**3.2.9 Optimum method for cross-linking**

182 Recent Advances in Arthroplasty

better large-strain mechanical properties than those irradiated at room temperature with

Although ultra-high-molecular-weight polyethylene (UHMWPE) has low friction and dampening properties, it has one major disadvantage, which is adhesive and abrasive wear. The UHMWPE particles produced from cyclical loading are thought to play a major role in particle-induced osteolysis and secondary implant failure and loosening. Osteolysis results from the intrusion of polyethylene debris between the implant surface and bone, inducing a macrophage response. Polyethylene also can fail as a result of third-body wear secondary to polyethylene degradation. A polyethylene wear rate of 0.10 mm per year is the threshold for the development of osteolysis. In addition, conventional polyethylene thickness should not decrease below 6 to 8 mm, or accelerated wear polyethylene failure and osteolysis will

Highly cross-linked polyethylene was introduced to reduce the polyethylene wear rate in THA.(McKellop 1999) It is well documented that young age, male gender, and high activity level increase the risk of wear, osteolysis, and mechanical failure. However, these different factors that influence wear rate in conventional polyethylene have no significant influence when using highly cross-linked polyethylene.(Rohrl 2007) In vitro studies on highly crosslinked polyethylene have shown that wear can be reduced by 42% to 100% compared with that of conventional polyethylene, and, multiple in vivo studies support many of these in vitro results. The annual linear wear for highly cross-linked polyethylene has been reported to be 45% that of the conventional liner at 5 years after implantation. This wear rate is well below the threshold for lysis. Long-term results (10 to 20 years) of highly cross-linked polyethylene support that the low wear rates reported approach that of MOM and ceramic-

Today, polyethylenes are sterilized by different techniques with regard to the fact that sterilization process can affect the physical properties of polyethylene. Some brands of the polyethylene components are now sterilized without irradiation, using either ethylene oxide or gas plasma, in order to minimize the oxidative degradation. Because these methods do not generate free radicals in the polyethylene, they completely avoid the potential for immediate and long-term oxidative degradation of the mechanical properties and wear resistance. However, because ethylene oxide or gas-plasma do not induce cross-linking, these polyethylenes do not take advantage of improving the wear resistance following crosslinking.(Digas 2003) Other brands of polyethylenes are still sterilized with gamma radiation, but for doing so, the polyethylene components are sealed in some type of low-oxygen atmosphere, including vacuum, inert gas, or with an oxygen scavenger. In addition, one manufacturer anneals the polyethylene acetabular cups after sterilization by heating them in the nitrogen packaging at 37° to 50°C (well below the melt temperature of 135°C to avoid distorting the components) for about 6 days to reduce the level of residual free radicals that were induced by the gamma radiation. These modifications in sterilization techniques can markedly reduce but not necessarily eliminate the oxidation that would otherwise occur

identical radiation dose levels.(Affatato 2005)

**3.2.6 Conventional polyethylene surfaces** 

**3.2.7 Highly cross-linked polyethylene surfaces** 

on-ceramic bearings.(Digas 2007; Atienza 2008)

**3.2.8 Sterilization techniques** 

develop.(Lundberga 2006)

In addition to the optimum method of cross-linking, the optimum amount of cross-linking to use is also a subject of current debate. Because increasing the level of crosslinking causes a progressively greater reduction in some mechanical properties, such as ultimate strength, ductility, fracture toughness, and fatigue strength, one extreme is to avoid cross-linking altogether (for example, by simply sterilizing with ethylene oxide or gas plasma). This avoidance is to retain the maximum values of strength, elongation, and fracture toughness, despite the fact that it results in substantially higher wear of the polyethylene. Among those who advocate crosslinking, the particular dose used represents that manufacturer's approach to balancing reduced wear against the need to maintain other mechanical properties well above that needed for acceptable clinical performance. Those using the high levels of cross-linking (9.5 to 11 Mrads), about 3 to 4 times the typical dose used historically to sterilize polyethylene components, maintain that these high levels are justified to obtain the additional 5% to 10% improvement in wear over that provided by a moderate dose, despite the corresponding reduction in other physical properties. In contrast, advocates of a moderate cross-linking dose, such as 5 Mrads, maintain that the corresponding reduction in wear to 85% below that of a noncross-linked polyethylene will be sufficient to avoid an osteolytic reaction in even the most active patients, without unnecessarily reducing other physical properties. Clearly, it is not desirable to use a dose that will result in mechanical failure of the polyethylene components.(Moore 2008; Bradford 2004)

### **3.3 Metal-on-metal bearings**

The first widely used total hip replacements featured cobalt-chromium alloy bearing against itself, primarily the McKee-Farrar (Howmedica, Limerick, Ireland) design, along with the Mueller (Sulzer AG, Winterthur, Switzerland), Ring (Downs, Ltd, Mitchham, England) and others. Because of a relatively high rate of early failure, the first-generation metal-on-metal hips were largely supplanted by the Charnley prosthesis, which featured a stainless steel ball and a polyethylene socket. Disregarding the early failures, the long-term survivorship of the early metal-on-metal designs has been comparable to that of the metal-onpolyethylene Charnley. In particular, the steady-state wear rates have been on the order of a

The Bearing Surfaces in Total Hip Arthroplasty – Options, Material Characteristics and Selection 185

vitro study, Affatato et al. demonstrated that femoral heads of 36 mm in diameter work in the mixed-lubrication regime (λ > 1) and 28-mm diameter heads were associated with a λ ratio of <1, thus showing their aptitude to work in the boundary-lubrication regime, with substantially higher volume depletion due to wear. With the 32-mm heads, the λ ratio varied in a range around 1, suggesting the possibility of their operating in the mixedlubrication regime.(Affatato 2008) Hip-joint-simulator tests on metal-on-metal total hip replacements have shown a transition from boundary to mixed lubrication as the diameter of the femoral head increased. An increase in diameter up to 54 mm resulted in a marked reduction in wear rates, which was attributable to growing support from the fluid-film action in a mixed-lubrication regime. Reducing the clearance between the femoral and acetabular components also reduces the amount of wear debris. Significant reduction in the wear debris associated with large-diameter (54-mm) heads has been noted when clearances were reduced. When clearances are designed, cup flexibility needs to be taken into account so that the minimum clearance is always more than the maximum flexibility of the cup. Thick (8 to 12-mm) chromium nitride (CrN) and chromium carbon nitride (CrCN) coating leads to twenty-two-fold lower wear rates as compared with those associated with standard metal-on-metal couples. Also, CrN wear particles were found to be less cytotoxic than standard metal-on-metal bearings when co-cultured with macrophage and fibroblast cells. The substantially lower ion concentration released by these surface-engineered components along with the reduced wear volumes and the reduced cytotoxicity per unit volume of wear points to the potential for clinical application of this technology. However, there is concern

The interplay of material(s), macrogeometry (diameter and clearance), microgeometry (surface topography), and lubrication influences the wear of MOM-bearing THRs to a far greater degree than MOP-bearing ones. Alloys of cobalt (Co) and chromium (Cr) have been preferred for MOM bearings in THR because of their hardness. High chromium content provides good corrosion resistance. Cast Co-Cr-Mo alloy, which was used to manufacture the majority of first-generation MOM hips, has a relatively high carbon content of 0.2% to 0.3% and contains primarily Cr and Mo carbides, which result in asperities on the polished surface. The macrogeometry of a MOM bearing can be described in terms of the diameter(s) of the ball and the socket and the clearance of the resultant bearing couple. Clearance is the size of the gap between the surfaces at the equator of the bearing. For hemispherical bearings, clearance is a function of the difference in the diameters of the surfaces of the ball and socket. A ball and socket of exactly equal diameter mated together would have zero clearance and a maximal contact area for that size bearing. Contact area can be increased by increasing the size (diameter) of the bearing surfaces and/or by decreasing the clearance. Conversely, for a given diameter, increasing clearance decreases contact area. Contact stresses are a function of material properties and are inversely proportional to contact area. Clearance also influences lubrication, as the size of the gap has implications for the amount and type of lubrication. Smaller clearances encourage fluid film lubrication. Large clearances lead to a reduced contact area, loss of effective lubrication, and more rapid wear.(Reiker 2005) However, too little clearance may lead to equatorial contact, very high frictional

about the durability of this coating.(Williams 2004)

forces, high torque, and loosening of the implant.(Chan 1996)

**3.3.2 Materials and design considerations** 

few micrometers per year compared with the average of 100 to 200/µm of polyethylene wear per year typically reported for metal-on-polyethylene hips. In view of the growing awareness of the problem of extensive osteolysis caused by polyethylene wear debris, a number of second-generation metal-on-metal implants have been developed, including conventional total hips and surface replacements. The first to be widely used clinically was the Metasul (Sulzer AG, Winterthur, Switzerland) hip. Hip simulator studies and clinical retrievals of modern metal-on-metal bearings also typically have shown steady-state wear rates on the order of a few micrometers per million cycles (with one million cycles being the equivalent of about 1 year's use in a patient of average activity). It is also apparent that metal-on-metal implants have the ability to self-heal, that is, to polish-out isolated surface scratches caused by third-body particles or subluxation damage. The overall clinical performance of second-generation metal-on-metal hips to date has been comparable to that of conventional metal-on-polyethylene hips. Clinical and laboratory wear studies have indicated that metal-on-metal implants often exhibit 10 to 20 times greater wear-in during the initial 1 to 2 years of clinical use, or one to two million cycles in a hip simulator. In addition, some first-generation and second-generation metal-on-metal hips have exhibited extensive surface micropitting, possibly due to a fatigue-corrosion mechanism associated with the smaller carbides.

### **3.3.1 Tribology**

The principle constituents of alloys for MOM bearings are cobalt, chromium, and molybdenum, but they can be differentiated as being either high carbon (0.20% to 0.25%) or low carbon (0.05% to 0.08%), with the former usually demonstrating lower in vitro wear rates. MOM hip bearings produce wear particles by some combination of the four classic mechanisms of adhesion, abrasion, corrosion, and surface fatigue, with an emphasis on abrasion and surface fatigue.(Wimmer 2003) Wear in metal-on-metal total hip arthroplasty implants is known to consist of two distinct phases. A relatively high-wear running-in phase, which lasts for between 0.5 and 2 x 106 cycles, is followed by a steady-state phase, during which the wear rate is constant and much lower. This decrease in wear has been demonstrated to be secondary to the so-called self-polishing effect of metal-on-metal bearing surfaces. Opposing metallic bearing surfaces are thought to run-in in the contact zone as surface asperities are polished or removed, improving the initial surface finish, with a commensurate reduction in the wear rate as the lubricating fluid film becomes more coherent. Metal-on-metal total hip arthroplasty implants can operate in the mild mixedlubrication regime in which much of the applied load is supported by elastohydrodynamic films. Correct tribological design leads to remarkably low steady-state wear rates. Promotion of the most effective elastohydrodynamic films calls for the largest possible head diameters and the smallest clearances that can be reasonably adopted consistent with fine surface finishes, good sphericity, polar contact, and minimal structural elastic deformation of the cup on its foundations.(Udofia 2003) With current manufacturing techniques, a surface roughness of 5 to 15 nm is typically achieved. This results in improved sphericity and surface finish. The presence of depressions or holes (rather than scratches) with smooth (rather than sharp) edges seems to improve the lubrication and wettability properties. The wear of metal-on-metal bearings is affected by various design parameters, such as the clearance or surface roughness. The predictive role of the λ ratio in the evaluation of wear behavior of metal-on-metal total hip replacements has also been well documented. In an in vitro study, Affatato et al. demonstrated that femoral heads of 36 mm in diameter work in the mixed-lubrication regime (λ > 1) and 28-mm diameter heads were associated with a λ ratio of <1, thus showing their aptitude to work in the boundary-lubrication regime, with substantially higher volume depletion due to wear. With the 32-mm heads, the λ ratio varied in a range around 1, suggesting the possibility of their operating in the mixedlubrication regime.(Affatato 2008) Hip-joint-simulator tests on metal-on-metal total hip replacements have shown a transition from boundary to mixed lubrication as the diameter of the femoral head increased. An increase in diameter up to 54 mm resulted in a marked reduction in wear rates, which was attributable to growing support from the fluid-film action in a mixed-lubrication regime. Reducing the clearance between the femoral and acetabular components also reduces the amount of wear debris. Significant reduction in the wear debris associated with large-diameter (54-mm) heads has been noted when clearances were reduced. When clearances are designed, cup flexibility needs to be taken into account so that the minimum clearance is always more than the maximum flexibility of the cup.

Thick (8 to 12-mm) chromium nitride (CrN) and chromium carbon nitride (CrCN) coating leads to twenty-two-fold lower wear rates as compared with those associated with standard metal-on-metal couples. Also, CrN wear particles were found to be less cytotoxic than standard metal-on-metal bearings when co-cultured with macrophage and fibroblast cells. The substantially lower ion concentration released by these surface-engineered components along with the reduced wear volumes and the reduced cytotoxicity per unit volume of wear points to the potential for clinical application of this technology. However, there is concern about the durability of this coating.(Williams 2004)

### **3.3.2 Materials and design considerations**

184 Recent Advances in Arthroplasty

few micrometers per year compared with the average of 100 to 200/µm of polyethylene wear per year typically reported for metal-on-polyethylene hips. In view of the growing awareness of the problem of extensive osteolysis caused by polyethylene wear debris, a number of second-generation metal-on-metal implants have been developed, including conventional total hips and surface replacements. The first to be widely used clinically was the Metasul (Sulzer AG, Winterthur, Switzerland) hip. Hip simulator studies and clinical retrievals of modern metal-on-metal bearings also typically have shown steady-state wear rates on the order of a few micrometers per million cycles (with one million cycles being the equivalent of about 1 year's use in a patient of average activity). It is also apparent that metal-on-metal implants have the ability to self-heal, that is, to polish-out isolated surface scratches caused by third-body particles or subluxation damage. The overall clinical performance of second-generation metal-on-metal hips to date has been comparable to that of conventional metal-on-polyethylene hips. Clinical and laboratory wear studies have indicated that metal-on-metal implants often exhibit 10 to 20 times greater wear-in during the initial 1 to 2 years of clinical use, or one to two million cycles in a hip simulator. In addition, some first-generation and second-generation metal-on-metal hips have exhibited extensive surface micropitting, possibly due to a fatigue-corrosion mechanism associated

The principle constituents of alloys for MOM bearings are cobalt, chromium, and molybdenum, but they can be differentiated as being either high carbon (0.20% to 0.25%) or low carbon (0.05% to 0.08%), with the former usually demonstrating lower in vitro wear rates. MOM hip bearings produce wear particles by some combination of the four classic mechanisms of adhesion, abrasion, corrosion, and surface fatigue, with an emphasis on abrasion and surface fatigue.(Wimmer 2003) Wear in metal-on-metal total hip arthroplasty implants is known to consist of two distinct phases. A relatively high-wear running-in phase, which lasts for between 0.5 and 2 x 106 cycles, is followed by a steady-state phase, during which the wear rate is constant and much lower. This decrease in wear has been demonstrated to be secondary to the so-called self-polishing effect of metal-on-metal bearing surfaces. Opposing metallic bearing surfaces are thought to run-in in the contact zone as surface asperities are polished or removed, improving the initial surface finish, with a commensurate reduction in the wear rate as the lubricating fluid film becomes more coherent. Metal-on-metal total hip arthroplasty implants can operate in the mild mixedlubrication regime in which much of the applied load is supported by elastohydrodynamic films. Correct tribological design leads to remarkably low steady-state wear rates. Promotion of the most effective elastohydrodynamic films calls for the largest possible head diameters and the smallest clearances that can be reasonably adopted consistent with fine surface finishes, good sphericity, polar contact, and minimal structural elastic deformation of the cup on its foundations.(Udofia 2003) With current manufacturing techniques, a surface roughness of 5 to 15 nm is typically achieved. This results in improved sphericity and surface finish. The presence of depressions or holes (rather than scratches) with smooth (rather than sharp) edges seems to improve the lubrication and wettability properties. The wear of metal-on-metal bearings is affected by various design parameters, such as the clearance or surface roughness. The predictive role of the λ ratio in the evaluation of wear behavior of metal-on-metal total hip replacements has also been well documented. In an in

with the smaller carbides.

**3.3.1 Tribology** 

The interplay of material(s), macrogeometry (diameter and clearance), microgeometry (surface topography), and lubrication influences the wear of MOM-bearing THRs to a far greater degree than MOP-bearing ones. Alloys of cobalt (Co) and chromium (Cr) have been preferred for MOM bearings in THR because of their hardness. High chromium content provides good corrosion resistance. Cast Co-Cr-Mo alloy, which was used to manufacture the majority of first-generation MOM hips, has a relatively high carbon content of 0.2% to 0.3% and contains primarily Cr and Mo carbides, which result in asperities on the polished surface. The macrogeometry of a MOM bearing can be described in terms of the diameter(s) of the ball and the socket and the clearance of the resultant bearing couple. Clearance is the size of the gap between the surfaces at the equator of the bearing. For hemispherical bearings, clearance is a function of the difference in the diameters of the surfaces of the ball and socket. A ball and socket of exactly equal diameter mated together would have zero clearance and a maximal contact area for that size bearing. Contact area can be increased by increasing the size (diameter) of the bearing surfaces and/or by decreasing the clearance. Conversely, for a given diameter, increasing clearance decreases contact area. Contact stresses are a function of material properties and are inversely proportional to contact area. Clearance also influences lubrication, as the size of the gap has implications for the amount and type of lubrication. Smaller clearances encourage fluid film lubrication. Large clearances lead to a reduced contact area, loss of effective lubrication, and more rapid wear.(Reiker 2005) However, too little clearance may lead to equatorial contact, very high frictional forces, high torque, and loosening of the implant.(Chan 1996)

The Bearing Surfaces in Total Hip Arthroplasty – Options, Material Characteristics and Selection 187

and promotes the formation of a passive oxide layer, whereas molybdenum is included in the alloy for corrosion resistance. The size and distribution of the carbide phase contributes to the hardness and mechanical behavior of the alloy. The properties of CoCr alloys depend upon their processing history—whether fabricated by investment casting or by a thermomechanical forming process. During an investment casting process, a wax replica of the prosthesis is first fabricated and coated with a silicate slurry. The replica and slurry are raised above the melting point of the wax (100–150oC) and burned away in a furnace, leaving only the slurry mold with an interior cavity having the dimensions of the replica. The cavity is then filled with molten alloy and, after cooling, the mold is broken to remove the cast implant. Mechanical forming, on the other hand, can be accomplished by a wide range of thermomechanical manufacturing methods, including rolling, forging, swaging, and drawing, all of which lead to a wrought CoCr alloy. Both cast and wrought CoCr alloys are subjected to annealing heat treatments for optimization of microstructure and mechanical properties. Historically, the first-generation metal-on-metal components were fabricated by casting(Hernandez-Rodriguez 2005), whereas contemporary secondgeneration metal-on-metal implants are manufactured from wrought CoCr alloys. (Barnes

Wrought CoCr alloys have significantly enhanced hardness, yield strength, and ultimate properties as compared with cast alloys, resulting from their more uniform carbide microstructure. All CoCr alloys used in orthopaedics contain trace constituents of carbon (0.35%), nitrogen (max 0.25%), nickel (max 1%), iron (max 0.75%), silicon (max 1.0%), and manganese (max 1.0%). Cast CoCr alloys may also contain tungsten (max 0.20%), phosphorus (max 0.020%), sulfur (max 0.010%), aluminum (max 0.30%), and boron (max 0.01%). Of these secondary constituents, the effect of carbon content on CoCr alloys has been examined most closely by researchers in the orthopaedic community because of the link between carbon content, carbide formation, and properties of the bearing surface. CoCr alloys may be further classified as either low-carbon (<0.05%) or high-carbon (>0.20%). The majority of contemporary metal-on-metal bearings are fabricated from a high-carbon CoCr alloy. In contemporary hip simulator studies, high- and low-carbon CoCr alloys have

Ion levels have been measured in whole blood, serum, and erythrocytes. Metal ions from hip implant materials include the following: titanium; vanadium; chromium; cobalt; and molybdenum. In reports of metal ion concentrations in urine, for maximal accuracy and to minimize the effect of variations in the level of hydration that may occur throughout the day, a 24-hour urine collection should be performed. Healthy controls have been demonstrated to have a mean serum cobalt level of 0.24 mg/L (median, 0.23 mg/L; range, 0.08 to 0.50 mg/L) and a mean serum chromium level of 0.28 mg/L (median, 0.29 mg/L; range, 0.06 to 0.93 mg/L). Currently there are two methods for measuring trace metals: inductively coupled plasma mass spectroscopy (ICP-MS) and graphite furnace atomic absorption spectrophotometry (GFAAS), with the former method being favored because of higher sensitivity to detect low levels of metal ions, allowing multielement determinations.(Engh 2009) Numerous studies have reported an increase in the levels of metal ions in serum, urine, and blood of patients with MOM implants. However, serum/plasma levels of cobalt and chromium ions appeared to be higher than those in red blood cells, suggesting that serum may reflect better the true ion levels when using blood to

exhibited comparable wear mechanisms and wear rates. (Isaac 2006)

2008)

**3.3.4.1 Metal ions** 

The limitations of current mass production manufacturing set the lower limit of clearance at about 20 µm. Wear rate, particularly during the initial run-in, increases rapidly with clearances above about 150 µm. In addition to the contact area, another important variable is where the contact occurs. Given bearings of equivalent diameters, equatorial contact is associated with higher frictional torques than the same contact area in a more polar location. Equatorial bearing may have been a factor associated with failure of some early MOM THRs, and this is supported by retrieval studies. Consequently, relatively polar contact is preferred.(Dowson 2004; Isaac 2006) (Fig.9)

Fig. 9. Too much reduction in the "clearance" can result in conversion a polar contact to an equatorial contact resulting in the articulation seizing

### **3.3.3 Lubrication**

The lubricating fluid and conditions are important variables which influence friction and wear. The synovial fluid of normal, osteoarthritic, and rheumatoid joints has been characterized to some degree. The fluid is thixotropic; viscosity is a function of the shear strain rate, or velocity of motion, for practical purposes. Less is known about the fluid formed around total joint arthroplasties, but the composition and rheology are likely variable. A favorable film thickness to surface roughness ratio (λ ratio) is desirable in order to maintain low friction between the articulating surfaces. Mixed-film lubrication appears to be the operative mechanism in most MOM hip joints. For a given load and surface velocity, fluid film thickness is dependent on the properties of the fluid but can be influenced by the properties of the bearing materials, the macrogeometry of the bearing (which is a function of diameter and clearance), and the surface microtopography (surface finish). (Affatato 2008) Within the apparent contact area, the surfaces make actual contact only at the tips of asperities, and a lubricant film can influence wear significantly. As wear proceeds, the contact area at the asperity tips increases, and such "running-in" can produce a more favorable microgeometry for lubricant films to separate the surfaces and reduce wear. Fluidfilm lubrication is encouraged by making the femoral head as large as practically possible and the clearance as small as practically possible. For MOM bearings, in distinction from PE bearings, larger diameters can actually produce lower wear rates for similar manufacturing parameters.(Smith 2001a; Chan 1999)

### **3.3.4 Composition and microstructure of CoCr alloys**

CoCr alloys consist of a primary cobalt alloy matrix phase and a secondary metal carbide phase. In the cobalt-based matrix, chromium enhances the mechanical properties of the alloy and promotes the formation of a passive oxide layer, whereas molybdenum is included in the alloy for corrosion resistance. The size and distribution of the carbide phase contributes to the hardness and mechanical behavior of the alloy. The properties of CoCr alloys depend upon their processing history—whether fabricated by investment casting or by a thermomechanical forming process. During an investment casting process, a wax replica of the prosthesis is first fabricated and coated with a silicate slurry. The replica and slurry are raised above the melting point of the wax (100–150oC) and burned away in a furnace, leaving only the slurry mold with an interior cavity having the dimensions of the replica. The cavity is then filled with molten alloy and, after cooling, the mold is broken to remove the cast implant. Mechanical forming, on the other hand, can be accomplished by a wide range of thermomechanical manufacturing methods, including rolling, forging, swaging, and drawing, all of which lead to a wrought CoCr alloy. Both cast and wrought CoCr alloys are subjected to annealing heat treatments for optimization of microstructure and mechanical properties. Historically, the first-generation metal-on-metal components were fabricated by casting(Hernandez-Rodriguez 2005), whereas contemporary secondgeneration metal-on-metal implants are manufactured from wrought CoCr alloys. (Barnes 2008)

Wrought CoCr alloys have significantly enhanced hardness, yield strength, and ultimate properties as compared with cast alloys, resulting from their more uniform carbide microstructure. All CoCr alloys used in orthopaedics contain trace constituents of carbon (0.35%), nitrogen (max 0.25%), nickel (max 1%), iron (max 0.75%), silicon (max 1.0%), and manganese (max 1.0%). Cast CoCr alloys may also contain tungsten (max 0.20%), phosphorus (max 0.020%), sulfur (max 0.010%), aluminum (max 0.30%), and boron (max 0.01%). Of these secondary constituents, the effect of carbon content on CoCr alloys has been examined most closely by researchers in the orthopaedic community because of the link between carbon content, carbide formation, and properties of the bearing surface. CoCr alloys may be further classified as either low-carbon (<0.05%) or high-carbon (>0.20%). The majority of contemporary metal-on-metal bearings are fabricated from a high-carbon CoCr alloy. In contemporary hip simulator studies, high- and low-carbon CoCr alloys have exhibited comparable wear mechanisms and wear rates. (Isaac 2006)

### **3.3.4.1 Metal ions**

186 Recent Advances in Arthroplasty

The limitations of current mass production manufacturing set the lower limit of clearance at about 20 µm. Wear rate, particularly during the initial run-in, increases rapidly with clearances above about 150 µm. In addition to the contact area, another important variable is where the contact occurs. Given bearings of equivalent diameters, equatorial contact is associated with higher frictional torques than the same contact area in a more polar location. Equatorial bearing may have been a factor associated with failure of some early MOM THRs, and this is supported by retrieval studies. Consequently, relatively polar contact is

Fig. 9. Too much reduction in the "clearance" can result in conversion a polar contact to an

The lubricating fluid and conditions are important variables which influence friction and wear. The synovial fluid of normal, osteoarthritic, and rheumatoid joints has been characterized to some degree. The fluid is thixotropic; viscosity is a function of the shear strain rate, or velocity of motion, for practical purposes. Less is known about the fluid formed around total joint arthroplasties, but the composition and rheology are likely variable. A favorable film thickness to surface roughness ratio (λ ratio) is desirable in order to maintain low friction between the articulating surfaces. Mixed-film lubrication appears to be the operative mechanism in most MOM hip joints. For a given load and surface velocity, fluid film thickness is dependent on the properties of the fluid but can be influenced by the properties of the bearing materials, the macrogeometry of the bearing (which is a function of diameter and clearance), and the surface microtopography (surface finish). (Affatato 2008) Within the apparent contact area, the surfaces make actual contact only at the tips of asperities, and a lubricant film can influence wear significantly. As wear proceeds, the contact area at the asperity tips increases, and such "running-in" can produce a more favorable microgeometry for lubricant films to separate the surfaces and reduce wear. Fluidfilm lubrication is encouraged by making the femoral head as large as practically possible and the clearance as small as practically possible. For MOM bearings, in distinction from PE bearings, larger diameters can actually produce lower wear rates for similar manufacturing

CoCr alloys consist of a primary cobalt alloy matrix phase and a secondary metal carbide phase. In the cobalt-based matrix, chromium enhances the mechanical properties of the alloy

preferred.(Dowson 2004; Isaac 2006) (Fig.9)

equatorial contact resulting in the articulation seizing

parameters.(Smith 2001a; Chan 1999)

**3.3.4 Composition and microstructure of CoCr alloys** 

**3.3.3 Lubrication** 

Ion levels have been measured in whole blood, serum, and erythrocytes. Metal ions from hip implant materials include the following: titanium; vanadium; chromium; cobalt; and molybdenum. In reports of metal ion concentrations in urine, for maximal accuracy and to minimize the effect of variations in the level of hydration that may occur throughout the day, a 24-hour urine collection should be performed. Healthy controls have been demonstrated to have a mean serum cobalt level of 0.24 mg/L (median, 0.23 mg/L; range, 0.08 to 0.50 mg/L) and a mean serum chromium level of 0.28 mg/L (median, 0.29 mg/L; range, 0.06 to 0.93 mg/L). Currently there are two methods for measuring trace metals: inductively coupled plasma mass spectroscopy (ICP-MS) and graphite furnace atomic absorption spectrophotometry (GFAAS), with the former method being favored because of higher sensitivity to detect low levels of metal ions, allowing multielement determinations.(Engh 2009) Numerous studies have reported an increase in the levels of metal ions in serum, urine, and blood of patients with MOM implants. However, serum/plasma levels of cobalt and chromium ions appeared to be higher than those in red blood cells, suggesting that serum may reflect better the true ion levels when using blood to

The Bearing Surfaces in Total Hip Arthroplasty – Options, Material Characteristics and Selection 189

It is important to recognize that in modern THR prostheses there may be several sources of metal particle and ion generation.(Savarino 2006) Systemic dissemination of soluble and particulate corrosion products from modular junctions has been described, including the presence of metallic particles in the lymph nodes, liver, and spleen. In modular total hips with a MOP bearing, with no or mild corrosion at the modular head and neck junction, serum and urine levels of Co averaged 0.94 ppb (range <0.54 to 1.65) and 0.92 ppb (range <0.3 to 1.14), respectively, and urine Cr levels averaged 1.0 ppb (range 0.54 to 1.92). With moderate or severe corrosion at the modular head and neck junction, the serum and urine levels of Co averaged 1.06 ppb (range 0.8 to 1.4) and 0.87 ppb (range <0.3 to 1.3), respectively, and the average urine Cr levels were 1.59 ppb (range 0.6 to 3.0).(Jacobs 1998) The levels of metal ions in serum and urine are further elevated in patients with MOM bearings. It appears that the ion levels are higher in the short term and decrease over time. This is consistent with a conditioning phase or running-in of the bearing. The ion levels in subjects with Co-Cr alloy MOM THRs that had been in situ for an average of 24 years have been reported. The average serum Co level was 0.9 ppb (range <0.3 to 2.0). The average serum Cr level was 1.28 ppb (range 0.21 to 2.56), and the average urine Cr level was 1.22 ppb (range 0.26 to 2.59).(Bitsch 2007) The longer-term serum and urine ion levels produced by MOM bearings are not much higher than those produced by the modular junctions of femoral components. Unfortunately, the toxicological importance of these trace metal elevations has not been established, and available data do not answer questions regarding the risks of ion hypersensitivity, toxicity, and carcinogenesis. Since wear of a MOM bearing cannot generally be measured on a radiograph, serum and urine metal ion concentrations may be useful indicators of patient activity and the tribological performance of these bearings.(MacDonald 2004) The distribution and histological effects of wear particles from a McKee-Farrar THR with an in vivo use of nearly 30 years has been described. Clinically, the implant was functioning well. Serum and urine levels of chromium and serum levels of cobalt, obtained 25 years after implantation and while the patient was healthy and active, were found to be 1.02 ppb, 0.51 ppb, and 0.66 ppb, respectively. Sections from multiple samples of the lymph nodes, spleen, liver, and kidney were examined with light microscopy. This tissue analysis did not reveal any evidence of end-organ damage or accumulation of metal particles. It appears that a patient with normal renal function is capable of clearing cobalt and chromium ions from his or her system. Potential additional

sources of ions include the acetabular morse taper connection.(Brodner 2003)

Although the periprosthetic tissue reactions surrounding MOM THA demonstrate some histological characteristics similar to those typically seen around MOPE THAs, they have been reported to be less intense with fewer histiocytes present in the tissues. It is still uncertain whether this difference can be attributed to variations in particle concentration, size, shape, or composition. The wear particles produced by MOM implants in vitro and in vivo are in the nanometer size range, mostly round to oval but with the presence of some needle-shaped particles as well, mainly depending on the number of cycles (for in vitro particles) and the implantation time (for in vivo particles). In addition, most of the particles contain Chromium (Cr) and Oxygen (O) but no Cobalt (Co) and are therefore most likely chromium oxides. Therefore, in addition to focusing on CoCrMo particles, attention to the effects of chromium oxide particles should be addressed in future in vitro and in vivo biologic studies. Finally, because of their nanometer size, those particles have a high surface

**3.3.4.2 Biological response** 

assess systemic ion levels.(Daniel 2007; Antoniou 2008) In a study looking specifically at the serum cobalt concentrations in patients during the first 5 years after an MOM THA, Brodner and colleagues showed a moderate concentration of 1 µg/L at 1 year and 0.7 µg/L at 5 years. They did not find any significant difference in the levels from 3 to 12 months and the subsequent measurements and, therefore, concluded that the serum cobalt concentrations did not reflect the higher run-in wear of MM implants.(Brodner 2003) In modern MM surface arthroplasty, Skipor and colleagues reported elevations in serum chromium that were 22-fold, 23-fold, and 21-fold higher at 3, 6, and 12 months postoperatively, respectively, than preoperatively. The corresponding serum cobalt levels were eightfold, sevenfold, and sixfold higher than preoperative levels, and the authors reported that the values seen with the current generation of surface arthroplasties were in the same range as those observed in association with conventional MOM THA.(Skipor 2002) Comparison of serum cobalt and chromium ion levels in patients with a MOM compared with a MOPE bearing and a MOM compared with a ceramic-ceramic bearing showed that there were always higher ion levels in the patients with the MOM bearings. In a study analyzing the whole blood cobalt and chromium ion levels in 68 patients, Hart and colleagues reported a decrease in cytotoxic CD8+ T-lymphocytes, which are involved in the defense against intracellular pathogens and cancerous cells.(Hart 2009) Of interest, Back and colleagues noted a significant decrease in the cobalt levels at 2 years compared with the levels at 6 months, as well as a decrease in the chromium levels at 2 years compared with the levels at 9 months. When looking at the effects of the size of the femoral head and the orientation of the acetabular component on the overall concentrations of metal ions in the whole blood of patients with MOM resurfacings, Langton et al. reported higher ion concentrations in patients with smaller components and when the inclination angle was greater than 45 degrees and the anteversion angle exceeded 20 degrees.(Langton 2008)

Wear particles from MOM bearings are nanometers in linear dimension, which is substantially smaller than PE wear particles. Light microscopic analysis of tissue obtained from around MOM joints showed particles with variable and irregular shape. The size of metal particles reported by scanning electron microscopy studies ranges from 0.1 to 5 µm. The studies have suggested that large metallic particles observed with light microscopy were agglomerates of the smaller particles. Transmission electron microscopy has demonstrated wear particles from Co-Cr-Mo bearings to be round to oval in shape with irregular boundaries. Most of the particles are smaller than 50 nm (range 6 nm to 1 µm). Additional analysis of these retrieved wear particles indicates that the particles have several different elemental compositions. There are Co-Cr-Mo particles, but there is an even greater number of chromium oxide particles. It has been hypothesized that the Co-Cr-Mo particles are produced by the wear of the carbides on the bearing surfaces and the prosthesis matrix, and that the chromium oxide particles come from the passivation layer on the implant surface and possibly from oxidized chromium carbides. There is little known about the rates of metallic particle production in vivo, lymphatic transport of metallic particles from the joint, and systemic dissemination. Utilizing information on volumetric wear rate and average particle size, it has been estimated that 6.7 x 1012 to 2.5 x 1014 metal particles are produced per year, which is 13 to 500 times the number of PE particles produced per year by a typical MOP joint. The aggregate surface area of these metal wear particles is substantial. (Malviya 2010)

It is important to recognize that in modern THR prostheses there may be several sources of metal particle and ion generation.(Savarino 2006) Systemic dissemination of soluble and particulate corrosion products from modular junctions has been described, including the presence of metallic particles in the lymph nodes, liver, and spleen. In modular total hips with a MOP bearing, with no or mild corrosion at the modular head and neck junction, serum and urine levels of Co averaged 0.94 ppb (range <0.54 to 1.65) and 0.92 ppb (range <0.3 to 1.14), respectively, and urine Cr levels averaged 1.0 ppb (range 0.54 to 1.92). With moderate or severe corrosion at the modular head and neck junction, the serum and urine levels of Co averaged 1.06 ppb (range 0.8 to 1.4) and 0.87 ppb (range <0.3 to 1.3), respectively, and the average urine Cr levels were 1.59 ppb (range 0.6 to 3.0).(Jacobs 1998) The levels of metal ions in serum and urine are further elevated in patients with MOM bearings. It appears that the ion levels are higher in the short term and decrease over time. This is consistent with a conditioning phase or running-in of the bearing. The ion levels in subjects with Co-Cr alloy MOM THRs that had been in situ for an average of 24 years have been reported. The average serum Co level was 0.9 ppb (range <0.3 to 2.0). The average serum Cr level was 1.28 ppb (range 0.21 to 2.56), and the average urine Cr level was 1.22 ppb (range 0.26 to 2.59).(Bitsch 2007) The longer-term serum and urine ion levels produced by MOM bearings are not much higher than those produced by the modular junctions of femoral components. Unfortunately, the toxicological importance of these trace metal elevations has not been established, and available data do not answer questions regarding the risks of ion hypersensitivity, toxicity, and carcinogenesis. Since wear of a MOM bearing cannot generally be measured on a radiograph, serum and urine metal ion concentrations may be useful indicators of patient activity and the tribological performance of these bearings.(MacDonald 2004) The distribution and histological effects of wear particles from a McKee-Farrar THR with an in vivo use of nearly 30 years has been described. Clinically, the implant was functioning well. Serum and urine levels of chromium and serum levels of cobalt, obtained 25 years after implantation and while the patient was healthy and active, were found to be 1.02 ppb, 0.51 ppb, and 0.66 ppb, respectively. Sections from multiple samples of the lymph nodes, spleen, liver, and kidney were examined with light microscopy. This tissue analysis did not reveal any evidence of end-organ damage or accumulation of metal particles. It appears that a patient with normal renal function is capable of clearing cobalt and chromium ions from his or her system. Potential additional sources of ions include the acetabular morse taper connection.(Brodner 2003)

### **3.3.4.2 Biological response**

188 Recent Advances in Arthroplasty

assess systemic ion levels.(Daniel 2007; Antoniou 2008) In a study looking specifically at the serum cobalt concentrations in patients during the first 5 years after an MOM THA, Brodner and colleagues showed a moderate concentration of 1 µg/L at 1 year and 0.7 µg/L at 5 years. They did not find any significant difference in the levels from 3 to 12 months and the subsequent measurements and, therefore, concluded that the serum cobalt concentrations did not reflect the higher run-in wear of MM implants.(Brodner 2003) In modern MM surface arthroplasty, Skipor and colleagues reported elevations in serum chromium that were 22-fold, 23-fold, and 21-fold higher at 3, 6, and 12 months postoperatively, respectively, than preoperatively. The corresponding serum cobalt levels were eightfold, sevenfold, and sixfold higher than preoperative levels, and the authors reported that the values seen with the current generation of surface arthroplasties were in the same range as those observed in association with conventional MOM THA.(Skipor 2002) Comparison of serum cobalt and chromium ion levels in patients with a MOM compared with a MOPE bearing and a MOM compared with a ceramic-ceramic bearing showed that there were always higher ion levels in the patients with the MOM bearings. In a study analyzing the whole blood cobalt and chromium ion levels in 68 patients, Hart and colleagues reported a decrease in cytotoxic CD8+ T-lymphocytes, which are involved in the defense against intracellular pathogens and cancerous cells.(Hart 2009) Of interest, Back and colleagues noted a significant decrease in the cobalt levels at 2 years compared with the levels at 6 months, as well as a decrease in the chromium levels at 2 years compared with the levels at 9 months. When looking at the effects of the size of the femoral head and the orientation of the acetabular component on the overall concentrations of metal ions in the whole blood of patients with MOM resurfacings, Langton et al. reported higher ion concentrations in patients with smaller components and when the inclination angle was greater than 45 degrees and the anteversion angle

Wear particles from MOM bearings are nanometers in linear dimension, which is substantially smaller than PE wear particles. Light microscopic analysis of tissue obtained from around MOM joints showed particles with variable and irregular shape. The size of metal particles reported by scanning electron microscopy studies ranges from 0.1 to 5 µm. The studies have suggested that large metallic particles observed with light microscopy were agglomerates of the smaller particles. Transmission electron microscopy has demonstrated wear particles from Co-Cr-Mo bearings to be round to oval in shape with irregular boundaries. Most of the particles are smaller than 50 nm (range 6 nm to 1 µm). Additional analysis of these retrieved wear particles indicates that the particles have several different elemental compositions. There are Co-Cr-Mo particles, but there is an even greater number of chromium oxide particles. It has been hypothesized that the Co-Cr-Mo particles are produced by the wear of the carbides on the bearing surfaces and the prosthesis matrix, and that the chromium oxide particles come from the passivation layer on the implant surface and possibly from oxidized chromium carbides. There is little known about the rates of metallic particle production in vivo, lymphatic transport of metallic particles from the joint, and systemic dissemination. Utilizing information on volumetric wear rate and average particle size, it has been estimated that 6.7 x 1012 to 2.5 x 1014 metal particles are produced per year, which is 13 to 500 times the number of PE particles produced per year by a typical MOP joint. The aggregate surface area of these metal wear particles is substantial.

exceeded 20 degrees.(Langton 2008)

(Malviya 2010)

Although the periprosthetic tissue reactions surrounding MOM THA demonstrate some histological characteristics similar to those typically seen around MOPE THAs, they have been reported to be less intense with fewer histiocytes present in the tissues. It is still uncertain whether this difference can be attributed to variations in particle concentration, size, shape, or composition. The wear particles produced by MOM implants in vitro and in vivo are in the nanometer size range, mostly round to oval but with the presence of some needle-shaped particles as well, mainly depending on the number of cycles (for in vitro particles) and the implantation time (for in vivo particles). In addition, most of the particles contain Chromium (Cr) and Oxygen (O) but no Cobalt (Co) and are therefore most likely chromium oxides. Therefore, in addition to focusing on CoCrMo particles, attention to the effects of chromium oxide particles should be addressed in future in vitro and in vivo biologic studies. Finally, because of their nanometer size, those particles have a high surface

The Bearing Surfaces in Total Hip Arthroplasty – Options, Material Characteristics and Selection 191

complex (MHC) class II restricted CD4+ T-lymphocytes by professional antigen-presenting cells (APCS). However, in addition to conventional peptide antigens, T cells can also react directly with metal ions. Regardless of the mechanisms, the consequences are that metal ions-sensitized CD4+ T-lymphocytes would elaborate cytokines, including interferon γ, that will attract and activate more macrophages, leading to the release of additional osteolytic cytokines and possibly causing a T-cell-mediated periprosthetic osteolysis.(Hallab 2001; Whittingham 2008) In a review study, Gawkrodger concluded that although early MOM displayed high incidences of hypersensitivity, no conclusion could be drawn about the new generation of MOM implants.(Gawkrodger 2003) Finally, it is still unclear if the loosening of the implants can be linked to an increased reactivity of the lymphocytes to metal particles or ions. When comparing tissues surrounding MOPE and MOM THAs, Campbell and colleagues showed more lymphocytes in the MOM THAs.(Campbell 2005) Although this infiltration of lymphocytes could be indicative of a delayed-type hypersensitivity response to the metal wear products and the development of a typical immunological response, Willert and colleagues also reported the presence of plasma cells, B-lymphocytes, and massive fibrin exudation, which are not characteristic of a type IV delayed-type hypersensitivity reaction. These authors described this reaction as an aseptic lymphocytedominated vasculitis-associated lesion (ALVAL) or as a lymphocyte-dominated immunologic (LYDIA) answer. Pandit et al. reported the presence of pseudotumors surrounding MOM surface replacements. These pseudotumors were characterized by an extensive necrosis of dense connective tissue, a focaly heavy macrophage and lymphocytic infiltration as well as the presence of plasma cells and eosinophils in some cases. The observed reaction was somewhat similar to ALVAL reaction reported by Willert et al. but with possibly a more diffuse lymphocyte infiltrate and the presence of extensive connective tissue necrosis.(Willert 2005) However, the specificity of such pseudotumors to MOM implants remains to be determined.(Pandit 2008) The potential adverse effects of MOM wear particles are mainly due to their local and systemic distribution and their capacity for releasing corrosion products, especially if they are in situ for several decades. Systemic dissemination of soluble and particulate corrosion products has been described, including the presence of metallic particles in the liver and spleen and raises questions about potential

Metal wear debris may cause DNA damage through the action of metal particles and/or through the action of metal ions by a chemical effect. Cobalt-chromium particles may exert a purely particulate effect that damages DNA by inducing free-radical formation in a way similar to the actions of nonmetallic particles. Alternatively, the cobalt-chromium particle might produce a chemical effect as a result of the release of cobalt and/or chromium ions. Cr3+ and Cr6+ cause DNA-strand breaks. Co2+ itself causes DNA-strand breaks, but in addition it inhibits DNA repair, therefore potentially augmenting the damage caused by chromium. Metal ions and wear debris may lead to genomic instability, similar to that induced by ionizing radiation, and this may increase the probability of DNA damage or

Temporary increases in the incidence of hematopoietic cancers at different follow-up points were observed in some cohorts. A study by Visuri et al. demonstrated a 3.77-fold increase in the rate of hematopoietic cancer following metal-on-metal total hip arthroplasty compared with the rate following metal-on-polyethylene total hip

genotoxicity.(Urban 2000)

**3.3.4.4 Local and systemic effects** 

mutation during normal cellular function.(Cobb 2006)

area and are subject to corrosion, releasing metal ions into the surrounding areas, and therefore the biologic response to metal ions also need to be considered.

The large aggregate surface area of metal wear particles may have both local and systemic effects. Surface area has been identified as a variable affecting the macrophage response to panicles. However, the local tissue reaction around MOM prosthesis, indicated by the number of histiocytes, is about one grade lower than that around MOP prostheses. A number of hypotheses have been proposed to explain this discrepancy. Since metal particles are considerably smaller than PE particles, histiocytes are able to store a larger number of metal particles, and therefore, the total number of histiocytes required to store the metal particles is lower. Very small particles may enter macrophages by pinocytosis instead of phagocytosis, which may alter the cellular response to the particles. It is also recognized that Co-Cr particles have greater potential for cytotoxicity than PE particles, and the cell may be incapable of the same inflammatory response. There may be a difference in the relative proportion of metal wear particles that are retained locally versus systemically distributed compared to PE wear particles. Dissolution of metal particles results in elevation of the cobalt and chromium ion concentrations in erythrocytes, serum, and urine. The metal particles have shown a dose-response effect. Low to moderate concentrations of metal particles stimulate the release of cytokines, such as interleukin-1 (IL-1), interleukin-6 (IL-6), tumor necrosis factor-α, and prostaglandin E1, that can lead to periprosthetic osteolysis and aseptic loosening. At higher concentrations, however, Co-Cr particles have been found to be cytotoxic, altering the phagocytic activity of macrophages and leading to cell death.(Kwon 2009; Huber 2010)

Although the volume of reactive periprosthetic inflammatory tissue associated with MOM bearings is less than with MOP, osteolysis can occur in hips with MOM bearings. The incidence of osteolysis associated with MOM bearings has not been well established but appears to be comparatively low.(Bharma 2006)

#### **3.3.4.3 Inflammatory versus immunological responses**

Because metal ions remain a major cause for concern in the widespread clinical use of MOM implants, in vitro studies have also been conducted on the effects of these metal ions. Co2+ and Cr3+ have been demonstrated to induce tumor necrosis factor (TNF)-α secretion. Petit and colleagues examined the effect of Co2+ and Cr3+ ions on protein oxidation in human U937 macrophages and showed that both Co2+ and Cr3+ ions induced a time- and dosedependent protein oxidation, reaching 6.5 and 2.9 times the control levels after 72 hours, respectively.(Bagchi 2002) A histological study of tissues surrounding MOPE and MOM THAs by Campbell and colleagues showed that fewer macrophages were present in the tissues surrounding MOM THAs than in the tissues surrounding MOPE THAs. Catelas and colleagues demonstrated that tissues surrounding failed MOM THAs with low to moderate quantities of metal particles could induce the production of potentially osteolytic cytokines. In addition to the potential for wear products to induce osteolysis through inflammatory response as observed surrounding conventional MOPE implants, a concern has emerged regarding the possibility of specific immunological responses. These immunological responses could be due to the release of metal ions that act as antigens and stimulate an allergic (hypersensitivity) reaction when they form organometallic complexes with proteins. Hypersensitivity reactions can be mediated by antibodies (type I, II, or III) or can be cellmediated (type IV, i.e., delayed type hypersensitivity). Type IV reactions to metal ions involve the activation and clonal expansion of metal-ion-specific major histocompatibility complex (MHC) class II restricted CD4+ T-lymphocytes by professional antigen-presenting cells (APCS). However, in addition to conventional peptide antigens, T cells can also react directly with metal ions. Regardless of the mechanisms, the consequences are that metal ions-sensitized CD4+ T-lymphocytes would elaborate cytokines, including interferon γ, that will attract and activate more macrophages, leading to the release of additional osteolytic cytokines and possibly causing a T-cell-mediated periprosthetic osteolysis.(Hallab 2001; Whittingham 2008) In a review study, Gawkrodger concluded that although early MOM displayed high incidences of hypersensitivity, no conclusion could be drawn about the new generation of MOM implants.(Gawkrodger 2003) Finally, it is still unclear if the loosening of the implants can be linked to an increased reactivity of the lymphocytes to metal particles or ions. When comparing tissues surrounding MOPE and MOM THAs, Campbell and colleagues showed more lymphocytes in the MOM THAs.(Campbell 2005) Although this infiltration of lymphocytes could be indicative of a delayed-type hypersensitivity response to the metal wear products and the development of a typical immunological response, Willert and colleagues also reported the presence of plasma cells, B-lymphocytes, and massive fibrin exudation, which are not characteristic of a type IV delayed-type hypersensitivity reaction. These authors described this reaction as an aseptic lymphocytedominated vasculitis-associated lesion (ALVAL) or as a lymphocyte-dominated immunologic (LYDIA) answer. Pandit et al. reported the presence of pseudotumors surrounding MOM surface replacements. These pseudotumors were characterized by an extensive necrosis of dense connective tissue, a focaly heavy macrophage and lymphocytic infiltration as well as the presence of plasma cells and eosinophils in some cases. The observed reaction was somewhat similar to ALVAL reaction reported by Willert et al. but with possibly a more diffuse lymphocyte infiltrate and the presence of extensive connective tissue necrosis.(Willert 2005) However, the specificity of such pseudotumors to MOM implants remains to be determined.(Pandit 2008) The potential adverse effects of MOM wear particles are mainly due to their local and systemic distribution and their capacity for releasing corrosion products, especially if they are in situ for several decades. Systemic dissemination of soluble and particulate corrosion products has been described, including the presence of metallic particles in the liver and spleen and raises questions about potential genotoxicity.(Urban 2000)

### **3.3.4.4 Local and systemic effects**

190 Recent Advances in Arthroplasty

area and are subject to corrosion, releasing metal ions into the surrounding areas, and

The large aggregate surface area of metal wear particles may have both local and systemic effects. Surface area has been identified as a variable affecting the macrophage response to panicles. However, the local tissue reaction around MOM prosthesis, indicated by the number of histiocytes, is about one grade lower than that around MOP prostheses. A number of hypotheses have been proposed to explain this discrepancy. Since metal particles are considerably smaller than PE particles, histiocytes are able to store a larger number of metal particles, and therefore, the total number of histiocytes required to store the metal particles is lower. Very small particles may enter macrophages by pinocytosis instead of phagocytosis, which may alter the cellular response to the particles. It is also recognized that Co-Cr particles have greater potential for cytotoxicity than PE particles, and the cell may be incapable of the same inflammatory response. There may be a difference in the relative proportion of metal wear particles that are retained locally versus systemically distributed compared to PE wear particles. Dissolution of metal particles results in elevation of the cobalt and chromium ion concentrations in erythrocytes, serum, and urine. The metal particles have shown a dose-response effect. Low to moderate concentrations of metal particles stimulate the release of cytokines, such as interleukin-1 (IL-1), interleukin-6 (IL-6), tumor necrosis factor-α, and prostaglandin E1, that can lead to periprosthetic osteolysis and aseptic loosening. At higher concentrations, however, Co-Cr particles have been found to be cytotoxic, altering the phagocytic activity of macrophages and leading to cell death.(Kwon

Although the volume of reactive periprosthetic inflammatory tissue associated with MOM bearings is less than with MOP, osteolysis can occur in hips with MOM bearings. The incidence of osteolysis associated with MOM bearings has not been well established but

Because metal ions remain a major cause for concern in the widespread clinical use of MOM implants, in vitro studies have also been conducted on the effects of these metal ions. Co2+ and Cr3+ have been demonstrated to induce tumor necrosis factor (TNF)-α secretion. Petit and colleagues examined the effect of Co2+ and Cr3+ ions on protein oxidation in human U937 macrophages and showed that both Co2+ and Cr3+ ions induced a time- and dosedependent protein oxidation, reaching 6.5 and 2.9 times the control levels after 72 hours, respectively.(Bagchi 2002) A histological study of tissues surrounding MOPE and MOM THAs by Campbell and colleagues showed that fewer macrophages were present in the tissues surrounding MOM THAs than in the tissues surrounding MOPE THAs. Catelas and colleagues demonstrated that tissues surrounding failed MOM THAs with low to moderate quantities of metal particles could induce the production of potentially osteolytic cytokines. In addition to the potential for wear products to induce osteolysis through inflammatory response as observed surrounding conventional MOPE implants, a concern has emerged regarding the possibility of specific immunological responses. These immunological responses could be due to the release of metal ions that act as antigens and stimulate an allergic (hypersensitivity) reaction when they form organometallic complexes with proteins. Hypersensitivity reactions can be mediated by antibodies (type I, II, or III) or can be cellmediated (type IV, i.e., delayed type hypersensitivity). Type IV reactions to metal ions involve the activation and clonal expansion of metal-ion-specific major histocompatibility

therefore the biologic response to metal ions also need to be considered.

2009; Huber 2010)

appears to be comparatively low.(Bharma 2006)

**3.3.4.3 Inflammatory versus immunological responses** 

Metal wear debris may cause DNA damage through the action of metal particles and/or through the action of metal ions by a chemical effect. Cobalt-chromium particles may exert a purely particulate effect that damages DNA by inducing free-radical formation in a way similar to the actions of nonmetallic particles. Alternatively, the cobalt-chromium particle might produce a chemical effect as a result of the release of cobalt and/or chromium ions. Cr3+ and Cr6+ cause DNA-strand breaks. Co2+ itself causes DNA-strand breaks, but in addition it inhibits DNA repair, therefore potentially augmenting the damage caused by chromium. Metal ions and wear debris may lead to genomic instability, similar to that induced by ionizing radiation, and this may increase the probability of DNA damage or mutation during normal cellular function.(Cobb 2006)

Temporary increases in the incidence of hematopoietic cancers at different follow-up points were observed in some cohorts. A study by Visuri et al. demonstrated a 3.77-fold increase in the rate of hematopoietic cancer following metal-on-metal total hip arthroplasty compared with the rate following metal-on-polyethylene total hip

The Bearing Surfaces in Total Hip Arthroplasty – Options, Material Characteristics and Selection 193

MOM bearings apply to ceramic-on-ceramic (COC) bearings. However, ceramics have two important properties that make them an outstanding material regarding friction and wear. Ceramics are hydrophilic, permitting a better wettability of the surface. This ensures that the synovial fluid-film is uniformly distributed over the whole bearing surface area. Secondly, ceramic has a greater hardness than metal and can be polished to a much lower surface roughness. Although the better wettability results in a slightly thinner fluid-film than with MOM bearings it is compensated by the reduced size of the asperities on the surface. Overall, this results in a favorable higher λ ratio and in a reduced coefficient of friction. This bearing combination is the most likely to achieve true fluid-film lubrication. However, because of the hardness of ceramics, the wear characteristics are sensitive to design, manufacturing, and implantation variables. Rapid wear has also been observed, generally associated with suboptimal positioning of the implants. COC bearings currently in clinical use are made of alumina. Developments in the production process (sintering) have improved the quality of the material. Modern alumina ceramics have a low porosity, low grain size, high density, and high purity. Thus, its hardness, fracture toughness, and burst strength increased. The alumina on alumina bearing is considered the standard ceramic on ceramic articulation.(D'Antonio 2009) Alumina ceramic bearings have been in clinical use for more than three decades, and significant basic science and clinical research support their use. The small amount of ceramic particulate debris that is generated has been noted to be much less biologically reactive than metal or polyethylene particles. Most important, the incidence of osteolysis associated with use of ceramic on ceramic bearings appears to be

Alumina ceramics are classified as hard, stiff, and brittle materials. Because alumina ceramics are highly oxidized, they are biologically inert and resistant to further oxidation. The hardness of alumina creates a product with significant resistance to surface damage, and ceramics are much harder than other materials routinely used in orthopedic surgery. The hardness of alumina makes it very abrasive and wear resistant. In addition, the hardness of alumina increases its resistance to scratching, and it is much less likely to scratch than titanium or cobalt chromium alloys. In fact, the only material capable of scratching alumina is diamond. Clinically, this is important because alumina can resist third-body wear and is not scratched by retained cement particles or bone. Although alumina has poor bending characteristics, it is extremely strong in compression. This lack of bending strength has currently limited its use in total hip arthroplasty to the femoral head and cup liner. Because alumina is very stiff, it does not deform under high loads. Therefore very precise production techniques are needed in order to ensure proper fit of the head within the socket. Polyethylene will mold around a femoral head if there is an initial, small incongruity, which is not true with ceramics, and poor manufacturing can lead to high wear rates. The lack of ceramic deformation makes the contact areas between the head and socket smaller as compared with metal on polyethylene articulations. In order to maximize the contact surface area, clearance must be optimized. Alumina is more than 300 times stiffer than cancellous bone and almost 200 times stiffer than polymethlymethracrylate. Because of this significant modulus mismatch, cemented ceramic components have been found to be associated with higher cement fracture and loosening rates than all-polyethylene components. Alumina is very brittle and under compression will deform linearly until fracture. No plastic deformation occurs before fracture. By definition, its fracture toughness is considered to be

minimal or nonexistent. (Bierbaum 2002)

**3.4.1 Mechanical properties** 

arthroplasty, but this difference was not significant.(Visuri 2006) Recent evidence from studies of patients who have undergone metal-on-metal hip resurfacing has shown raised circulating levels of cobalt and chromium ions to be associated with T-cell lymphopenia, with the lymphocyte counts of these patients differing from those of control subjects with a total hip arthroplasty that did not produce metal wear debris. The effect was consistent only for CD8+ T-cell lymphocytes.(Hart 2006, 2009)

Another concern regarding increased metal-ion levels following metal-on-metal total hip arthroplasty is that the ions may cross the placental barrier. Initially, it was thought that the placenta was an effective barrier to metal ions, but a subsequent study showed that the placenta actually exerts a modulatory effect on the rate of metal ion transfer. The transfer rate at low maternal metal-ion levels is higher than that at higher maternal levels. The transplacental transfer rate was in excess of 95% in the controls for both metals, but only 29% for chromium and 60% for cobalt in study patients, suggesting that the placenta exerts a modulatory effect on the rate of metal ion transfer.(Ziaee 2007) On the basis of these findings and the lack of comprehensive knowledge regarding the potential effects of metal ions on fetal development, metal-on-metal total hip arthroplasty probably should not be performed in women of child-bearing age until additional information is available.(Brodner 2004)

#### **3.3.4.5 Cancer risk**

Co and Cr wear particles have been shown to induce carcinoma in animal models, giving rise to the concern that such alloys could have the same effect in human tissues if present in sufficient amounts for a sufficient length of time. The first well-documented case of cancer associated with total joint replacement was in a patient who developed a malignant fibrous histiocytoma 3.5 years after a MOM THR performed in December 1969. There have been, at least, 24 additional cases reported in the English literature of malignancy occurring in association with a total hip or knee prostheses. Of the 25 reported cases of cancer following a total joint replacement, 21 involved sarcomas.(Gillespie 1988) The risk of cancer after MOM THR has been assessed specifically in only one epidemiological study. In that study, the relative risk of cancer was reported to be 0.95, suggesting that there is no apparent increased risk of cancer development after MOM total hip arthroplasty. In addition, the risk of sarcoma after MOM THR was found to be 0.00. However, the same study found the relative risk of hematopoietic cancer to be 1.59 following MOM THR and 3.77 for leukemia when MOM implants were compared with MOP implants. From an epidemiological perspective, these data are limited because of the small number of patients who underwent MOM THR. Furthermore, the majority of patients in these reports have less than 10 years of follow-up. The latency of known carcinogens, such as tobacco, asbestos, and ionizing radiation, is several decades. Longer follow-up of large patient groups is needed to better assess the risk of cancer with any implant system.(Tharani 2001) Since the goal of more wear resistant bearings is to reduce the need for reoperation, theoretical risks should be weighed against the risks of revision THR. In the Medicare population, the 90-day mortality following revision THR is 2.6%, which is significantly and substantially higher than that of primary total hip and directly related to the revision procedure. Rigorous long-term studies are needed to assess the relative risk to benefit ratios for total hip bearings.

### **3.4 Ceramic bearings**

Ceramic bearings, made of alumina, have demonstrated the lowest in vivo wear rates to date of any bearing combination. The same principles of friction and lubrication reported for MOM bearings apply to ceramic-on-ceramic (COC) bearings. However, ceramics have two important properties that make them an outstanding material regarding friction and wear. Ceramics are hydrophilic, permitting a better wettability of the surface. This ensures that the synovial fluid-film is uniformly distributed over the whole bearing surface area. Secondly, ceramic has a greater hardness than metal and can be polished to a much lower surface roughness. Although the better wettability results in a slightly thinner fluid-film than with MOM bearings it is compensated by the reduced size of the asperities on the surface. Overall, this results in a favorable higher λ ratio and in a reduced coefficient of friction. This bearing combination is the most likely to achieve true fluid-film lubrication. However, because of the hardness of ceramics, the wear characteristics are sensitive to design, manufacturing, and implantation variables. Rapid wear has also been observed, generally associated with suboptimal positioning of the implants. COC bearings currently in clinical use are made of alumina. Developments in the production process (sintering) have improved the quality of the material. Modern alumina ceramics have a low porosity, low grain size, high density, and high purity. Thus, its hardness, fracture toughness, and burst strength increased. The alumina on alumina bearing is considered the standard ceramic on ceramic articulation.(D'Antonio 2009) Alumina ceramic bearings have been in clinical use for more than three decades, and significant basic science and clinical research support their use. The small amount of ceramic particulate debris that is generated has been noted to be much less biologically reactive than metal or polyethylene particles. Most important, the incidence of osteolysis associated with use of ceramic on ceramic bearings appears to be minimal or nonexistent. (Bierbaum 2002)

### **3.4.1 Mechanical properties**

192 Recent Advances in Arthroplasty

arthroplasty, but this difference was not significant.(Visuri 2006) Recent evidence from studies of patients who have undergone metal-on-metal hip resurfacing has shown raised circulating levels of cobalt and chromium ions to be associated with T-cell lymphopenia, with the lymphocyte counts of these patients differing from those of control subjects with a total hip arthroplasty that did not produce metal wear debris. The effect was consistent

Another concern regarding increased metal-ion levels following metal-on-metal total hip arthroplasty is that the ions may cross the placental barrier. Initially, it was thought that the placenta was an effective barrier to metal ions, but a subsequent study showed that the placenta actually exerts a modulatory effect on the rate of metal ion transfer. The transfer rate at low maternal metal-ion levels is higher than that at higher maternal levels. The transplacental transfer rate was in excess of 95% in the controls for both metals, but only 29% for chromium and 60% for cobalt in study patients, suggesting that the placenta exerts a modulatory effect on the rate of metal ion transfer.(Ziaee 2007) On the basis of these findings and the lack of comprehensive knowledge regarding the potential effects of metal ions on fetal development, metal-on-metal total hip arthroplasty probably should not be performed in women of child-bearing age until additional information is available.(Brodner 2004)

Co and Cr wear particles have been shown to induce carcinoma in animal models, giving rise to the concern that such alloys could have the same effect in human tissues if present in sufficient amounts for a sufficient length of time. The first well-documented case of cancer associated with total joint replacement was in a patient who developed a malignant fibrous histiocytoma 3.5 years after a MOM THR performed in December 1969. There have been, at least, 24 additional cases reported in the English literature of malignancy occurring in association with a total hip or knee prostheses. Of the 25 reported cases of cancer following a total joint replacement, 21 involved sarcomas.(Gillespie 1988) The risk of cancer after MOM THR has been assessed specifically in only one epidemiological study. In that study, the relative risk of cancer was reported to be 0.95, suggesting that there is no apparent increased risk of cancer development after MOM total hip arthroplasty. In addition, the risk of sarcoma after MOM THR was found to be 0.00. However, the same study found the relative risk of hematopoietic cancer to be 1.59 following MOM THR and 3.77 for leukemia when MOM implants were compared with MOP implants. From an epidemiological perspective, these data are limited because of the small number of patients who underwent MOM THR. Furthermore, the majority of patients in these reports have less than 10 years of follow-up. The latency of known carcinogens, such as tobacco, asbestos, and ionizing radiation, is several decades. Longer follow-up of large patient groups is needed to better assess the risk of cancer with any implant system.(Tharani 2001) Since the goal of more wear resistant bearings is to reduce the need for reoperation, theoretical risks should be weighed against the risks of revision THR. In the Medicare population, the 90-day mortality following revision THR is 2.6%, which is significantly and substantially higher than that of primary total hip and directly related to the revision procedure. Rigorous long-term studies are

needed to assess the relative risk to benefit ratios for total hip bearings.

Ceramic bearings, made of alumina, have demonstrated the lowest in vivo wear rates to date of any bearing combination. The same principles of friction and lubrication reported for

only for CD8+ T-cell lymphocytes.(Hart 2006, 2009)

**3.3.4.5 Cancer risk** 

**3.4 Ceramic bearings** 

Alumina ceramics are classified as hard, stiff, and brittle materials. Because alumina ceramics are highly oxidized, they are biologically inert and resistant to further oxidation. The hardness of alumina creates a product with significant resistance to surface damage, and ceramics are much harder than other materials routinely used in orthopedic surgery. The hardness of alumina makes it very abrasive and wear resistant. In addition, the hardness of alumina increases its resistance to scratching, and it is much less likely to scratch than titanium or cobalt chromium alloys. In fact, the only material capable of scratching alumina is diamond. Clinically, this is important because alumina can resist third-body wear and is not scratched by retained cement particles or bone. Although alumina has poor bending characteristics, it is extremely strong in compression. This lack of bending strength has currently limited its use in total hip arthroplasty to the femoral head and cup liner. Because alumina is very stiff, it does not deform under high loads. Therefore very precise production techniques are needed in order to ensure proper fit of the head within the socket. Polyethylene will mold around a femoral head if there is an initial, small incongruity, which is not true with ceramics, and poor manufacturing can lead to high wear rates. The lack of ceramic deformation makes the contact areas between the head and socket smaller as compared with metal on polyethylene articulations. In order to maximize the contact surface area, clearance must be optimized. Alumina is more than 300 times stiffer than cancellous bone and almost 200 times stiffer than polymethlymethracrylate. Because of this significant modulus mismatch, cemented ceramic components have been found to be associated with higher cement fracture and loosening rates than all-polyethylene components. Alumina is very brittle and under compression will deform linearly until fracture. No plastic deformation occurs before fracture. By definition, its fracture toughness is considered to be

The Bearing Surfaces in Total Hip Arthroplasty – Options, Material Characteristics and Selection 195

articulation are due to its low surface roughness (secondary to small grain size), hardness, enhanced wettability, and fluid film lubrication. It has been shown that there are two wear phases during in vitro testing. The "run-in" phase is the first phase and involves the first million or so cycles. Volumetric wear rates for alumina against alumina bearings during this run-in phase measure 0.1 to 0.2 mm3 per million cycles. The second phase is called the "steady-state" phase. During this period, volumetric wear rates decrease to less than 0.02 mm3 per million cycles. Compared with metal on polyethylene couples, during both the run-in and steady-state phases, wear is reduced up to 5000 fold. Under certain clinical conditions, accelerated wear can occur with alumina on alumina couples. One phenomenon called "stripe wear" occurs when accelerated wear is present over a discrete area. Stripe wear may be associated with separation of the ball from the socket such as during the swing phase of gait or when the ball is levered out of the socket by impingement. In vitro testing under the conditions of separation of the femoral head from the socket leads to increased volumetric wear. It was noted that wear as high as 1.24 mm3 per million cycles could occur with separation and stripe wear results. A bimodal distribution of particle size was also noted in this study, with nanometer-sized particles (1 to 35 nm) probably associated with polishing of the articulation and micrometer-sized particles (0.05 to 10 µm) that likely originated from stripe wear and transgranular fracture of the alumina ceramic. Numerous retrieval studies of ceramic bearings have been performed, and the results are interesting. One study examined retrieved alumina components associated with aseptic loosening of the socket at a mean of 11 years after implementation. Components were classified into three groups: (1) low wear with no visible signs of material loss; (2) stripe wear with a visible oblong worn area on the femoral head and a penetration rate below 10 µm/year; and (3) severe wear with visible loss of material on both components and maximum penetration higher than 150 µm.(Fig.10 and 11) Evaluation of these 11 components revealed massive, severe wear on two devices. The remaining nine components had liner wear rates less than 15 µm/year. The authors concluded that two different types of wear are associated with ceramic on ceramic couples-one that is limited and has negligible effect on long-term performance of the implant, and a second type that catastrophically leads to rapid destruction of the bearing surface. Published wear rates examining clinical performance of the alumina on alumina bearing surface have reported wear to range from 0.3 µm/year to 5.0 mm/year. These variations may be related to implanted material and design issues or surgical and patient factors. However, it is important to note that most catastrophic wear

Fig. 10. Acetabular liner showing gross damage to the liner

its resistance to fracture. The initial flaws in the material determine the risk of ceramic fracture, and flaws are related to the purity and density of the ceramic. A combination of improvements, including improved processing with smaller grain sizes, fewer impurities, laser etching, and proof testing, have led to a lower incidence of material fracture. The burst strength of alumina components improved from 38 kilonewtons in 1977 to 98 kilonewtons in 1998. (Tateiwa 2008)

### **3.4.2 Alumina and zirconia**

Medical-grade alumina is consolidated from a high-purity (>99%) alumina powder through a hot isostatic pressing step. Modifications to the sintering process over the past three decades have resulted in increased density, hardness, and strength of medical-grade alumina. Modifications to the sintering also led to a decrease in the average grain size from 4.5 µm to about 1.8 µm, with a narrower grain size distribution. The improved regulation of the grain size did not lead to any noticeable changes in the fracture toughness of alumina. While alumina remains the most widely used ceramic in the total hip, zirconia is also utilized as a replacement for alumina femoral heads, but only for applications where the counterface is UHMWPE. Zirconia is stronger than alumina; therefore it can withstand higher stresses, which may be induced when smaller and thinner implants are used. During deformation, pure zirconia undergoes a crystalline-phase transformation from monoclinic to tetragonal, with the former occupying more space. This transformation adversely affects the mechanical properties of pure zirconia, leading to internal stresses within the deformation zone ahead of propagating cracks. Zirconia can be stabilized with the addition of either magnesium oxide (MgO) or yttrium oxide (Y2O3) to avoid the detrimental effects of deformation-induced phase transformation.(De Aza 2002)

Alumina was the first ceramic to be widely used in hip arthroplasty, following its introduction in 1970. Aluminium oxide (Al203) can be highly polished to produce a very low coefficient of friction and is also highly resistant to abrasion. It has a more stable structure than the more recently developed Zirconia so the properties are more predictable. Improved manufacturing techniques have resulted in smaller grain size and smoother finish helping to reduce the fracture risk that was encountered when using earlier generations. (D'Antonio 2009)

Zirconium oxide (ZrO2)(Zirconia) was introduced in 1985 as an alternative to alumina. Its superior mechanical strength was attractive with a theoretical reduction in fracture risk. It also allowed smaller femoral head sizes whilst maintaining excellent wear characteristics when coupled with UHMWPE. Clinical trials have shown that UHMWPE wear when coupled with a zirconia femoral head is at least as good as that with a Co-Cr femoral head and in many cases is much better.(Affatato 2001) However Zirconia undergoes phase transformation. The strongest tetragonal phase is also the most unstable, so, for medical purposes, a stabilizer, Yttrium oxide, is used. In vivo studies have shown that phase transformation still occurs at the bearing surface secondary to temperature and pressure changes which transforms zirconia to the more stable monoclinic phase. This results in a 3% increase in volume producing surface roughness. Hence it performs poorly when articulating against itself and can also result in accelerated polyethylene wear bearing on UHMWPE. As a result it has fallen out of favour.(Fernandez-Fairen 2007)

### **3.4.3 Tribologic properties**

In vitro wear studies have proved that alumina on alumina is a very low friction couple, and wear is significantly reduced. The outstanding tribologic properties of the alumina

its resistance to fracture. The initial flaws in the material determine the risk of ceramic fracture, and flaws are related to the purity and density of the ceramic. A combination of improvements, including improved processing with smaller grain sizes, fewer impurities, laser etching, and proof testing, have led to a lower incidence of material fracture. The burst strength of alumina components improved from 38 kilonewtons in 1977 to 98 kilonewtons in

Medical-grade alumina is consolidated from a high-purity (>99%) alumina powder through a hot isostatic pressing step. Modifications to the sintering process over the past three decades have resulted in increased density, hardness, and strength of medical-grade alumina. Modifications to the sintering also led to a decrease in the average grain size from 4.5 µm to about 1.8 µm, with a narrower grain size distribution. The improved regulation of the grain size did not lead to any noticeable changes in the fracture toughness of alumina. While alumina remains the most widely used ceramic in the total hip, zirconia is also utilized as a replacement for alumina femoral heads, but only for applications where the counterface is UHMWPE. Zirconia is stronger than alumina; therefore it can withstand higher stresses, which may be induced when smaller and thinner implants are used. During deformation, pure zirconia undergoes a crystalline-phase transformation from monoclinic to tetragonal, with the former occupying more space. This transformation adversely affects the mechanical properties of pure zirconia, leading to internal stresses within the deformation zone ahead of propagating cracks. Zirconia can be stabilized with the addition of either magnesium oxide (MgO) or yttrium oxide (Y2O3) to avoid the detrimental effects of

Alumina was the first ceramic to be widely used in hip arthroplasty, following its introduction in 1970. Aluminium oxide (Al203) can be highly polished to produce a very low coefficient of friction and is also highly resistant to abrasion. It has a more stable structure than the more recently developed Zirconia so the properties are more predictable. Improved manufacturing techniques have resulted in smaller grain size and smoother finish helping to reduce the

Zirconium oxide (ZrO2)(Zirconia) was introduced in 1985 as an alternative to alumina. Its superior mechanical strength was attractive with a theoretical reduction in fracture risk. It also allowed smaller femoral head sizes whilst maintaining excellent wear characteristics when coupled with UHMWPE. Clinical trials have shown that UHMWPE wear when coupled with a zirconia femoral head is at least as good as that with a Co-Cr femoral head and in many cases is much better.(Affatato 2001) However Zirconia undergoes phase transformation. The strongest tetragonal phase is also the most unstable, so, for medical purposes, a stabilizer, Yttrium oxide, is used. In vivo studies have shown that phase transformation still occurs at the bearing surface secondary to temperature and pressure changes which transforms zirconia to the more stable monoclinic phase. This results in a 3% increase in volume producing surface roughness. Hence it performs poorly when articulating against itself and can also result in accelerated polyethylene wear bearing on

In vitro wear studies have proved that alumina on alumina is a very low friction couple, and wear is significantly reduced. The outstanding tribologic properties of the alumina

fracture risk that was encountered when using earlier generations. (D'Antonio 2009)

UHMWPE. As a result it has fallen out of favour.(Fernandez-Fairen 2007)

1998. (Tateiwa 2008)

**3.4.2 Alumina and zirconia** 

**3.4.3 Tribologic properties** 

deformation-induced phase transformation.(De Aza 2002)

articulation are due to its low surface roughness (secondary to small grain size), hardness, enhanced wettability, and fluid film lubrication. It has been shown that there are two wear phases during in vitro testing. The "run-in" phase is the first phase and involves the first million or so cycles. Volumetric wear rates for alumina against alumina bearings during this run-in phase measure 0.1 to 0.2 mm3 per million cycles. The second phase is called the "steady-state" phase. During this period, volumetric wear rates decrease to less than 0.02 mm3 per million cycles. Compared with metal on polyethylene couples, during both the run-in and steady-state phases, wear is reduced up to 5000 fold. Under certain clinical conditions, accelerated wear can occur with alumina on alumina couples. One phenomenon called "stripe wear" occurs when accelerated wear is present over a discrete area. Stripe wear may be associated with separation of the ball from the socket such as during the swing phase of gait or when the ball is levered out of the socket by impingement. In vitro testing under the conditions of separation of the femoral head from the socket leads to increased volumetric wear. It was noted that wear as high as 1.24 mm3 per million cycles could occur with separation and stripe wear results. A bimodal distribution of particle size was also noted in this study, with nanometer-sized particles (1 to 35 nm) probably associated with polishing of the articulation and micrometer-sized particles (0.05 to 10 µm) that likely originated from stripe wear and transgranular fracture of the alumina ceramic. Numerous retrieval studies of ceramic bearings have been performed, and the results are interesting. One study examined retrieved alumina components associated with aseptic loosening of the socket at a mean of 11 years after implementation. Components were classified into three groups: (1) low wear with no visible signs of material loss; (2) stripe wear with a visible oblong worn area on the femoral head and a penetration rate below 10 µm/year; and (3) severe wear with visible loss of material on both components and maximum penetration higher than 150 µm.(Fig.10 and 11) Evaluation of these 11 components revealed massive, severe wear on two devices. The remaining nine components had liner wear rates less than 15 µm/year. The authors concluded that two different types of wear are associated with ceramic on ceramic couples-one that is limited and has negligible effect on long-term performance of the implant, and a second type that catastrophically leads to rapid destruction of the bearing surface. Published wear rates examining clinical performance of the alumina on alumina bearing surface have reported wear to range from 0.3 µm/year to 5.0 mm/year. These variations may be related to implanted material and design issues or surgical and patient factors. However, it is important to note that most catastrophic wear

Fig. 10. Acetabular liner showing gross damage to the liner

The Bearing Surfaces in Total Hip Arthroplasty – Options, Material Characteristics and Selection 197

periprosthetic membranes obtained during revision for aseptic component loosening with an alumina on alumina couple. These were compared with a series of membranes obtained from revisions of a metal on polyethylene bearing. In the alumina on alumina group, the cellular reaction, which was generally mild, was determined to be in response to the zirconia ceramic particles used in the cement as an opacifying agent. No cellular reaction to the alumina particles was noted.(Lerouge 1997) This contrasted with the significant cellular activity noted in the metal on polyethylene group with reaction to the polyethylene debris. Osteolysis associated with alumina on alumina total hip arthroplasty has been infrequently reported. In one study, when an implant made with large-grain-size ceramics, low density, and high porosity was used, large production of debris resulted and osteolysis occurred. Tissue obtained from failed hips with an alumina on alumina couple was shown to have significantly lower prostaglandin E2 (PGE2) levels compared with tissue obtained from hips with metal on polyethylene articulation. Both alumina and polyethylene debris stimulate cellular release of tumor necrosis factor (TNF)-α. However, polyethylene particles cause more release of TNF-α, and in fact the stimulation may be 8 to 10 times greater. Of importance, alumina particles induce macrophage apoptosis, which leads to decreased macrophage activity. This induced apoptosis explains the decreased levels of TNF-α associated with alumina and may also account for the paucity of ceramic-related osteolysis. Ceramic debris may not be bio-inert as initially assumed, because osteolysis has been described in some patients with a COC bearing.(Nam 2007) Some studies describe inflammatory and cytotoxic reactions on the cellular level, but the relationship to material, size, and particle number remains uncertain. It seems that there is less inflammatory reaction compared to MOM or MOP bearings in well-functioning prostheses. In contrast with polyethylene or metallic particles, foreign body reactions are routinely observed. Ion

In a simulator study, Firkins et al demonstrated superior wear for alumina ceramic-on-metal articulations. Femoral heads 28 mm in diameter made of medical-grade alumina were articulated against acetabular cups manufactured from medical-grade, high-carbon-wrought cobalt-chromium alloy. The implants were tested for 5 million cycles. Wear and surface analyses were performed every 1 million cycles. The results were compared with metal-onmetal articulations that used medical-grade, low-carbon, cobalt-chromium alloy femoral heads against medical-grade, high-carbon, cobalt-chromium alloy sockets. The metal-onmetal bearings initially showed high bedding-in and subsequently developed a steady-state volumetric wear rate of 1.23 mm3 per 1 million cycles. No bedding-in period was observed with the ceramic-on-metal, and a volumetric wear rate of 0.01 mm3 per 1 million cycles was reported. This represents more than a 100-fold decrease in wear with the ceramic-on-metal articulation. Surface analysis of the ceramic head showed no signs of wear or change in surface roughness. Metal particles from both articulations were of nanometer size (6 to 30 nm). In addition, the ceramic-on-metal articulations produced slightly smaller particles although they were far fewer in number. (Firkins 2001) Clinical studies are currently under way to assess the performance of these bearings. The potential advantage of this novel ceramic-on-metal bearing is lower wear and the generation of significantly fewer metal particles compared with currently available metal-on-metal bearing surfaces. This bearing combination allows for the use of large femoral heads, similar to metal-on-metal bearings. With larger ceramic heads, the fracture risk is reduced, and an increased number of femoral

toxicity is not an issue with ceramics.(Mehmood 2008)

**3.4.6 Ceramic-on-metal articulation** 

has been reported with products produced before 1990. In recent years, wear rates below 15 µm/year have been consistently reported. Many investigators believe that severe wear is related to clinically exceptional circumstances and that with properly implanted bearing surfaces, catastrophic wear is essentially nonexistent.(Capello 2008)
