July 26, 2008

Biocompatible Alumina Ceramic for Total Hip Replacements

By Zeng, P

Their resistance to wear and biocompatibility make ceramics ideal materials for medical applications, such as implants. For over 30 years, pure alumina has been the dominant material for ceramic hip prostheses. Interest in alumina hip prostheses continues to grow, due to the relatively short life of polymer/metal prostheses, mainly resulting from osteolysis and aseptic loosening caused by polymer wear debris. Since its introduction by Boutin in the 1970s, substantial improvements have been achieved in the microstructure of medical grade alumina by improving purity and processing to give complete densification and fine, uniform grain sizes. A brief review is given of the types of alumina used in total hip replacement, the development of medical grade alumina, and methods of in vivo and in vitro investigation of alumina prostheses, with a focus on current knowledge of the damage observed on alumina prostheses. Particular attention is paid to wear mechanisms and the influence of materials properties on wear behaviour. A region of relatively severe wear, known as stripe wear, is widely observed on retrieved alumina hip prostheses. This type of wear can now be replicated in vitro in joint simulators by the introduction of a 'microseparation' motion during the test cycle. Finally, the future of ceramic hip prostheses and development of the next generation of ceramics for hip prostheses is discussed. Keywords: Microstructure, Alumina, Wear, Hip prostheses

This review is a revised version of a commended entry in the 2007 Literature Review Prize of the Institute of Materials, Minerals and Mining, set up to encourage the preparation of critical literature reviews by students as an essential part of study for a higher degree in the materials field, and to make the best of these available to a wider readership.

Total hip replacements

About 2% of people suffer hip problems that lead to the need to replace their natural hip joints. In the UK at least 50 000 hip replacement surgeries are carried out every year, and are highly successful in reducing the pain and disability of worn or damaged hip joints.1 Total hip replacement (THR) is one of the most successful applications of biomaterials.

The basic design of a THR involves a femoral head, an acetabular cup and a metallic stem (Fig. 1). The review by Dowson3 summarised six alternative existing or potential combinations, according to the materials used for the femoral head and acetabular cup, as: metal- on-polymer, such as the Charnley prosthesis (Fig. 1a), metal on metal, such as the McKee-Farrar prosthesis (Fig. 1b), ceramic on ceramic (Fig. 1c), ceramic on polymer (Fig. 1d), ceramic on metal (Fig. 1e), and metal on ceramic. The first four combinations are already in the market, the latter two being potential combinations that are still at the laboratory stage.6

The first metal on polymer THR, also the first true artificial hip joint, implanted by Charnley in 1959,7 contained a stainless steel femoral head and a poly(tetrafluoroethylene) (PTFE) acetabular cup. Because of the toxicity caused by PTFE wear debris, Charnley shortly afterwards changed the PTFE to ultrahigh molecular weight polyethylene (UHMWPE).8 Almost at the same time, McKee and Watson 9 developed the metal on metal THR using a cobalt-chromium alloy (Fig. 1b). The first ceramic on ceramic THR was introduced by Boutin in 1970 and developed by Mittlemeier in 1974. Later in 1975, alumina on polymer THRs were implanted in Switzerland.10

The metal on UHMWPE THR, namely, the Charnley low friction arthroplasty (LFA), has dominated the THR markets for four decades and is widely referred to as the 'gold standard'.3 Nevertheless, the survival rate of the Charnley LFA is not satisfactory in the long- term (>15 years), dropping from 94-2% at 10 years follow-up to 88- 7% at 15 years follow-up.11 It is estimated that 15% of hip replacement surgeries are second replacements,1 also known as revision surgeries, due to the failure of the first THR. The revision operation always takes longer, and is harder, resulting in an even lower success rate. Therefore, some surgeons even reserve THRs for patients aged over 60 years to reduce the rate of THR failure. However, with the increasing life expectancy of the population, there is a requirement for long-term (>15 years) performance THRs. In addition, there is a significant, and increasing, number of patients younger than 50 years old who have hip damage from severe sports injuries. It is important to reduce the wear of THRs and improve the lifespan of the artificial hip prostheses.

1 Existing or potential materials combinations for THRs classified by materials used for femoral head and acetabular cup: a- d are already in market, whereas e is still in laboratory stage

The main causes of failure of THRs are loosening of the joint due to wear, and inflammation caused by a reaction to wear particles from the artificial joint surfaces that have been absorbed by surrounding tissue. Osteolysis (resorption of periprosthetic bone), which is believed to be a result of polymer wear debris, has been widely observed and has become a major problem limiting the lifespan of polymer THRs.10,12 Therefore, there is renewed interest in hard on hard THRs, such as ceramic on ceramic and metal on metal.

Ceramic on ceramic total hip replacements

Ceramics are regarded as favourable materials for THRs due to their wear resistance and consequent reduction of wear particles. Alumina is the most widely used ceramic for hip prostheses. Compared with other combinations, such as metal on polymer and alumina on polymer, alumina on alumina THRs show the lowest wear rate under laboratory conditions13,14 (Fig. 2). Additionally, alumina is a bioinert material that reduces the chance of osteolysis. However, the early clinical performance of alumina on alumina THRs was controversial. High fracture rates of alumina femoral heads, from 6.9 to 13.4%, were cited15-17 during the 1970s, although over a later period, fracture rates as low as 0-4% were reported.18-20 It is worth noting that the aluminas with high fracture rate were based on formulations developed for industrial applications,15 l7 whereas the low fracture rates were observed from medical grade alumina.18- 20

2 Comparison of volumetric wear rates in pin on disc wear test (mm^sup 3^/million cycles) of different materials combinations:13 note that alumina on alumina pairs show lowest wear rates under laboratory conditions

3 Microstructures of alumina for THRs: note enrichment of glass phases in grain boundaries and coarse grain size in a

Apart from the material itself, poor hip design and operation skills were also responsible for the early failure of alumina on alumina hip prostheses. Unlike polymers, there is a high risk of fracture during an operation due to the low toughness of alumina. Fracture of alumina on alumina hip prostheses is a major problem for surgeons and restricted the application of alumina on alumina THRs in the early years (see below). Interestingly, research on alumina on alumina hip prostheses in the first 30 years was mainly a European development, with some clinical research in Australia and Japan, because the USA Food and Drug Administration (FDA) banned the use of alumina hip prostheses before 2003 due to the high fracture rates encountered in the 1970s.

Early alumina on alumina THRs used aluminas developed for industrial applications, the microstructure of which is poor (Fig. 3a): insufficient purity, low density and coarse grain size (compared with latest alumina for hip prostheses, Fig. 3b and c). The importance of developing alumina materials for medical use was rapidly realised and the first standard for alumina for hip prostheses, ISO 6474,22 was established in 1984. ISO 6474 qualifies medical grade alumina and decreases the risk of fracture in medical applications. With the improvement of medicalgrade alumina material and better hip design, alumina on alumina THRs achieved great success in Europe. The work of Sedel et al.23 showed increasing survival rates of alumina on alumina THRs for patients younger than 50 years and it is now well accepted by surgeons that alumina on alumina THRs are the first choice for the patients younger than 50 years. Based on the successes in Europe and good results of trials on modern alumina-on-alumina hip prostheses in the USA,24 the FDA withdrew the ban on alumina on alumina hip prostheses in 2003. Research on alumina on alumina hip prostheses is now a global enterprise and the likelihood of widespread adoption of ceramic on ceramic hip joints is high.

Development of medical grade alumina

As indicated above, the pioneering alumina materials for hip joints were based on ceramics developed for industrial applications, which had insufficient mechanical strength for in vivo applications and hence poor reliability and fracture rates.25 As a result, ISO 6474 was set up to qualify ceramics used in hip prostheses.22

Significant improvement of medical-grade alumina was achieved in Germany in the early 1970s, as part of the development of a range of modern engineering ceramics.26,27 The development of medical grade alumina has been rationalised in terms of three generations of ceramics (Table 1). Developments have been concentrated on purity, grain size and density in view of the close correlation between mechanical strength and these characteristics. Table 1 Mechanical properties of medical grade alumina27

4 Wear rate-grain size dependence in wet erosive wear test of pure polycrystalline alumina:28 note wear rate decreases with decreasing grain size

Glassy phases are commonly found on the grain boundaries of ceramic materials as a result of impurities in the raw materials. The glassy phase tends to degrade in the body and cause the material to age, i.e. to loose its mechanical strength.28-30 For the latest, third generation of alumina, purity is as high as 99.9%26 and glassy phases are not observed even in TEM (Fig. 3c). However, it is worth noting that some second phases are useful, such as magnesium oxide (MgO). It is believed that the grain size of alumina can be effectively controlled by doping with MgO.26

Therefore, a reduction in grain size is an essential requirement in reducing wear. Reducing the grain size is also believed to be beneficial in delaying the transition to higher wear rates (Fig. 6),30,31 and this has been reflected in the development trends for medical-grade alumina (Table 1).

Porosity (i.e. density) also plays an important role in wear of alumina.32,33 Pores, especially intergranular pores, can act as sources of slip or twinning and promote wear.32,33 Third generation aluminas are hot isostatically pressed (hipped) to give increased density of 3.96 kg m^sup -3^ (Table 1). Since the temperatures used in hipping are lower than those for sintering, grain growth is limited, which also benefits grain size.

5 Schematic illustration of a development of intergranular microcracklng and b wear process as summation of crack growth and delay steps28

6 Wear data for nominally 'pure' alumina ceramics of three grain sizes: room temperature data for rotating silicon nitride sphere, 12 mm in diameter, 450 N load, on flat specimen, paraffin oil lubricant; note initial slow, steady increase in scar diameter with sliding time, followed by abrupt transition to severe wear at critical sliding time; sliding time for onset of transition diminishes significantly for larger grain size materials; vertical dashed lines are theoretical predictions of transition times31

7 Summarised fracture rates of alumina on alumina hip prostheses as function of ceramic brand and implantation period: note dramatic decrease of fracture rates in later 1970s and early 1980s15- 18,24,27,34-46

The colour of medical grade alumina is related to its quality and chemical purity.26 Alumina ceramics with low impurity levels, but which have been doped with magnesium oxide (MgO) as specified by ISO 6474, are not white, but ivory in the unsterilised form. When the ceramic ball heads are sterilised using gamma rays, the material turns brown, which can be explained in terms of the absorption due to the different valences of aluminium (3), oxygen (2) and magnesium (2).24

In vivo wear performance of alumina on alumina THRs

In vivo study of hip prostheses, involving clinical evaluation of prostheses retrieved from bodies, is the most direct method to investigate the wear performance. However, because of the limited supply of retrieved hip prostheses and the need to protect patients, it is extremely difficult for researchers to access in vivo materials. Most prostheses in the work reviewed below are manufactured from first and second generation medical grade alumina, with relatively limited information from retrieved third generation prostheses. It is also worth noting that most observations have been made by surgeons who have direct access to the retrieved prostheses.

Fracture rate

The results of fracture analyses of the alumina hip prostheses are summarised in Table 2 and Fig. 7.

Fracture rates as high as 13-4% were observed for first generation alumina on alumina hip prostheses in the early years.16 After 1980, a dramatic decrease in fracture rates occurred, with several authors reporting 0% failure.34-37 No official data for third generation alumina have been published; however, a lower fracture rate is indicated. It can be concluded that fracture is no longer the problem for surgeons that it was in the 1970s and modern alumina on alumina hip prostheses are unlikely to fail during hip operations.

Wear rate survival rate

The wear rate is the most direct way to evaluate wear performance of prostheses. Alumina on alumina hip prostheses exhibit lower wear rates, deduced from in vivo investigation, than either metal on polymer or metal on metal hip prostheses. Reported linear wear rates range from 0.5 x 10^sup -3^ m/year, with a mean value of 5.6 x 10^sup -5^ m/year.25 On this basis, ceramic on ceramic prostheses could function in vivo without noticeable abrasion for nearly 10 years.47

Although the wear rate is low, the published survival rates of alumina on alumina hip prostheses are not satisfactory. These are summarised in Table 3 and Fig. 8. It can be seen from Table 3 that the survival rates of early alumina hip prostheses, before ISO 6474, are lower than those for the Charnley LFA.3 The limited data for second generation alumina hip prostheses show improved survival rates of 100% on 5 years follow-up and 95-1 and 94-3% on 7 years follow-up for two different hip design.44 However, there is a lack of data for third generation prostheses as a result of their restricted time in service.

Table 2 Reported fracture rates of alumina on alumina hip prostheses

Table 3 Reported survival rates of alumina on alumina hip prostheses

Table 4 Published data on revisions and causes

It is worth noting the work of Sedel et al.23 which showed increasing survival rates of alumina hip prostheses for patients younger than 50 (87% for patients younger than 50 year on 15 year follow-up, against an overall survival rate of 70%). Although there is uncertainty on the long term performance of alumina on alumina hip prostheses due to the limited data, alumina hip prostheses show the best performance in the patient group younger than 50 years. It has been suggested23 by surgeons that for young and more active patients, alumina on alumina hip prostheses are good choice; however, the standard Charnley LFA may still be most appropriate for older patients.8

Reasons for revision

Table 4 summarises the published data on the reasons for revisions. It can be concluded that fracture of alumina is no longer the main cause of failure; since the introduction of medical grade alumina, aseptic loosening has become the main reason for revision surgeries. However, the mechanism leading to this loosening is not clear, i.e. it has not been established whether wear is a factor in the loosening process. Therefore, there remains a need for further studies on the wear mechanisms of alumina hip prostheses.

8 Summary of survival rates of alumina on alumina hip prostheses in literature as function of implantation period23,34,39,44,48

Wear debris

Polymer wear debris is believed to be the main cause of osteolysis as a result of biological reaction with tissue.10,12 However, the role of ceramic wear debris remains unclear and only limited data involving ceramic wear debris are published.

Numerous particles with a mean size of 5 [mu]m were observed in the pseudosynovial tissue obtained from revisions due to mechanical failure.52 In addition, wear debris in the size range 5-90 nm (Fig. 9) was revealed in the tissue from alumina on alumina hip prostheses, together with wear debris 0.5-3.2 [mu]m in size.53 This was the first description of nanometre sized alumina wear particles in retrieval tissues.

Two mechanisms were proposed to generate the two types of alumina wear debris:53

(i) relief polishing, producing nanometre sized alumina wear debris under normal articulating conditions

(ii) intergranular and intragranular fracture due to edge loading under condition of microseparation of the head and cup on rim contact, generating larger wear particles.

Yoon et al.54 reported the only case of osteolysis caused by ceramic debris. However, some surgeons argued that this was an exceptional case resulting from poor design of the THR.8 Therefore, clinical results concerning osteolysis of alumina hip prostheses are required to clarify the observation.

In vitro wear performance of alumina on alumina THRs

In vivo experience is the most direct and convincing means of obtaining data on long term hip prostheses performance; however, studies are based on failure analysis of retrieved prostheses, largely ignoring successful implants. In addition, supply of in vivo samples is limited and is not helpful to validate new materials. Therefore, in vitro studies are often carried out as a complement to in vivo studies.

In vitro tests performed outside a living organism, often in conditions designed to simulate in vivo service, are an effective way to investigate prospective new materials for THRs. Pin on disc tests, reciprocating pin on plate tests and joint simulators are the most widely accepted in vitro methods for characterising the tribological behaviour of hip prostheses.3 The first two tests are widely applied in materials science to study the tribology of materials, e.g. to compare different material combinations or the effect of different lubricants. A series of pin on disc tests indicated that the coefficient of friction of alumina on alumina hip prostheses is affected by the type of lubrication,55 being much smaller when a 1 wt-% aqueous solution of carboxymethyl cellulose sodium salt was used as lubricant instead of distilled water. Therefore, alumina on alumina hip joints might benefit from full fluid film lubrication with appropriate machining to produce a good surface finish, good fit and a proper lubricant.

9 Wear debris in size range 5-90 nm in tissue from worn alumina on alumina hip prosthesis50 In vitro studies are also used to analyse potential materials used for hip prostheses, such as zirconia/ alumina combinations. Although zirconia/zirconia pairs have shown poor wear performance,56 zirconia/alumina and alumina/ zirconia pairs exhibited similar wear performance to alumina/ alumina. On this basis, combinations of zirconia/alumina or alumna/ zirconia were proposed as potential materials for hip prostheses.56

Although pin on disc studies have proven effective to investigate the tribology of potential materials for hip prostheses, these conditions are not realistic in view of the complex environment, configuration and stresses experienced by human hip joints. This perception led in the 1970s to the development of a more sophisticated approach to joint tribology, based on joint simulators,3 which are now used for most in vitro investigations.

In joint simulators, alumina on alumina bearings show lower wear rates than metal on polymer or metal on metal bearings. However, the wear rates are significantly lower than those observed in vivo.34,50,57-61 Fluoroscopy studies60 have suggested that the discrepancy is a result of differences in joint separation between simulated and in vivo joints. Hip joint separation could be resolved in vivo if the amount of separation was >0.75 mm. For gait, the maximum and minimum separations were found to be 2.8 and 0.8 mm (average: 1.2 mm); and for abduction/adduction leg lift, 3.0 and 1.7 mm (average: 2.4 mm). It was hypothesised that this micro separation could occur with any hip prosthesis and could be a factor in the initiation of fracture based wear, leading to the characteristic 'stripe' observed on ceramic on ceramic hip prostheses.61 The small clearances between the head and socket (typical radial clearances are 30 [mu]m) mean that the femoral head may translate inferiorly and laterally if micro-separation occurs. These displacements would typically be

10 Schematic depiction of relative microseparation during gait cycle25

11 Schematic illustration of microseparation mechanism in in vitro simulation63

Wear of hipped alumina on alumina bearings for THR under microseparation conditions has been studied at the University of Leeds.62,63 During the first million cycles (bedding-in) of the microseparation tests, characteristic stripe wear was observed on all femoral heads, with a matching area on the rim of the acetabular inserts. Under mild microseparation conditions, produced by a swing phase load of 400 N, an average wear rate of 5.5 x 10^sup -10^ m^sup 3^/million cycles was observed during the initial million cycles, which reduced to a steady state level of 1.0 x 10^sup - 10^ m^sup 3^/million cycles. Under more severe conditions, produced by a swing phase load of 50 N, an average wear rate of 4.0 x 10^sup -9^ m^sup 3^/million cycles was observed during bedding-in, which reduced to a steady state level of 1.3 x 10^sup -9^ m^sup 3^/ million cycles. In contrast, a bedding-in wear rate of 1.1 x 10^sup -10^ m^sup 3^/ million cycles and steady state wear rate of 5 x 10^sup -11^ m^sup 3^/million cycles for the same material (hipped alumina) were observed under normal simulation with no microseparation.

It is clear that microseparation raises slightly the wear rate of alumina on alumina hip prostheses in simulated tests. Furthermore, under microseparation, the wear mechanisms and wear debris were similar to those observed in previous alumina retrieval studies with debris ranging from 10nm to 1 [mu]m in size. Replication of the stripe wear, found on retrieved in vivo alumina hip prostheses, was observed for the first time (Fig. 12). The in vitro performance of hip joints with microseparation is much more similar to the in vivo performance than in previous simulations.

12 Stripe wear of hipped alumina components (marked):63 left image shows pair of alumina hip prostneses explanted after 1 year and right Image pair of alumina hip prostheses following In vitro microseparation

13 Schematic depiction of wear on retrieved in vivo alumina hip prostheses

Microstructure of worn alumina on alumina THRs

Plastic deformation, cracking and chemical reaction are the main wear mechanisms of alumina. It is well documented that wear behaviour is strongly affected by microstructural features, such as grain size and pores.30-33,64 However, research to date on alumina on alumina hip prostheses has concentrated on the wear performance of the overall system rather than the wear mechanism of the material. Most observations were made by surgeons, and there is a lack of quantitative data on either microstructure or mechanical properties. Only limited work has been found in literature concerning microstructures of worn alumina on alumina hip prostheses.21,46,49,50,62,65,66 While knowledge of wear rate or survival rate may be sufficient for surgeons or patients wishing to choose between existing implants, it does not provide the fundamental understanding of wear mechanisms crucial to the development of new ceramic materials.

Walter and Plizt21 investigated the wear surfaces of 29 retrieved in vivo first generation alumina hip prostheses. Wear was observed in two different zones (Fig. 13): a 'pole' zone (Fig. 13a), namely, at the top of the femoral heads or bottom of the acetabular cups, and an equatorial zone near the edges of the acetabular cups or corresponding zones of the femoral heads (Fig. 13b). The emergence of these two wear zones can be attributed to edge loading due to microseparation, for example, Fig. 10c for the pole zone wear and Fig. 10b for the equatorial zone. It was proposed that low level wear begins when the surfaces first came into contact; this appears to progress to severe wear in the contact areas of the component associated with high stresses and poorer lubrication.50

It is worth noting that the equatorial zone has subsequently been widely observed.46,50,62,65,66 However, no other reports of pole zone wear have been found in the literature. Furthermore, some authors66 argue that wear always occurs on the rim of the socket and never on the apex (equivalent to pole zone wear). Although pole zone wear is feasible by a microseparation mechanism, further observations are required to support the hypothesis.

Walter and Plizt21 also observed little difference in the microstructure corresponding to different wear patterns. Loss of intergranular cohesion led to severe wear in either the pole zone or equatorial zone. However, outside the wear areas, the loss of one or more grains was observed in regions of the originally highly polished bearing surfaces (caused by occasional grain pull-out in the mild wear zone). They proposed that the severe wear might be explained not only by the 'first point of contact' mechanism but also by dry scratching of nonpolished bearing surfaces.21

Interestingly, energy dispersive microanalysis revealed a region of high alkaline earth and silicon concentrations in the grain boundaries after thermal etching21 (Fig. 3a). These regions were sensitive to corrosion, especially in zones where microcracks occurred, which might be the initiation point for later fracture of the component. Note, however, that this work21 investigated first generation alumina hip prostheses implanted between 1975 and 1978, and the higher purity third generation materials would not be expected to show such impurity concentrations at grain boundaries.

Further investigations were made on the worn surface of retrieved alumina on alumina hip prostheses, for which clinical observations had given a range of reasons for failure.21 Although different wear patterns were observed on the retrieved prostheses, plastically deformed, agglomerated alumina wear debris was believed to play an active role in the enhancement of the avalanche effect, associated with exaggerated wear in certain alumina/alumina Autophor (Mitterlmeier) hip joints.21

14 Images (SEM) of 'stripe' wear

The Leeds group50,61 made significant progress in characterising the microstructures of worn alumina on alumina hip prostheses. They analysed retrieved in vivo first and second generation alumina prostheses implanted between 1977 and 1994. Few differences were observed between the worn surfaces of first and second generation aluminas. Three regions were distinguished, according to the wear pattern:

(i) low wear: little visible wear, surface remained polished

(ii) stripe wear: an elliptical wear stripe visible on head, with roughing of part of the cup surface sometimes visible

(iii) severe wear: both head and cup showed large areas of wear and loss of sphericity.

The unworn areas outside of the main contact area (low wear area) had a mean roughness of 0.005 [mu]m R^sub a^. Wear stripes had a typical roughness of 0.1-0.2 [mu]m and severely worn areas were usually in the range 0.2-0.4 [mu]m R^sub a^. It was proposed, as above, that low level wear began on initial surface contact, progressing to stripe wear in contact areas of the component associated with high stresses and hence poorer lubrication.51 In the low wear regime, the mechanism appears to be polishing with occasional loss of surface grains by a grain boundary fracture mechanism (Fig. 14a). Stripe wear areas have sharply defined edges with a low wear region outside the stripe and more severe surface damage within (Fig. 14b and c). The spherical wear debris (0.1-0.5 [mu]m in diameter) seen in the pores left by grain removal was believed to be a result of a third body wear mechanism. The regions of severe wear were similar in appearance to the areas of stripe wear but on a larger scale. However, no transgranular fracture was observed in the severe wear regime. Some boundary lubrication (provided by adsorbed proteins), which gave the surfaces a degree of protection, was also proposed: for short contact durations, it is likely that the proteins could protect the surfaces; under harsher conditions such as rising from a seated position (squeeze lubrication effects), the boundary layer would not have been sufficient to protect the surface. Work on retrieved in vivo third generation alumina hip prostheses has also been published.46 Stripe wear due to edge loading was still visible and grain pull-out with fine scale debris trapped in the pits was revealed in the wear scar. In the centre of the scars, evidence of repolishing of the surface was observed. Wider stripes were seen on the femoral heads than on the acetabular cups.

Future of ceramic hip prostheses

Pure alumina hip prostheses have been used for more than 30 years and have been the dominant ceramic hip implant. The latest high performance alumina product provides a competitive solution of high reliability and excellent wear resistance. However, alumina is a brittle material and subject to a small but persistent probability of fracture.67 The challenge for ceramic engineering is to develop an improved material which maintains the advantageous properties of third generation alumina but allows new applications that require high mechanical load bearing capability.68

One approach to this goal is a composite material based on an alumina matrix with reinforcement phases to increase toughness and hardness, such as zirconia toughened alumina (ZTA). In 2000, CeramTec AG launched a mixed alumina-zirconia ceramic under the trade name Biolox-delta which contained Al^sub 2^O^sub 3^ (75 vol.- %), ZrO^sub 2^ (24 vol.-%) and mixed oxides (1 vol.-% CrO^sub 2^+ SrO).68 Compared with third generation alumina, this ZTA shows a potential doubling of strength relative to unreinforced alumina, although the hardness was slightly lower. Therefore, the door is open for new ceramics based on alumina.


On the basis of the critical review of research on alumina hip prostheses presented above, the following conclusions can be drawn:

1. The chronic problems with osteolysis caused by polymer wear debris from metal on polymer hip prostheses means that interest in alumina on alumina THRs continues to grow.

2. The wear performance of alumina hip joints is related both to the alumina itself (grain size, manufacturing process, etc.) and to the prosthesis design. Significant improvement of medical grade alumina has been achieved in Germany leading to the third generation aluminas used for current hip prostheses.

3. Investigations of alumina on alumina prostheses recovered from bodies show lower wear rates compared with other material combinations; however, occasional fractures are still experienced.

4. In vitro tests on alumina prostheses in joint simulators show excellent performance of third generation alumina prostheses, compared with the in vivo performance of first and second generation prostheses.

5. The introduction of a 'microseparation' movement between head and cap into the simulator test cycle has improved the ability to replicate in vivo service conditions and hence assess the performance of hip prostheses in vitro.

6. Microstructure plays an important role in the wear mechanisms of alumina and is a key issue in engineering new alumina ceramics for THRs. However, there is a fundamental lack of quantitative microstructural and mechanical property data for alumina prostheses, since most observations have been made by surgeons. Further investigations on hip prosthesis microstructures are required.

7. The next generation of ceramics for hip prostheses are likely to be based on alumina composites containing second phases to increase toughness, such as ZTA.


The author would like to thank Dr B. J. Inkson and Professor W. M. Rainforth for their support, guidance and encouragement. Studentship support from Overseas Research Scholarship and EPSRC is also gratefully acknowledged.


1. BestTreatments and NHS Direct: 'Hip replacement: an operation to replace your hip with an artificial one', October 2006, BMJ Publishing Group Limited, available at: http:// www.bestreatments.co.uk/btuk/pdf/18618.pdf (accessed June 2007).

2. W. H. Harris: 'Advanced concepts in total hip replacement'; 1985, Thorofare, Slack Inc.

3. D. Dowson: Proc. Inst. Mech. Eng. H, 2001, 215H, 335-358.

4. S. Sodha, J. P. Garino, M. Christians, B. Daniel and C. Faustin: Univ. Pennsylv. Orthop. J., 2001, 14, 1-4.

5. UK Science & Innovation: 'Artificial hip advances could keep boomers on the move', November 2004, available at: http:// www.britainusa.com/science/advanced_engineering_materials/ article.asp?a=7647 (accessed June 2007).

6. P. J. Firkins, J. L. Tipper, E. Ingham, M. H. Stone, R. Farrar and J. Fisher: J Biomech, 2001, 34, 1291-1298.

7. J. Charnley: Lancet, 1961, 1, 1129-1132.

8. L. Sedel, R. Nizard and P. Bizot: J. Bone Joint Surg. A, 2000, 82A, (10), 1519.

9. G. K. Mckee and J. Watson-Farrar: J. Bone Joint Surg. B, 1966, 48B, (2), 245.

10. M. Semlitsch and H. G. Willert: Proc. Inst. Mech. Eng. H, 1997, 211H, 73-88.

11. R. V. Joshi, N. S Eftekhar, D. J. Mcmahon and O. A. Nercessian: J. Bone Joint Surg. B, 1998, 80B, (4), 585-590.

12. H. Amstutz, P. Campbell, N. Kossovsky and I. Clarke: Clin. Orthop. Relat. Res., 1996, 276, 7-18.

13. M. Slonaker and T. Goswami: Mater. Des., 2004, 25, 395-405.

14. H. J. Fruh, G. Willmann and H. G. Pfaff: Biomaterials, 1997, 18, 873-876.

15. P. Boutin, P. Christel and J. M. Dorlot: J. Biomed Mater. Res., 1988, 22, 1203-1232.

16. P. Griss and G. Heimke: Arch. Orthop. Trauma Surg., 1981, 98, 157-164.

17. K. Knahr, M. Bohler, P. Frank, H. Plenk and M. Salzer: Arch. Orthop. Trauma Surg., 1987, 106, 297-300.

18. C. T. Trepte, E. F. Gauer and B. M. Gartner: Z. Orthop., 1985, 123, 239-244.

19. H. Mittelmeier and J. Heisel: Clin. Orthop., 1992, 282, 64- 72.

20. J. Heisel and E. Schmitt: Z. Orthop., 1987, 125, 480-490.

21. A. Walter and W. Plitz: in 'Ceramic in surgery', (ed. P. Vincenzini), 253-259; 1983, Amsterdam, Elsevier.

22. ISO 6474: 'Implants for surgery - Ceramic materials based on high purity alumina', 1984, Geneva, ISO.

23. L. Sedel, R. S. Nizard, L. Kerboull and J. Witvoet: Clin. Orthop. Relat. Res., 1994, 298, 175-183.

24. J. D'Antonio, W. Capello, M. Manley and B. Bierbaum: J. Arthropl, 2000, 17, (4), 390-397.

25. J. Nevelos, E. Ingham, C. Doyle, R. Streicher, A. Nevelos, W. Walter and J. Fisher: J. Arthropl., 2000, 15, (6), 793-795.

26. G. Willmann: J. Mater. Process. Technol., 1996, 56, 168-176.

27. G. Willmann: Clin. Orthop., 2000, 37, 22-28.

28. R. W. Davidge and F. L. Riley: Wear, 1995, 186-187, 45-49.

29. F. Xiong, R. R. Manory, L. Ward, M. Terheci and S. Lathabai: J. Am. Ceram. Soc., 1997, 80, 1310-1312.

30. W. M. Rainforth: J. Mater. Sci., 2004, 39, 6705-6721.

31. S. J. Cho, B. J. Hochey, B. R. Lawn and S. J. Bennison: J. Am. Ceram. Soc., 1989, 72, (7), 1249-1252.

32. R. W. Rice: Ceram. Eng. Sci. Proc., 1985, 16, (7-8), 940- 958.

33. B. Gueroult: J. Can. Ceram. Soc., 1994, 63, (2), 132-141.

34. M. Hamadonch, P. Boutin, J. Daussange, M. E. Bolander and L. Sedel: J. Bone Joint Surg. A, 2002, 84A, (1), 69-77.

35. J. F. O'Leary, T. H. Mallory, T. J. Kraus, A. V. Lombardi and C. L. Lye: J. Arthropl., 1988, 3, 87-96.

36. S. A. Hoffinger, K. J. Keggi and L. E. Zatorski: Orthopedics, 1991, 14, 523-531.

37. J. P. Garino: Clin. Orthop., 2000, 332, 41-47.

38. R. J. E. D. Higgs: J. Bone Joint Surg. B, 1990, 72B, 1101.

39. R. S. Nizard, L. Sedel and P. Christel: Clin. Orthop., 1992, 282, 53-63.

40. M. Winter, P. Griss, G. Scheller and T. Moser: Clin. Orthop., 1992, 282, 73-80.

41. A. Burckard and C. Berberat: Arch. Orthop. Trauma Surg., 1993, 112, 215-218.

42. E. W. Fritsch and M. Gleitz: Clin. Orthop., 1996, 328, 129- 136.

43. M. Boehler, H. Plenk and M. Salzer: Clin. Orthop., 2000, 332, 85-93.

44. P. Bizot, M. Larrouy, J. Witvoet, L. Sedel and R. Nizard: Clin. Orthop., 2000, 332, 134-142.

45. B. E. Bierbaum, J. Nairus, D. Kuesis, J. C. Morrison and D. Ward: Clin. Orthop., 2002, 334, 158-163.

46. W. L. Walter, G. M. Insley, W. K. Walter and M. A. Tuke: J. Arthropl., 2004, 19, (4), 402-413.

47. I. Bos and G. Willmann: J. Bone Joint Surg., 2001, 83B, (Suppl. II), 130.

48. R. Nizard, L. Sedel, D. Hannouche, M. Hamadouche and P. Bizot: J. Bone Joint Surg., 2005, 87B, (6), 755-758.

49. A. B. Nevelos, P. A. Evans, P. Harrison and M. Rainforth: Proc. Inst. Mech. Eng., 1993, 207, 155-162.

50. J. E. Nevelos, E. Ingham, C. Doyle, J. Fisher and A. B. Nevelos: Biomaterials, 1999, 20, 1833-1840.

51. J. P. Garino: Proc. BIOLOX Symp., 2005, 10, 157-168.

52. F. J. Kummer, S. A. Stuchin and V. H. Frankel: J. Arthropl., 1990, 5, 28-33.

53. J. L. Tipper, A. Hatton, J. E. Nevelos, E. Ingham, C. Doyle, R. Streicher, A. B. Nevelos and J. Fisher: Biomaterials, 2002, 23, 3441-3448.

54. T. R. Yoon, S. M. Rowe, S. T. Jung, K. J. Seon and W. J. Maloney: J. Bone Joint Surg. A, 1998, 80A, 1459-1468.

55. Y. S. Zhou, M. Ohashi and K. Ikeuchi: Wear, 1997, 210, 112- 119.

56. Y. Morita, K. Nakata and K. Ikeuchi: Wear, 2003, 254, 147- 153.

57. R. J. A. Bigsby, D. D. Auger, Z. M. Jin, D. Dowson, C. S. Hardaker and J. Fisher: J. Biomech., 1998, 31, 363-369.

58. J. L. Tipper, P. J. Firkins, A. A. Besong, P. S. M. Barbour, J. Nevelos, M. H. Stone, E. Ingham and J. Fisher: Wear, 2001, 250, 120-128.

59. S. Affatato, G. Bersaglia, I. Foltran, D. Emiliani, F. Traina and A. Toni: Wear, 2004, 256, 400-405.

60. A. V. Lombardi, T. H. Mallory, D. A. Dennis, R. D. Komistek, R. A. Fada and E. J. Northcut: J. Arthropl., 2000, 15, (6), 702- 709.

61. J. E. Nevelos, F. Prudhommeaux, M. Hamadouche, C. Doyle, E. Ingham, A. Meunier, A. B. Nevelos, L. Sedel and J. Fisher: J. Bone Joint Surg. B, 2001, 83B, (4), 598-603. 62. T. Stewart, J. Tipper, R. Streicher, E. Ingham and J. Fisher: J. Mater. Sci., Mater. Med, 2001, 12, 1053-1056.

63. T. D. Stewart, S. Williams, J. L. Tipper, E. Ingham, M. H. Stone and J. Fisher: Proc. 29th Leeds-Lyon Symp. on 'Tribology', 291- 296; 2003, Amsterdam, Elsevier.

64. W. M. Rainforth: Ceram. Int., 1996, 22, 365-372.

65. J. M. Dorlot, P. Christel and A. Meunier: J. Biomed Mater. Res., 1989, 23, 299.

66. J. M. Dorlot: Clin. Orthop. Relat. Res., 1992, 282, 47-52.

67. P. Merkert and M. Kuntz: Proc. BIOLOX Symp., 2006, 11, 283- 288.

68. I. C. Clarke, D. Green, P. Williams, T. Donaldson and G. Pezzotti: Proc. BIOLOX Symp., 2006, 11, 189-205.

P. Zeng*

Department of Engineering Materials, University of Sheffield, Mappin Street, Sheffield S1 3JD, UK

* Corresponding author, email [email protected]

Copyright Institute of Materials May 2008

(c) 2008 Materials Science and Technology; MST. Provided by ProQuest Information and Learning. All rights Reserved.