Modern trends in total joint arthroplasty indicate a sharp increase in demand with the inclusion of younger and more active patients. Implant survival curves suggest increased implant durability and previously described wear patterns are less relevant.¹ The introduction of highly cross-linked polyethylene two decades ago has resulted in less particulate wear and decreased osteolysis. Nevertheless, aseptic loosening is still the prevalent cause of implant failure, significantly outpacing the rates of periprosthetic infection, instability, and mechanical breakage.² Persistent development and changes in implant surface design and biocompatibility have had a relatively small impact on outcomes. The original Charnley hip has shown 85% survival at 25 years, which is very impressive considering the use of small diameter heads, cement, and early generation polyethylene formulations.³ Early data on modern polyethylene show no penetration on radiographs at 10 years, suggesting very limited wear.⁴ Does that suggest that implant wear, osteolysis, and loosening are problems of the past? To answer that question, the current chapter will focus on the biology of osteolysis, cellular interactions, molecular signaling cascades, and potential therapies for osteolysis and associated bone loss.

Even though implants are generally designed to be relatively inert, our bodies interact with them at the cellular and molecular levels. A foreign-body reaction is mediated by a cascade that results from (1) associated injury, (2) local hematoma, (3) recruitment of host inflammatory cells, (4) acute to chronic inflammation, and (5) fibrosis or granuloma formation. Foreign-body reactions can originate at the time of implant insertion through immediate tissue damage, or later due to wear particles and debris. Wear may be defined as loss of material volume with generation of material particles resulting from motion between two opposed surfaces.⁵ In the case of joint replacement implants, such wear occurs at the bearing surface, where most motion takes place, or at the implant-bone interface. The inflammation produced locally leads to bone loss termed osteolysis. The initial loss of bone support around the implant leads to increased mechanical strains and micromotion. Cyclical micromotion then acts as a pump with subsequent pressurization that carries particles to other areas of the interfaces and can produce massive and extensive osteolysis.

The original design of total hip arthroplasty involved a polytetrafluorethylene (Teflon) bearing surface, and Charnley documented a very high failure rate of 95% between 1959 and 1962, mostly from osteolysis and implant loosening. Nevertheless, the process of osteolysis was not recognized as a biologic event until both Willert and Harris separately
described the lysis derived from macrophages around loose implants, which was later mistakenly termed “cement disease” by Jones and Hungerford in 1987.⁶ The so called “cement disease” was thought to be the product of cement particles produced around a total implant from abnormal motion and poor cement technique.⁷ Methyl methacrylate, a monomer of acrylic resin, has revolutionized total joint arthroplasty by allowing early fixation of implants into bone. Nevertheless, loosening at the implant-cement interface or cement-bone interfaces allowed for generation of small debris that initiated the inflammatory cascade and led to extensive osteolysis. In severe cases of “cement disease,” the entirety of the proximal femur and pelvic columns can show resorption and associated bone loss (Figure 1). The severity of the problem and the hypothesis that it was related to the cement itself led to the design of noncemented total hip arthroplasty in the 1980s, only to see the persistence of osteolysis. Osteolysis remains a major mechanism of failure in total joint arthroplasty, now reported around 21.8% in revision hip arthroplasty⁸ compared with around 40% several decades ago.⁹

Osteolysis and concomitant loosening of implants can result from a variety of causes in both cemented and noncemented implants. The term “cement disease” has been replaced by “particle disease” through better understanding of the mechanical and physiological processes associated with orthopaedic implants. The typical response to particles can be split into three specific phases. In response to debris, a chronic activation of the immune system takes place through both cellular and cytokine responses. An increased number of macrophages, fibroblasts, and osteoclasts then disrupts the periprosthetic bone environment causing tissue breakdown. Finally, this process disrupts bone regeneration and remodeling by direct cytotoxic effect on mesenchymal osteoprogenitor cells.

Wear debris are constantly produced at joint bearing surfaces, modular interfaces, and other implant-implant or implant-bone interfaces. The particle size, composition, and shape are dependent on the material, with a wide range of particles identified. The accumulation of particles at the bone-implant interface stimulates the recruitment of immune cells, which interact with the particulate debris through direct surface antibodies or phagocytosis by macrophages or lymphocytes. Particles in the 0.2 to 10 µm range involve a macrophage driven response, while smaller metal particles in the nanometer range can elicit a lymphocyte response.¹⁰ Ultimately, particle volume and rate of accumulation leads to a nonspecific immune response with an expansive foreign body reaction and chronic inflammation that mimics chronic infection. Furthermore, in vitro studies have documented a linear increase in transcriptional activity and cytokine responses to particulate volume. A threshold effect has been proposed for wear rate and associated osteolysis, for example, wear rates above 0.1 mm per year are at significant risk of osteolysis.¹¹

Different materials provide for different particulate debris characteristics, including size, composition, and shape, that influence the interactions with the local cellular environment. Polyethylene is the most common source for particle matter around bearings. Polyethylene is a commonly used plastic dating back to 1898 when Hans von Pechmann prepared it serendipitously from diazomethane.¹² The manufacturing process directly affects the material characteristics including shaping by machining or compression molding, irradiation, and thermal processing, as well as end point sterilization and packaging. Conventional polyethylene was sterilized by gamma radiation in air which led to oxidative damage and increased wear and particle production contributing to long-term osteolysis. Contemporary orthopaedic devices take advantage of ultra-high-molecular-weight polyethylene (UHMWPE) which contains extremely long chains of two to six million monomer units. UHMWPE increases cross-linking which improves wear resistance at a cost of decreased mechanical properties and increased free radical formation. In the latest generation polyethylene, advances in radiation and annealing techniques as well as adjuvants such as vitamin E reduce free radicals and improve mechanical and fatigue properties. This does not mean that UHMWPE is less likely to produce osteolysis.

Understanding failure mechanisms is further complicated by the different mechanical environments encountered at various interfaces. The polyethylene wear particles are different in the knee compared with the hip. Failed total knee arthroplasties have large-size debris, which are associated with polyethylene delamination, pitting, and fatigue wear less likely to produce a large macrophage response (Figure 2). The prevalence of osteolysis after total knee arthroplasty ranges from 5% to 20% at 15 years. The knee differs from the hip due to the complex geometry and less congruent surfaces that produce multiaxial movement including sliding, rolling, and rotation. In contrast, the hip is largely exposed to cyclical motion that
leads to wear mainly through adhesion and abrasion mechanisms. Backside wear is not common at the hip, while it remains an important failure mechanism at the knee.

To address the debris problems of polyethylene, alternative bearing surfaces have been proposed including the use of ceramic and metal. Ceramic bearings have the lowest wear rates of any bearing combination, with less than 0.5 to 2.5 µm per year. However, ceramic-on-ceramic bearings have a much smaller window of error in component position, with early designs showing increased squeaking and breakage. Unfortunately, breakage of a ceramic bearing leads to disastrous particulate debris formation that produces extensive osteolysis, with revisions showing poor results due to the persistent particulate presence, third body wear, and continued osteolysis. A broken ceramic liner has to be replaced with another ceramic liner, as other materials such as polyethylene are softer and fail rapidly in response to ceramic debris. Alternatively, ceramic on polyethylene bearings have similar results to metal on polyethylene, even though possible decreased wear in the range of less than 150 µm has been reported. The low wear would suggest ceramic-on-polyethylene to be an ideal bearing, but subsequent possible damage to the ceramic head can lead to increased abrasive wear. Specifically, ceramic heads can show stripe wear as well as lead pencil-like wear, which increase surface roughness but also predispose the head to fracture in overweight or higher activity patients.

Metal-on-metal bearings produce smaller particles and lower wear rates compared with bearings that include polyethylene. An annual linear wear rate of less than 5 µm per year has been reported. The lower wear rates as well as improved material properties allowed for larger heads, increased stability, and theoretically a more forgiving surface. Interestingly, some small head metal-on-metal bearings had acceptable results with long-term follow-up of over 20 years. Unfortunately, some large head metal-on-metal bearings that followed had a high rate of failure due to significant adverse reactions both around the implant as well as the systemic toxicity of plasma metal levels.¹³ The one aspect not accounted for was the significantly higher particulate presence. The metal particles are very small, in the range of 50 nm, but their numbers are up to 500 times higher compared with metal on polyethylene bearings. This significant volume of particulate matter leads to aggressive recruitment of the immune system and development of aseptic lymphocyte-dominated vasculitis-associated lesions (ALVAL) and pseudotumors.¹⁴ This response is significantly different than the macrophage-driven response in polyethylene particle wear. The response to metal particles can be massive, involving extensive tissue necrosis, macrophagic response with phagocytosed metallic debris, granulomas, progressive perivascular lymphocytic infiltrate and ALVAL, lymphoid neogenesis, and large invasive fluid collections. Just like ceramic bearings, the component position is an important factor in bearing surface performance, with overabducted cups showing higher serum ion levels. Cases of adjacent tissue invasion have been widely described, including damage to nearby vascular and nervous planes. Ultimately, the high rate of failure in registries across the world lead to a quick response and recall of some designs, with most patients either now revised or remaining asymptomatic. The problem of metal interfaces has persisted, with fretting corrosion wear now described with modular implants at dual taper neck junctions and trunnions.¹⁵