic motion of amplitude +/−24 degree was applied to the femoral head in the flexion-extension plane with a frequency of 1 Hz. The friction torque was measured at 0, 0.8, 1.3, 1.9, 4 and 5.5 million cycles using 6 different viscosities of 25% new born calf serum.

The results show that the friction factors (f) of small clearance were generally higher than those of big clearance and this difference became wider with the progress of wear. The lower f of big clearance, especially in the lower range of Sommerfeld number (z) after 5.5 million cycles, is significant and will affect the ultimate performance of prostheses as this range has closer rheological properties to synovial fluid and represents long term wear conditions. At the same time, the friction factors were always higher every time when measured from high z to low z, although this difference became slightly smaller with the progress of wear, which indicates that there is still direct contact between the bearings. The lower friction factor when increasing z, is due to the wear and bedding-in with the progress of the measurement. It is concluded that large clearance has lower friction factor than small clearance, and full fluid film lubrication is unlikely to have developed between the MOM bearings in this study, even with a small clearance and high viscosity.

Introduction and Aims: The extremely low wear rates of third generation alumina-alumina bearings in traditional hip simulators are not reflected in vivo. Separation of the bearing during swing phase and edge loading with heel strike is reported to account for this discrepancy.

Method: We have had the opportunity to visually inspect 21 bearings at re-operation from a group of 1588 hip arthroplasties with third generation alumina ceramic-ceramic bearings. Re-operations were for heterotopic ossification (one), loosening (three), femoral fracture (six), psoas tendonitis (six), sepsis (three) and dislocation (two). There were no re-operations for bearing failure. Sixteen of these 21 bearings (16 heads and 12 inserts) were retrieved and analysed. We mapped the location and we measured the volume of the wear and we performed microscopy and measured roughness of worn and unworn areas.

Results: Eleven bearings had visual evidence of edge loading wear, making an incidence of 52% in the 21 patients having re-operations. These 11 bearings and five visually undamaged bearings were analysed. The wear on the insert was always located at the rim indicating edge loading. The location and orientation of the stripe on the head was not consistent with subluxation during normal gait but was consistent with subluxation and edge loading with the hip flexed at 90 degrees. The average wear volume was 0.7mm3 per year (heads plus liners). Longer service bearings had signs under SEM of repolishing of the wear area suggesting that the process of edge loading wear will be self-limiting. The heads without a wear scar showed very little damage: under SEM, a slight relief polishing of individual grains and minor pitting was noted.

Conclusion: The subluxation causing the stripe wear in these patients did not occur during normal walking gait. It probably occurred with rising from a chair. Simulator testing of third generation alumina-alumina components must include edge loading if it is to give a realistic indication of in vivo performance. Alumina-alumina bearings remain an excellent option for total hip arthroplasty, however more work is required to understand the clinical consequences.