A significant complication of implantation is micromotion and subsequent loosening of the prosthesis, which may be caused by both biologic and mechanical factors (Table 17.4, Fig. 17.2). Loosening has been estimated to occur in as many as 50 percent of cemented total hip replacements followed for more than 10 years, but can be a sequela of uncemented procedures as well. Mechanical instability may result from excessive activity by the patient or material failures. Biologic causes of loosening include infection and inflammation, the latter usually macrophage-mediated, reactions to debrided particulate matter from joint components. Osteolysis, which may follow all the previously mentioned factors, is another significant cause in that it perpetuates the cycle of loosening (Fig. 17.3).

There also is evidence that intracapsular pressure raised by the surgical procedure itself induces osteolysis (20). For example, the elevated intracapsular pressure that can accompany a hip with loose components may, by pressure, pump debris along interfaces and in and of itself cause osteolysis and loosening (21).

Bacterial endotoxins have also been proposed as a mechanism promoting loosening by modulating the biologic response to wear particles. Substances such as lipopolysaccharide from gram-negative organisms or lipoteichoic acid, and peptidoglycan from gram-positive organisms may originate from otherwise harmless and subclinical intestinal flora, minor infections, and dental procedures, and attach to wear particles acting synergistically to promote osteolysis (22).

Immunologically, metallic ions can serve as haptens and may elicit an adaptive immune response (23). Commonly used alloy metals, for example, nickel, cobalt, chromium, and, rarely, others such as titanium have been implicated as relevant allergens.

Patch-test sensitivity to metals in the normal population is estimated at 10 percent, with nickel, a common component of inexpensive jewelry, far more “allergic” than cobalt or chromium. In functioning implants, it has been assessed as high as 25 percent, and in a loosened implant, >60 percent. Extreme reactivity to metal, however, is unusual and estimated at approximately 1 percent of
the population. Chronically loose implants with their associated metallic debris may sensitize persons who were previously without sensitivity. However, significant immunologic reactions, including dermatitis and eczema, have not been widely reported.

It would nonetheless appear prudent to advise a patient who has a known dermal allergy to jewelry that such a reaction can attenuate the normal life span of the joint replacement.

Clinical criteria for “loosening” vary, and although many criteria have been postulated, there is still no general consensus on what constitutes clinically significant loosening.

Factors that may contribute to failure or loosening include wear fatigue of the materials used, debonding between tissue or implant interface planes (which further creates an environment for corrosion), and fatigue damage and wear secondary to radiation techniques used to sterilize implants. Both adhesive forces that literally pull wear particles from surfaces, and abrasive actions that occur between soft and hard elements of the implant may be further contributory factors (24).

In hips, osteolysis-associated loosening usually appears about 4 years after surgery. Sixty percent of lesions progress, with progression occurring slowly at about 0.89 mm/year.

Activation of macrophages by wear debris initiated by subclinically evident micromotion (25) is considered the leading cause of osteolysis (Fig. 17.4). However, wear debris may also be the result of many fundamental mechanisms of wear (26).

In order to bring clarity to the plethora of terms used to describe wear, McKellop et al. have proposed four categories of wear:

In addition, the term corrosion refers to electrochemical processes that can remove or add material and thus also generate damage. Corrosion can act alone or may interact with mechanical wear (26).

As defined by Bauer and Schils (27), adhesive wear results from transient bonding between the articulating surface under load, which may extract particles from the weaker surface. Abrasive wear occurs when surfaces of different hardness create a scratching effect. Fatigue wear occurs when accumulated stresses on the surface of an implant exceed intrinsic fatigue strength. Modes of wear include that between two surfaces or between two surfaces with an interposed third party particle such as bone, cement, metal, or polyethylene (PE).

Corrosion products can be generated from metal ions released from an implant surface, especially in metal-on-metal implants. Wear damage may be grossly visible as scratching, burnishing, delamination, and/or pitting (Fig. 17.5).

Important characteristics of wear include wear rates roughly estimated at 0.14 mm/year for cobalt-chromium and 0.10 mm/year for PE. Wear rates above 0.2 mm/year may be a critical level above which patients are at risk for wear-related complications. Most wear debris is <1 µm in size, often below the detectable limit on transmission light microscopy, with the mean size of a PE particle 0.5 µm. Thus, there are certainly more wear particles than apparent to the routine diagnostic pathologist. Particulate size dictates the pathologic reaction (28). Particles <7 µm are phagocytosable by mononuclear macrophages, larger ones by foreign body giant cells, and those much larger (mean, 67 µm) encapsulated by fibrosis (26). Bioreactivity of particulate debris is also related to composition and concentration. Because it has been estimated that billions of particles are produced during the life of an implant, the sheer scope of possible particle-induced pathology is substantial.

Macrophages engulf particulate wear debris (PE more than metal or cement) from prosthetic materials and release. Proinflammatory factors include:

In addition, the NOD-like receptor protein, NALP3, located in macrophages initiates cleavage of pro-IL1 β into its secreted form, triggering an inflammatory cascade (31). The specific nature of the macrophage response depends on many parameters including the composition, size, shape, volume, and surface area of the particulate debris.

These can mediate the differentiation of osteoclast precursor cells and activate mature osteoclasts. The complicated role played by substances such as IL-6 and interferon-α is underscored by some studies that demonstrate wear debris inhibition of the antiosteoclastogenic signaling of these molecules (32). Recent investigations have focused on the role of blockading the osteoclast activation path by interfering with RANKL (the receptor activator of NF-κB) (33).

RANKL is expressed on the cell surfaces of osteoblasts and bone marrow stromal cells, and it can stimulate the differentiation of osteoclast progenitor cells to form mature osteoclasts. This signaling is mediated through a receptor/activator termed RANK, which is located on the cell membrane of osteoclast precursors and mature osteoclasts (34).

A third molecule, osteoprotegerin, which negatively regulates this process by inhibiting RANK’s capacity to activate RANKL, may reflect a protective effect against osteolysis if shown to be elevated in serum (35).

Wear debris particles may be opsonized by host proteins and substances such as albumin, collagen, and fibronectin. Antibodies may introduce alterations in surface charges or composition, which affect local cellular activity (25). The additional role of bacterial products, such as endotoxin, remains unclear.

Wear debris may also activate osteoblasts, lymphocytes, synoviocytes, and fibroblasts with resultant proinflammatory cytokine activation and activation of metalloproteinases, stromelysin, and collagenase, a virtual alphabet soup of tissue resorptive substances (36). The levels of cathepsins in interface tissue have been measured at higher than that in rheumatoid synovium (37).

There are few, if any, laboratory markers that can indicate or correlate with the osteolysis around implants or the associated increased bone remodeling. Despite the fact that some studies have attempted to correlate loosening with increases in circulating cytokines (such as IL-1 or IL-6) (30,38) or in bone remodeling serum markers such as cross-linked N-terminal telopeptide (NTX) and procollagen I C-terminal extension peptide (PICP), the current diagnosis of aseptic loosening remains clinical and radiologic (39,40).

In summary, osteolysis can be seen as the end result of a number of mechanisms, including:

Whatever the cause of osteolysis, it may be first evidenced by pain or the appearance of radiolucent lines, zones, or regions between the implant or implantable material and the host bone (Fig. 17.6).

In the ideal situation, the introduction of an implant and the subsequent reaming of the bone for implant placement would lead to healing, not unlike that seen in normal fracture healing. The implant would stimulate the differentiation of osteoblasts in adjacent tissue, with eventual osseous anchorage. In actual practice, however, micromotion may lead to fibrosis rather than bone, the final result being the formation of a fibrous or, in some instances, synovial and bursa-like membrane, similar to that seen in pseudoarthroses (Fig. 17.7).

Clinically and roentgenographically, this fibrous membrane may appear as a loose prosthesis with a radiolucent line at the implant-bone interface (Fig. 17.6). However, radiolucency does not necessarily mean clinical loosening. In general, it seems difficult to establish a clear correlation between
radiographic and clinical loosening. Indeed, radiolucent zones between implants and host tissue may be either physiologic or pathologic (41).

The accuracy of comparing different imaging modalities in the detection of osteolysis has been studied extensively. In a cadaveric model, the sensitivity for detecting lesions was 51.7 percent for computed tomography and 95.4 percent for magnetic resonance imaging (MRI) (42).

When a roentgenographic radiolucency is suspected, the histologic correlation is usually that of a fibrous membrane. In a small percentage of cases, radiolucent lines have as a histologic correlation not fibrous membranes but endosteal resorption of cortical bone (43,44).

In some hip-retrieval studies, the dominant correlation of radiolucencies on the acetabular side has been fibrosis and fibrous membranes. However, in the femoral stem, endosteal bone
remodeling predominates, suggesting different mechanisms at work at different anatomical sites (44).

Fornasier et al. (45) evaluated the bone-cement interface in successful asymptomatic total hip replacements retrieved at autopsy, and reviewed the previous understanding of the stages of development of this interface. An initial stage of postoperative tissue necrosis and damage induced by the operative procedure was followed by a second stage in which a fibroblastic ingrowth led to remodeling, repair, and replacement of the former necrotic bone. A final stage, occurring several years after implantation, led to the production of a membrane consisting of acellular, noninflammatory fibroblastic tissue. Where it was present, foreign body reactions to cement particles were thought to decrease with time. The authors concluded that the cellular constituents of the membrane include histiocytes apparently reactive to particulate PE debris, a reaction that increases with time but progresses at different rates in different patients at different sites. The histiocytic response to particulate matter may lead to healthy bone remodeling and osteoclast activation as well as macrophage handling and transportation of particulate matter away from the bone surface. This fibrous membrane, a cellular engine, may occur in both asymptomatic and symptomatic cemented arthroplasties, and both symptomatic and asymptomatic joints. Debated points included the role of micromotion in contributing to the development of the fibrous membrane and the well-known association in some patients of a marked osteolysis, aggressive granulomatous lesions, significant bone lysis, and granulomatous pseudotumors, which clearly represent an aberrant and exaggerated histiocytic and giant

cell reaction to an otherwise possibly harmless reactive process (46,47,48,49). In conclusion, wear debris contributes to a cycle characterized by both histiocytic and macrophage activation and fibrous membrane proliferation.

In one experimental model, cobalt debris was found to be toxic to fibroblasts, but particulate titanium, chromium, and titanium-aluminum alloy actually stimulated fibroblast proliferation (50). The resultant fibrous membrane was postulated to be deleterious, by acting as a conduit for transporting metallic debris and, in turn, stimulating further fibroblast- and macrophage-generated osteolysis, and thus loosening.

In a recent histopathologic study of late aseptic loosening of cemented total hip prostheses, Williams and McQueen (51) described in detail the microanatomy of implant fibrous membranes. Clinically and roentgenographically, the hip prosthesis interfaced with a densely fibrous membrane comprising three zones. The region adjacent to the cement consisted of a pseudosynovial lining, and the area immediately adjacent to that was filled with methyl methacrylate pearls, round-to-oval shaped bodies. A layer of foamy histiocytes was also noted, which ultrastructurally revealed cytoplasm filled with cellular debris and numerous lipid vacuoles. Neuronal bodies were also present, indicating that the histiocytes were phagocytosing cellular debris. In contrast, the histiocytes not in the middle layer were filled with abundant endoplasmic reticulin and protein vacuoles, these latter cells showing dense plaques and contractile filaments similar to those of myofibroblasts, suggesting that the membrane was a metaplastic response stimulated by particulate debris and motion around the loosened prosthesis. The authors concluded that late aseptic loosening is caused by fragmentation of the methyl methacrylate mantle grossly and microscopically. Methyl methacrylate particles then stimulate histiocytes adjacent to them to produce and release lytic enzymes, which in turn cause resorption of adjacent bone (51).

It is now well established that stromal cells including fibrous tissue and other inflammatory cells at the radiolucent zone may be associated with the production of potent osteolytic factors (52,53). Osteolytic factors that have been studied include IL-1α/β and TNFα, IL-6, macrophage-colony stimulating factor (M-CSF), and PGE2. In addition, fibroblasts in periprosthetic membrane tissue are capable of inducing the differentiation of normal human peripheral blood monocytes to mature osteoclasts by a mechanism that involves both RANKL and TNFα (54,55).

Macrophages may also induce bone loss directly by the release of oxide radicals and hydrogen peroxide. Clearly, giant cells and macrophages in other inflammatory conditions, such as rheumatoid arthritis, may lead to directly observed macrophage resorption of tissue, including articular cartilage and bone.

In summary, numerous factors may lead to the clinically observable decline in the mechanical stability of implanted joints after 10 years, particularly in aging and in excessively active patients. These are usually associated with the development of a fibrous membrane, which may or may not contain histiocytes reacting to particulate debris. However, experimental evidence would indicate that macrophages and histiocytes join in the reactive process of metallic wear debris, and may in and of themselves stimulate an ongoing inflammatory reaction in which numerous direct or indirect bone-resorbing and osteolytic materials, including cytokines, PGE2, and collagenase, may be released. This inflammatory reaction may be dramatic, with marked osteolysis and pseudogranulomatous tumors mimicking true tumors (Figs. 17.3A and 17.4B).

A broad range of materials have been used in joint arthroplasties (Fig. 17.8) (Table 17.5). These include most notably:

In the United States in 2012, 93 percent of THA were cementless, with 56 percent of THA bearings being metal highly cross-linked polyethylene (HXLPE) and 39 percent being ceramic-HXPLE. Ninety-nine percent of acetabular cups were modular. In regard to the femoral heads, 61 percent were metal and 39 percent were ceramic, and 51 percent were 36 mm and 28 percent were 32 mm. THA implant usage in the United States tends to favor cementless fixation using modular acetabular cups and large-diameter femoral heads with metal-on-polyethylene or ceramic-on-polyethylene bearings.

Interestingly, the majority of THR femoral stems are cementless in the United States, as opposed to many European countries. In Sweden, with a long-standing tradition of arthroplasty registries to record complications, the majority are cemented.

The stainless steel implant is usually made of an iron-based alloy that contains an array of elements, including iron, chromium, and nickel, often with added substances such as manganese, carbon, molybdenum, and silicone. The cobalt-chromium alloys are similarly complex mixtures of substances. Although the percentage of potentially toxic or even carcinogenic substances within these alloys is small, toxicity has been demonstrated with metals such as chromium and vanadium, and hypersensitivity with cobalt and nickel. One type of titanium contains aluminum, which in excess is known to be associated with several diseases (57).

In addition to various metals, implanted devices in orthopaedic arthroplasty include plastic components such as the PE acetabular cup often used in hip replacements. The traditional fixation device is a PMMA polymer (or bone cement). All these substances have been associated with documented tissue reactions in human implants. Although some of the debrided material may be seen on routine histologic preparations, in most cases the particulate matter in failed prostheses is too small to be detected under light microscopy.

An interesting complication associated with ceramic on ceramic implants is actually audible. “Squeaking” has been most intensely studied, but clicking, snapping, and grinding can also occur (58). Thought to involve 0.5 to 20 percent of ceramic on ceramic implants, a likely cause is edge loading. Squeaking may also be related to impingement, component malposition, separation between liner and shell, and third-body noise.