Damage of Implant Surfaces in Total Knee Arthroplasty

Ebru Oral, PhD

Markus A. Wimmer, PhD

Orhun Muratoglu, PhD

INTRODUCTION

Contemporary total knee arthroplasty (TKA) uses metallic, ceramicized, and polymeric materials to reconstruct the knee joint. The femoral component is, in general, manufactured from alloys of cobalt-chrome. The femoral component articulates against a tibial insert made of ultrahigh-molecular-weight polyethylene (UHMWPE)— in this chapter, UHMWPE and polyethylene terminology is used interchangeably and refers to the same polymer (i.e., UHMWPE). The tibial insert is typically metal-backed by titanium base alloy. If used, the patellar components are also fabricated from UHMWPE. These polyethylene components form the primary articulations present in the total knee (i.e., tibiofemoral and patellofemoral), and it is at these articulations where most damage is initiated. When the damage is extensive, the performance of the joint deteriorates and eventually leads to revision surgery.

The primary damage mode active at tibiofemoral and patellofemoral articulations is the wear of the polyethylene component. Articular wear is a highly complex process encompassing several different wear mechanisms. The wear mechanisms describe the mechanical, physical, and chemical interaction of the articulating elements in the joint. The two most active wear mechanisms in total knee replacement (TKR) are delamination and adhesive wear. Delamination occurs by the generation of subsurface cracks and their propagation to the surface, leading to the removal of large (more than 0.5 mm) pieces of polyethylene wear debris in the form of flakes (Figs. 13-1, 13-2 and 13-3). This mechanism has been typically facilitated by the embrittlement (e.g., through oxidation) of polyethylene. Adhesive wear is initiated by the orientation and strain hardening of the polymer and the formation of microjunctions between the articulating bodies. During mechanical action, these microjunctions are torn off and fragments become small particles, usually on the order of a few micrometers or less in size. Abrasion is another wear mechanism that is typically observed in knee joints. Here, hard asperities on the femoral component and hard third body particles, such as bone chips or bone cement particles, generate wear debris by cutting and removing the softer polyethylene articular surface.

Besides wear, creep and plastic deformation also contribute to the damage seen on the polyethylene surface. Creep, which is the flow of the polymeric material under continued stress in loaded regions, leads to permanent deformation. This deformation is a function of time, the majority of which accumulates during approximately the first 2 years of in vivo use, at which time it reaches a steady state.

Among the earlier-described damage modes, delamination was recognized as the primary precursor of most device failures. Once initiated, delamination rapidly led to the loss of geometric conformity at articulation, disrupted the intended load distribution in the joint, and eventually led to implant failure in the form of tibial insert disengagement and fracture of polyethylene component. Thanks to improvements in UHMWPE processing techniques, sterilization, and packaging procedures in the last two decades, oxidation and oxidative embrittlement have been minimized and this wear mechanism is much less prominent. While late failure due to osteolysis and adhesive-abrasive wear was the most prevalent reason of failure in the late 1990s, this is no longer the case (Table 13-1).¹,² According to the 2018 report of the American Joint Replacement Registry,³ the two major reasons for TKR failure are now mechanical complications (23.7%) and aseptic loosening (21%). Wear of the articulating surface is listed in fifth place (2.2%).

Bone resorption secondary to particulate debris is at least still partially responsible for aseptic loosening, especially in the later stages.⁴,⁵ Bone resorption or periprosthetic osteolysis is initiated by osteoclasts, which are stimulated by inflammatory mediators that are produced through the macrophage response to wear debris. Particularly submicron size wear debris (less than 1 µm) is thought to be easily phagocytosed by the macrophages. The progressive growth of inflammatory/granulomatous tissue and the increased bone resorption at the bone-implant interface eventually lead to loosening of the implant. Under certain circumstances, the amount of bone loss after TKR has been great, leading to component loosening or fracture.⁶ Unfortunately, revision surgery in these cases is made particularly difficult, often necessitating the use of special augments to the components and bone grafts.⁷

FIGURE 13-1 An example of delamination on the articular surface of a surgically retrieved polyethylene tibial insert. The delaminations (arrows) are well-accepted to have a strong correlation with oxidation-induced embrittlement secondary to gamma sterilization.

FIGURE 13-2 An example of delamination on the articular surface of a surgically retrieved polyethylene tibial insert. This explant also exhibited a yellow discoloration, due most likely to the penetration of synovial fluid into the cracks. Note that this implant shows pitting (arrows) on the new surface formed by delamination.

FIGURE 13-3 An example of delamination on the articular surface of a surgically retrieved polyethylene patellar component. Delamination, mostly on the lateral facet of the dome, is evident. There are also radial cracks apparent around the periphery of the explant. The inset is a high-magnification photograph of the highlighted superior-lateral region showing numerous cracks.

Delamination, adhesive-abrasive wear, creep, and plastic deformation contribute to the formation of the scar on articular surfaces of the polyethylene components. Unlike total hip replacement, in which the in vivo wear rate of polyethylene is well-quantified through various methods, the true rate at which the TKR damage modes proceed in vivo has only been recently quantified. It is impossible to quantify in vivo wear radiographically, mainly because of the complex geometry of the knee implants. The information on polyethylene damage in TKR available as of 2019 is largely based on the analysis of retrievals, which are inherently a selection of failures and do not necessarily represent the uncomplicated in vivo behavior. Yet, these explants are telling in determining the factors that can accelerate damage mechanisms. This chapter concentrates on the factors that affect the damage modes and the in vivo performance of the total knee. First, the historic design changes directed for the reduction of in vivo damage of tibial inserts and patellar components are outlined. The basic structure of polyethylene, mechanisms of in vivo damage, and factors that affect these damage modes are reviewed. Last, newly emerging technology of antioxidant stabilization of cross-linked polyethylene, modification of articular surfaces, and other technologies for use in the total knee are presented.

TIBIAL INSERTS

Total knee designs in the 1950s and 1960s were predominantly hinged devices such as the Guepar prosthesis.⁸,⁹ These knees, by design, had a high degree of constraint between the tibial and femoral components, which led to high bone-cement interface shear stresses and, subsequently, a very high incidence of early failure secondary to aseptic loosening and infection.⁹ To remedy this situation, resurfacing arthroplasty was introduced for the knee in the early 1970s by Gunston¹⁰ with the polycentric knee. This design resurfaced the distal femur and proximal tibia with metallic and polyethylene runners, respectively. Among the downfalls of the polycentric knee were the failure to address the patellofemoral joint and a lack of instrumentation to allow reproducible implant insertion. The introduction of the total condylar design by Insall and associates in the mid-1970s revolutionized TKA.¹¹ This design was the predecessor of modern designs, as it allowed resurfacing of all three compartments of the knee, provided instrumentation for more reproducible insertion, and accommodated the collateral ligaments about the knee as no previous design had done. Subsequent to the total condylar design, knees fell into one of the two categories: posterior cruciate retaining or posterior cruciate substituting. The posterior cruciate-substituting designs substituted for the posterior cruciate ligament with a polyethylene post on the tibial component that articulated with a cam on the femoral component. This divergence in philosophy was fueled by controversy
regarding the function of the posterior cruciate ligament with respect to knee kinematics and anteroposterior stability after TKR. Both designs are still used and both have equally good long-term results to date. The advances in surgical technique, instrumentation, and component design (posterior stabilized and cruciate retaining) have made the operation more reproducible and markedly reduced the incidence of aseptic loosening.

TABLE 13-1 Historical and Current Ranking of Top Reasons for Revision

	Sharkey et al, 2002¹		Sharkey et al, 2014²
Reason for Revision	Ranking (Early)	Ranking (Late)	Ranking (Early)	Ranking (Late)
UHMWPE	4	1	4	4
Loosening	3	2	2	1
Instability	2	3	3	3
Infection	1	4	1	2

Further advances in the total knee designs accompanied the use of modularity. Initially, tibial components were either all-polyethylene, or the polyethylene tibial surface was directly compression molded onto a tibial base plate.¹² Both types of tibial components were fixed using bone cement. Modularity allowed the surgeon to exchange the polyethylene tibial insert at revision surgery instead of revising the tibial metal plate in the event that the tibial component was well-fixed and aligned. Furthermore, the use of modular trial inserts and spacer blocks has allowed the refinement of accurate joint line placement and flexion-extension gap balancing. Today, the use of modular components is widespread and polyethylene inserts are made through a variety of methods that are discussed later (see Ultrahigh-Molecular-Weight Polyethylene).

PATELLAR COMPONENTS

In the early total knee designs, resurfacing of the patella was not common.¹³,¹⁴ The long-term benefits of patellar resurfacing are still controversial. Although some surgeons always resurface the patella, some never perform resurfacing, and some resurface based on clinical or intraoperative findings.¹⁵,¹⁶ The incidence of postoperative anterior knee pain is believed to be lower with the resurfacing of the patella.¹⁷,¹⁸ In a 2018 systematic review and subsequent meta-analysis, the postoperative Knee Society (pain) score was significantly higher in the patellar resurfacing compared with the nonresurfacing group (OR 1.52, P = .004). Also, the percentage for reoperation is lower for the patellar resurfacing group (1% vs. 7%).¹⁹

Patellar complications are a source of poor outcome and require revision surgery after TKA.²⁰,²¹,²²,²³ Excessive wear, fracture, and maltracking of the patellar components have been associated with such complications²⁴,²⁵,²⁶,²⁷,²⁸ and have been related to the surgical alignment and design.²⁹ Berger et al³⁰ have shown the importance of surgical alignment with a study of the ramifications of femoral rotation on patellar tracking. The combined internal rotation of the tibial and femoral components, judged from computed tomography scan measurements, correlated directly with the severity of patellofemoral complication in a group of 30 patients, suggesting that rotational malalignment may be the most important cause of patellofemoral complications. Among the design aspects that stimulate patellar complications are the polyethylene thickness, the nature of the backing (cement versus metal), and the articular surface geometry.³¹,³²,³³ Furthermore, femoral design has been shown to affect the function of the patellofemoral joint. A femoral design with a trochlear groove that extends proximally to engage the patella in full extension, a deepened trochlear groove, and a gradual anterior to distal transition are desirable to optimize patellofemoral function after TKA.

ULTRAHIGH-MOLECULAR-WEIGHT POLYETHYLENE

UHMWPE is the material of choice for the fabrication of tibial inserts and patellar components and is used in all arthroplasties on the tibial side of the articulation regardless of the choice of counterface material on the femoral side. UHMWPE is synthesized by the polymerization of ethylene (CH₂=CH₂) gas leading to the formation of long-chain molecules with the chemical formula of (-C₂H₄-)n, where n is the degree of polymerization. As the degree of polymerization increases, so does the molecular weight of the polymer. The nomenclature of UHMWPE has changed considerably in the past five decades beginning in the early 60s as a form of HDPE³⁴ and it became possible to produce even higher-molecular-weight polyethylenes.

Today, UHMWPE is defined by ASTM D 4020 as a linear polyethylene with an average molecular weight of greater than 3.1 million g/mol³⁵ and greater than 1 million g/mol by ISO 11542. The first use of UHMWPE in orthopedics was in 1962 with Charnley,³⁶ who used the RCH 1000 resin—a grade of polyethylene available at the time—for his original acetabular components.
The available surgical grades of UHMWPE resins today are GUR1020 and GUR1050 (Celanese, Houston, TX) and have molecular weights of about 3.5 and 6 million g/mol, respectively.³⁷ Today, there is also a version of GUR1020 supplied premixed with 1000 ppm of vitamin E (GUR1020E). Until the disassembly of their production line in 2002, two other resins, 1900 and 1900H, which had a molecular weight in the range of 2 to 4 million g/mol, were supplied by Hercules Powder Company and Montell Polyolefins (Wilmington, DE, USA). The primary difference between the currently available resins is their molecular weight, which can affect their mechanical properties and adhesive-abrasive wear resistance. The Himont 1900 resin, although it exhibited the lowest mechanical properties among the three resins, also found widespread use in TKR devices primarily for its low incidence of delamination when consolidated through direct compression molding. The clinical comparison of the same implant made with the Himont 1900 resin and another resin such as the GUR4150 in MG-I and MG-II prostheses showed more delamination and oxidative damage with the latter resins³⁸ and suggested that resin type, consolidation method, or a combination of the two was effective in decreasing the oxidation potential. Muratoglu et al showed that oxidation is the highest at flake boundaries, in which calcium stearate contained in GUR4150 accumulated during consolidation.³⁹,⁴⁰ Calcium stearate was added to GUR 4150 resin to protect the consolidation equipment from corrosion. It also acted as a lubricant and a release agent. While the Himont 1900 resin is no longer in production, the currently available GUR UHMWPEs do not contain calcium stearate anymore to avoid associated complications.

At the nanometer scale, UHMWPE is a semi-crystalline polymer with its crystalline domains embedded within an amorphous matrix. At room temperature, UHMWPE molecules pack themselves in a long-range order, forming crystalline lamellae, typically 10 to 50 nm in thickness and 10 to 50 µm in length (Fig. 13-4). There is a wide distribution of lamellar sizes within the polymer with the surrounding amorphous phase consisting of randomly oriented and entangled polymer chains traversed by tie molecules which interconnect lamellae and provide resistance to mechanical deformation.

After synthesis, UHMWPE is in the form of fine white powder resin particles which go through a consolidation process for the long polymer chains at the granule boundaries to diffuse effectively into the neighboring granules and fuse the polymer into stock forms. Consolidation into a stock material is done using ram extrusion, slab molding, or direct compression molding methods (Fig. 13-5). For the early designs of tibial inserts, for example, the MG-I and Insall-Burstein I (IB-I), the UHMWPE powder was directly molded into the metal base plate. Today, with the popularity of modular designs, the tibial inserts and patellar components are mostly manufactured by machining a ram-extruded or slab compression-molded stock material or by direct compression molding of the near-net or final shape.

FIGURE 13-4 A transmission electron micrograph (A) and a schematic (B) of ultrahigh-molecular-weight polyethylene show the lamellae embedded in an amorphous matrix. The long-chain polyethylene molecules assume a random orientation in the amorphous regions. In the crystalline lamellae, the molecules are oriented in a long-range order.

The mechanical and wear properties of UHMWPE may be significantly affected by the manufacturing processes, such as the temperature and pressure cycles during molding and as such these processes are proprietary. The mechanical properties of UHMWPE are strongly related to the crystalline content and structure and the amount of interaction between its crystalline and amorphous phases. For example, the mechanical properties of consolidated UHMWPE vary as a function of the cooling rate during consolidation,⁴¹ which affects the crystallization kinetics and may lead to lower crystallinity at higher cooling rates. Final implant shapes are further fashioned
by either extensive machining stock material or minimal machining of the near-net shaped components obtained from direct compression molding. The final process in implant manufacture is sterilization. While gamma irradiation is the most common method of sterilization, ethylene oxide (EtO) and gas plasma sterilization are also used. Radiation processes including gamma irradiation produce long-lived free radicals in UHMWPE, the reactions of which with oxygen are a major factor in the long-term oxidation and degradation of the implant. One of the improvements in the technologies used in total knee implants in the last two decades has been radiation cross-linking of the UHMWPE-bearing surface to improve wear resistance.⁴² We will discuss these processes and associated treatments to reduce oxidation in detail in further sections. Today, EtO and gas plasma sterilization, which do not introduce long-lived free radicals in UHMWPE, are also used in combination with radiation cross-linked materials to avoid increasing their oxidation potential (Table 13-2).

FIGURE 13-5 Schematic of ram extrusion (A), slab molding (B), and direction compression molding of ultrahigh-molecular-weight polyethylene powder (C). During ram extrusion, resin powder is consolidated to form bar stock. During slab molding, the mold is filled with resin powder and consolidated under high pressure and high temperature. Direct compression molding allows the consolidation of resin powder into the final geometry of the tibial insert. In the case of ram extrusion and slab molding, the consolidated stock is machined to fabricate tibial inserts and patellar components.

DAMAGE ON ARTICULATING AND NONARTICULATING SURFACES

During the activities of daily living, the articular surfaces of polyethylene components accumulate a scar. On tibial inserts, the scar is commonly localized to the articular condylar surfaces. Occasionally, damage may occur elsewhere, such as the tibial eminence. The scarring of the eminence may be induced by medial-lateral translation or, more typically, by a malrotated tibial component.⁴³ Rotation of the femoral component after progressive damage on the articular surface of the tibial insert is another mechanism that has been discussed in the literature.⁴⁴

Another location where polyethylene surface damage occurs is the undersurface of tibial inserts. This damage is primarily a consequence of micromotion between the insert and baseplate, which produces small particles through adhesive wear, and has been commonly called “backside wear.” It is estimated to contribute as much as 30% to the total wear volume in TKR.⁴⁵ In addition, the screw holes present on the tibial baseplate are often imprinted on the undersurface because of the effect of creep deformation of the polyethylene tibial insert into these openings.

The tibial post-cam articulation of posterior-stabilized (PS) designs has also been shown to contribute to the generation of particles. Puloski⁴⁶ reported on 23 retrieved PS tibial inserts of four different manufacturers (nine different designs), which had been implanted for a mean duration of 36 months, and performed quantitative and qualitative analysis of wear. About 30% exhibited severe damage with gross loss of polyethylene. Although the wear was primarily posterior at the post-cam articulation, damage of the medial, lateral, and anterior surfaces of the post was observed. Factors such as cam-post mechanics and geometry may influence the wear at this articulation. On occasion, gross wear and even fracture of the post have also been observed.⁴⁶,⁴⁷ Post fracture is a rare event and may be accelerated by the oxidative embrittlement of polyethylene.

WEAR MECHANISMS

There are several mechanisms of wear underlying the formation and progress of the scar on the polyethylene tibial and patellar components. The four main mechanisms are surface fatigue, adhesion, abrasion, and tribochemical reactions as defined by ASTM standard G40-05 and reported in the orthopedic literature.⁴⁵,⁴⁸

Surface fatigue results from repeated stress cycles in the subsurface of a material, during repeated sliding or rolling over the same wear track. The application of load through the femoral component on the tibial insert leads to compressive stresses immediately under the contact area and tensile stresses outside the contact.⁴⁹,⁵⁰ The peak of the shear stresses occurs approximately 1 mm below the surface (Fig. 13-6A). Because of the rolling-sliding motion of the femoral component, the contact region of the polyethylene is subjected to a fluctuating stress environment (Fig. 13-6B), leading to crack initiation in the material. The formed microcracks can grow below the surface, and when they reach the surface cause detachments of large flaky particles, also known as delamination (Figs. 13-3 to 13-3). Delamination differs from other wear types as it is not a gradual release of material, but rather accumulated damage followed by surface destruction that leads to device failure in vivo. Pitting is also a fatigue-related phenomenon and is the outcome of the coalescence of shallow cracks or cracks initiated from the surface. Yet, this mode of damage mostly occurs on the articular surfaces without substantial material loss. In the absence of surface delamination, the adhesive-abrasive mechanisms prevail as the dominant wear processes in total knees.

Adhesion occurs when there is local bonding between asperities of both articulating surfaces under pressure. During relative motion, these microjunctions are torn off and fragments may form particles or are being transferred to the counterbody. Lubricant chemistry and film thickness between the two surfaces are important factors allowing the asperities to come into contact and bond. The rate of adhesive wear depends strongly on the relative motion of the metallic counterface against the polyethylene articular surface.⁵¹,⁵²,⁵³ For example, cross-shear increases wear by an order of magnitude by shifting from a unidirectional, reciprocating wear path to a rectangular, bidirectional wear path when it is rubbed against an implant surface finish cobalt-chrome disc in the presence of bovine serum.⁵⁴ The unidirectional motion used in this pin-on-disc (POD) experiment is thought to orient and strain-harden the surface of polyethylene, resulting in low
wear. On the other hand, with the rectangular, bidirectional motion, the surface strain hardens (strengthens) in the rubbing direction and weakens in the transverse direction. The metallic counterface encounters a weaker surface during changes in the direction of motion, leading to high wear. The microscopic analysis of the wear surfaces of surgically retrieved tibial inserts has shown striated wear patterns with ripples oriented perpendicular and parallel to the primary direction of anterior-posterior motion.⁵⁵ These striations (Fig. 13-7) represent amorphous and crystalline regions of polyethylene⁵⁶ and develop during adhesive-type wear of polyethylene in the total knee.

TABLE 13-2 Current Alternatives of Medium to Highly Crosslinked Virgin (No Antioxidant) Polyethylenes in Total Knee Replacements

Trade Name	Manufacturer (Location)	Radiation Temperature (°C)	Radiation Dose (kGy^a)	Radiation Type	Postirradiation Treatment	Sterilization Method	Total Radiation Dose Level (kGy^a)
Prolong	Zimmer Biomet (Warsaw, IN)	˜125	65	E-beam	Melted at 150°C	GP	65
ArCom XL	Zimmer Biomet (Warsaw, IN)	N/A	˜50		Mechanical deformation	EtO	˜50
Durasul	Zimmer Biomet (Warsaw, IN)	˜120-125	95	E-beam	Melted at 150°C	EtO	95
X3	Stryker (Mahwah, NJ)	N/A	˜90	Gamma	Sequentially annealed (˜130°C) after each 30 kGy	GP	˜90
XLK	Depuy Synthes (Warsaw, IN)	N/A	˜50	Gamma	Melted at 155°C	GP	˜50
XLPE	Smith & Nephew (Memphis, TN)	N/A	˜75	Gamma	Melted at 147°C	EtO	˜75
EtO, ethylene oxide; GP, gas plasma.
¹⁰kilograys (kGy) = 1 Mrad.

FIGURE 13-6 A: Shear stress contours shown cross section in a polyethylene implant. Note the maxima below the surface (FEM model from Mell et al²⁹⁹). B: Shear stress over time during stance phase of walking. Note the cyclic character of the stress pattern when viewed at a single location.

Abrasion defines the action of hard counterface asperities or hard debris (so-called “third bodies”) ploughing the softer material. Both hard asperities and third-body particles cause the release of polyethylene from the articular surfaces. Submechanisms include microcutting, microploughing, and microfatigue and are dependent on the material properties and sharpness of the asperities and/or third bodies. Third-body wear is common in TKR and can be generated by the fragmentation of bone cement used for fixation, bone chips, or metal debris from failure of metal-backed patellar components. Wasielewski reported on 55 tibial inserts retrieved at the time of revision surgery and found that 25% had severe insert damage related to third-body debris consisting of cement or metallic particles.⁵⁷ Third body can penetrate the articulation and accelerate the wear process, mainly in the form of material removal through extensive scratching of the polyethylene articular surfaces. With hard third-body particles, such as the barium sulfate additive in bone cement, scratching of the femoral counterface can occur, increasing the surface roughness. This in turn increases polyethylene wear.

Tribochemical reactions occur when surfaces in mechanical contact are activated due to friction and react with the interfacial medium, which results in the alternating formation and removal of chemical reaction products
at the surfaces. Whether this mechanism has any significance in TKR is currently unknown.

FIGURE 13-7 Striated (“lace”) morphology (A) and surface ripples (B) on the articular surface of a surgically retrieved polyethylene tibial insert. These features are the likely precursors to adhesive-abrasive wear on tibial inserts.

Adhesive and abrasive wear has been implicated in the generation of submicron polyethylene debris, capable of initiating adverse tissue reactions and periprosthetic osteolysis in vivo. The following section discusses the clinical observations on wear debris found in periprosthetic tissue and osteolysis in TKA.

WEAR APPEARANCE AND VOLUMETRIC MATERIAL LOSS

Unlike total hip replacement, in which the in vivo wear rate of polyethylene has been known through various methods, the true rate of TKR has only been established in the past few years based on retrieved UHMWPE components. There are no reliable radiographic methods developed for the quantification of in vivo wear, mainly because of the complex geometry of the knee implants.

Wear appearance, which describes the visible changes of surface texture, composition, or shape as a consequence of wear, has been used as a surrogate for cumulative damage in many research studies. “Wear features,” “wear patterns,” or “wear damage” has been used synonymously in the literature. Originally developed by Hood et al,⁵⁸ the methodology has been cited over 350 times and applied to multiple retrieval studies. Polishing, burnishing, scratching, pitting, delamination, striation, abrasion, embedded debris, and surface deformation are typical wear features on retrieved polyethylene tibial inserts (Fig. 13-8). While the appearance of damage holds clues regarding the identification of the acting wear mechanism, damage patterns are only moderate predictors for material loss (except for delamination) as recently demonstrated in a study by Knowlton et al.⁵⁹

FIGURE 13-8 Various damage modes seen on the tibial polyethylene surface.

The burnished feature occurs due to an adhesive wear mechanism that tears and pulls off fibrils during sliding of the femoral condoyle on the tibial insert. Scratches have linear features on the articulating surface, often due to the presence of third-body particles that act abrasively. Pitting is a common wear feature characterized by mostly round, small holes (<1 mm) at the surface.
Pits can either be caused by local surface fatigue and ejection of material or by plastic deformation when third bodies (e.g., cement particles) are indented into the surface. Delamination is the most extreme wear feature and caused by surface fatigue as described above. It frequently occurred on components that were sterilized in air, accelerating oxidation and causing embrittlement of UHMWPE.⁶⁰

Striations are a unique and frequent wear pattern that typically occurs within the first year after implantation of retrieved polyethylene inserts. The width of a single striation is about 70 to 100 µm and the darker elevations are separated from the brighter regions that form troughs in between.⁵⁵ Interestingly, it was recently found that striations accounted for the second highest material loss after delamination.⁶¹ While little is known about the actual mechanism that causes striations, a recent FTIR microscopy study by Rad et al⁵⁶ suggests that striations are an arrangement of crystalline and amorphous regions. Other damage features that are reported in the literature are “abrasion,” which is a rough and tufted region that can be visualized on the articulating surface of the tibial inserts, and “embedded debris,” characterized by color or texture differences that suggest embedded particles in the polyethylene surface.⁶² Deformation, which is a permanent change of the surface geometry due to plastic flow or creep, is also often mentioned. Obviously, this type of damage does not correspond to wear since no material is lost.⁶⁰

The appearance of the scar observed on surgically retrieved patellar components is similar to those observed on tibial inserts. They generally develop on the lateral facet and the dome of the patellar component. Cameron⁶³ reported this type of scarring in 11 surgically revised Freeman-Swanson and Tricon (Smith & Nephew, Memphis, TN) knees. Scarring of the lateral facet of the patellar component is presumably induced by lateral loading of the patellar surface and excessive lateral subluxation or tilt, or both, of the patella during flexion, often associated with internal rotation of the femoral and tibial component.

Volumetric loss of material has been difficult to measure in TKR for various reasons. First, the loss of material is relatively subtle when delaminated components are excluded. Pourzal et al⁶⁴ found up to 60% less wear in total knees when compared to total hip components made out the same polyethylene. Next, the volumetric loss is easily confounded by plastic deformation/creep. Third, the geometry of TKR is complex, and tolerances are not always tight. Therefore, techniques had to be developed that allowed measurements from reconstructed surfaces, in an attempt to estimate the original surface.⁶⁵ Such methods also allowed differentiating wear from conformational changes due to plastic deformation and creep. Other studies approximated wear by linear thickness measurements or stereoradiography. A summary of measurements that can be found in the literature is given in Table 13-3. It should be noted that some of the studies listed (e.g., CMM autonomous reconstructions) do not account for backside wear.

OSTEOLYSIS IN TOTAL KNEE REPLACEMENTS

Wear particles can trigger osteolysis in total joint arthroplasty.⁵ This is true for both hip and knee joints, and the appearance of tissues surrounding prosthetic devices that have failed aseptically is similar between total knees and hips. In fact, Goodman et al⁶⁶ reported that these tissues are similar regarding cellular profile, structure, chemokine/cytokine signaling, and enzymatic profiles.

Compared with total hip, total knee patients exhibit a reduced rate of osteolysis, which is thought to be a consequence of fewer particles due to a smaller wear volume, but also the larger and more elongated polyethylene wear particles produced in the total knee.⁶⁷,⁶⁸,⁶⁹ Shanbhag et al performed an analysis of particulate wear debris retrieved from interfacial membranes around 18 failed total knees.⁶⁹ The mean particle size was 1.7 ± 0.7 µm, and this increase in size may contribute to a decrease in the biologic activity of particles. Still, 30% of the particles were less than 1 µm in size. We know now that a high proportion of the small particles is generated by backside wear.⁴⁵,⁷⁰ With backside wear, the debris can find access to the periprosthetic interface through screw holes. Peters et al⁷¹ showed the development of lytic lesions around screw-bone interfaces induced by polyethylene and metallic debris from modular interfaces.

Ultimately, particle size and total volumetric material loss of the polyethylene liner are dependent on implant design.⁷² In addition, the particle load experienced by the tissue is in part a function of the joint space available, which makes the distribution of wear products less dense around knee arthroplasty when compared to hip arthroplasty. As a result, knee replacements demonstrate a lower number of particles per unit of surface compared to hip replacements even in the case where particle size and worn off material volume are identical. Therefore, osteolysis has been less frequent in TKR.

FACTORS AFFECTING DAMAGE AND SOLUTIONS TO MINIMIZE WEAR

The damage observed in tibial and patellar components is multifactorial and contributing factors will be discussed in detail below. Various damage mechanisms may require various solutions, and some may contradict each other. For example, to avoid surface fatigue, a large contact area with high conformity is needed. However, adhesion is reduced when the contact area is
minimized. Thus, an optimum between both should be chosen to balance these requirements.⁷³ The present section discusses factors that affect polyethylene wear and damage in detail.

TABLE 13-3 Volumetric Wear of Tibial Components

	Study	Knee Design (Type)	Assessment	Avg. Volume Loss [µL/y or /Mc]	No. of Studied Retrievals
In vivo	Kop et al²⁸⁹	APG, LCS (MB, FB)	CMM	85 77	7 10
	Atwood et al²⁹⁰	LCS (MB)	Thickness-based	54	100
	Engh et al²⁹¹	Sigma RP, LCS, PFC Sigma (MB, FB CR/PS)	Micro-CT	43 ± 25 74 ± 49	12 12
	Benjamin et al²⁹²	AMK (FB CR)	laser scan	794	33
	Lavernia et al²⁹³	PCA (FB-CR)	Thickness-based	32 ± 43	28 (PM)
Gill et al²⁹⁴	AGC (FB-CR)	RSA	100	6
	Pourzal et al¹³⁹	MGII (FB-CR)	CMM/autonomous reconstruction	18 ± 13 (no delamination)	17 (PM)
	Rad et al¹³⁷	NexGen (FB-CR)	CMM/autonomous reconstruction	13.8 ± 2.0	83
	Knowlton⁷⁷	NexGen (FB-CR/PS)	CMM/autonomous reconstruction	10.3 ± 6.5 (CR) 11.1 ± 4.2 (PS)	19 (PM) 25 (PM)
In vitro	Muratoglu et al⁴²	NKII (FB-CR)	CMM Gravimetric	8.3 ± 0.8 8.8 ± 1.5
	Schwenke et al²⁹⁵	MGII (FB-CR)	Gravimetric	22.4 (load control)−9.8 (display control)
	Teeter et al²⁹⁶	Genesis II (FB-CR)	Micro-CT	24.7
	Laurent et al²⁹⁷	NexGen (FB-CR)	Gravimetric	15.4 ± 1.2
	Popoola et al²⁹⁸	NexGen (FB-CR)	Gravimetric	15.4 ± 3
CR, cruciate-retaining; FB, fixed bearing; MB, mobile bearing; PM, postmortem; PS, posterior-stabilized.

Kinematics and Design

The kinematic conditions are determined based on the design, such as the retention or substitution of the posterior cruciate ligament, and conformity of the tibiofemoral articulation. In general, cruciate-retaining (CR) designs have less conforming tibiofemoral articulations than their posterior stabilized counterparts. This, in theory, allows the ligaments about the knee to exert their effect on the kinematics. In a video-fluoroscopy study comparing PS to CR knees, Dennis et al found a paradoxic anterior femoral translation with knee flexion in the CR design.⁷⁴ The posterior stabilized knees consistently exhibited posterior femoral rollback. In a similar study, Banks et al⁷⁵ studied the differences in the kinematics of CR and cruciate-sacrificing total knee designs and found that axial rotation and condylar translations decreased with the post-cam substitution of the posterior cruciate ligament. These design-related differences in tibiofemoral conformity and kinematics may be related to the distinct differences in wear behavior in CR and PS total knees described by Hirakawa et al,⁷⁶ who showed smaller scars and reduced apparent wear with PS TKA. However, in a recent study by Knowlton,⁷⁷ who compared the wear of CR and PS knees of the same manufacturer and design family (Nexgen, Zimmer Inc.), no volumetric wear differences were found. Both groups were comprised of postmortem retrievals (19 vs. 25, respectively), which were on average 9 years in situ. The wear rates were 11.9 ± 5.0 mm³/y and 11.1 ± 4.2 mm³/y for the CR and PS design, respectively (wear on post of PS design not considered). There was also no statistically significant difference regarding creep and plastic deformation, although the CR design displayed a higher amount on average (70 vs. 32 mm³).

The state of kinematics determines the wear path that the femoral component follows on the articular surface of polyethylene. In the knee, the sliding motion, which is induced by the flexion-extension (typically 50 to 90°) and anteroposterior translation (typically 5 to 20 mm), produces unidirectional motion on articulating surfaces of the tibial insert. When superimposed on the sliding motion, any level of tibial rotation generates multidirectional motion. Unidirectional motion produces very little wear, whereas multidirectional motion is what causes an increase in wear rates.⁵⁴ Schwenke and Wimmer⁷⁸ experimentally determined the relationship between cross-shear motion and UHMWPE wear using a cobalt-chromium
wheel that articulated in a gliding-rolling fashion on a flat polyethylene plateau. They found that it takes 6.4 times more work to remove a unit wear volume in the direction of principal motion (i.e., anteroposteriorly) than 90° perpendicular to it. This explains findings of Kawanabe et al,⁷⁹ Wang et al,⁸⁰ and Muratoglu et al⁸¹ who showed that the wear rate of polyethylene tibial inserts increases with increasing extent of internal-external rotation of the tibia on in vitro knee simulators. Specifically, the study by Kawanabe et al⁷⁹ demonstrated that the addition of a ±5-degree tibial rotation increased the wear rate from 1.7 mg per million cycles to 10.6 mg per million cycles on a cruciate retaining design (AGC Biomet). Wang’s study⁸⁰ showed a decrease in the rate of weight loss from 14.4 mm³ per million cycles with a 13.5-degree internal-external rotation to 3.9 mm³ per million cycles with no rotation of the tibia. Muratoglu’s study⁸¹ showed that the gravimetric weight loss of conventional polyethylene in a flat-on-flat design TKR (Natural-Knee II) increases from 0.4 to 23.0 mg per million cycles, with an increase of tibial rotation range from 5 to 14°. Excessive tibial rotation can also lead to damage on the intercondylar eminence further contributing to weight loss. Another kinematic consideration that has been discussed in the past is the potential separation of the femoral component from the tibial insert during articulation. Dennis et al⁸² have fluoroscopically measured the occurrence and the degree of femoral liftoff with posterior-stabilizing and CR knees. The femoral condylar liftoff on one side of the tibial insert could lead to an edge loading condition on the opposite side. If the mechanical properties of the polyethylene tibial insert have degraded owing to oxidative embrittlement, liftoff could conceivably accelerate surface fatigue and lead to rapid delamination of the component.

Although the tibiofemoral kinematics substantially affects the adhesive wear behavior of tibial inserts, patellofemoral kinematics, such as the tracking, and patellar rotation and tilt influence the damage accumulation on the patellar components. The improved understanding of patellofemoral kinematics has led to alterations in surgical technique to improve patellofemoral tracking. These include obtaining proper rotation of the femoral components, slight lateral placement of the femoral component, and medial placement of the patellar component.⁸³ In a recent simulator study⁸⁴ that was performed under a malaligned conditions, it demonstrated that a constant 5° external rotation applied to the patella button can result in delamination if the polyethylene has undergone embrittlement due to oxidation. However, many of the observed patellar complications could be attributed to the design of the patellofemoral articulation. Design modifications with respect to the patellofemoral joint have decreased the incidence of patellofemoral complications. Berger reported on a series of 172 MG-I and 109 MG-II cemented total knees with a mean follow-up of 11 years. The MG-I and MG-II are similar in terms of the tibiofemoral articulation; however, the MG-II has a more anatomic sagittal contour of the femoral sulcus, resulting in a decreased anterior radius of curvature of the patellofemoral articulation. In addition, the patellofemoral articulation was made more congruent, and an all-polyethylene patella was used instead of a metal-backed patellar component to decrease the stresses in the polyethylene. In this series, the MG-I group had a 9% prevalence of patellofemoral complications compared to 0% in the MG-II group.⁸⁵ In a similar series, Theiss and associates studied a group of 301 cemented primary total knees, 148 with the MG-I and 153 PFC prostheses followed for a minimum of 2 years postoperatively.⁸⁶ The patellofemoral complication rate was 10% for the MG-I group compared to 0.7% for the PFC group. From these studies it is apparent that patellofemoral complications are sensitive to design issues as well as rotational alignment of the femoral component. An anatomic sagittal contour of the femoral prosthesis, a congruent trochlear groove with a cemented all-polyethylene patellar component, and proper rotational alignment markedly reduced the prevalence of patellofemoral complications after TKA.

Oxidation

Oxidation and oxidative degradation constituted the most pervasive problem in total joint knee arthroplasty for several decades.⁸⁷,⁸⁸ Historically, the polyethylene used in total knees was sterilized using gamma irradiation in air with a typical dose of 25 to 40 kilograys (kGy). These implants had high wear, high rates of degradation, and high rates of revision.⁸⁹ In 1995, gamma sterilization in air of UHMWPE implants was abandoned and implants started being sterilized while packaged in inert gas.⁹⁰ Gamma sterilization in air led to marked oxidative changes in polyethylene, which resulted in the significant degradation of its mechanical properties. This degradation was characteristically manifested in the form of a reduction in strength and ductility and an increase in modulus.⁹¹,⁹²,⁹³,⁹⁴ In the mid-1990s, these sterilization-induced changes in polyethylene were shown to strongly correlate with delamination, subsurface cracking, and pitting of tibial inserts.⁸⁹,⁹⁵,⁹⁶,⁹⁷ Today, all gamma sterilization of polyethylene implants is carried out in an inert environment, such as nitrogen, argon, or vacuum. The oxidation rate observed in these implants has been much lower under shelf-storage than those sterilized in air; however, the oxidation rate in vivo still is measurable and progressive.⁹⁸,⁹⁹

Concerning gas sterilization methods, a study reported by Williams et al⁹⁵ demonstrated that, after ethylene oxide sterilization, UHMWPE-bearing surfaces showed no cracking or delamination, even after 15 years of in vivo use. Although very few implants are only processed by gas sterilization due to the general low wear resistance in nonirradiated surfaces,³⁰⁰ gas sterilization as a terminal process is considered, especially in cases where radiation can be detrimental to the performance of the implanted
material. The understanding of oxidation and oxidation potential in UHMWPE used for joint implants has been based largely on radiation-induced free radicals. During ionizing irradiation of UHMWPE (for sterilization as well as for cross-linking which will be discussed later), free radicals are formed, the most prevalent of which are the carbon free radicals resulting from the breakage of the C-H bonds.¹⁰⁰,¹⁰¹,¹⁰²,¹⁰³ Most of these free radicals recombine in the amorphous portion of the polymer,¹⁰⁴ resulting in cross-linking. The remaining free radicals are trapped in the crystalline lamellae.¹⁰⁵,¹⁰⁶ Oxygen reacts with the primary free radicals to form peroxy free radicals.¹⁰³,¹⁰⁷,¹⁰⁸,¹⁰⁹,¹¹⁰ These peroxy radicals abstract a hydrogen atom from other polyethylene chains, creating primary free radicals, which can then react with oxygen to further this cascade.¹¹¹,¹¹² The reactions of polyethylene free radicals with oxygen and the decay of the formed hydroperoxides eventually result not only in carbonyl-containing species, which are defined as “oxidation” in UHMWPE, but also in chain scission, lowering the molecular weight of the material and degrading its material properties.¹¹³ Oxidation is determined by spectroscopic techniques measuring carbonyl moieties on UHMWPE formed as hydroperoxides decay.¹¹⁴

Only gold members can continue reading. Log In or Register to continue