Osseointegration is defined as the direct contact of bone with the surface of the implant material (3). This bone–implant contact helps to stabilize the implant within the host bone and can prevent micromotion, or microscale movement of the implant relative to the host bone, which is thought to play an important role in the eventual failure of joint arthroplasties (15). Using design parameters and biologic treatments, scientists have been able to increase the bone formation which stabilizes implants (16,17). In the clinical setting, the most commonly used strategies to enhance osseointegration are to use an implant that is inherently stable mechanically with good initial contact between the implant and host bone and to use surface treatments favorable to extensive bone–implant ingrowth (16).

Bone remodeling adjacent to the implant begins almost immediately after surgery and overlaps chronologically with the injury response. In fact, the initial injury response is critical to establishing osseointegration (16). Thus, especially in cementless implants, the ability to obtain a strong mechanical bond between the implant and host bone depends upon the formation of woven bone at the interface, which is subsequently replaced by lamellar bone. Although cemented implants have their peak integration strength at the time of implant placement, it takes several weeks for implant fixation strength to fully develop in cementless implants and this process depends upon the ability of the body to form new bone and to form new bone and grow into or onto the implant surface (16).

The surgical procedure itself, beyond creating a stimulus for bone repair at the interface, also affects the surrounding host bone (18,19,20). A recent study showed that techniques used in femoral canal preparation affect the degree of porosity developed in the cortical bone, leading to an early loss in cortical bone mass (20). For instance, reaming increased porosity in a canine model by six-fold 6 weeks after surgery. The remodeling stimulus was presumably related to an interruption of the intramedullary blood supply, although heat generation may also contribute. A recent study from our laboratory showed that elevations in peri-implant trabecular and endocortical bone remodeling were transient in an unloaded rat intramedullary implant model (21). Thus, it is clear that canal preparation and implant placement trigger a remodeling response. It is not yet clear if these changes have downstream effects by directly or indirectly influencing later events such as osteolysis or stress shielding. It is possible that these remodeling transients may affect longitudinal changes in peri-implant bone mass and partially explain reported periods of initial bone loss, followed by recovery, often times followed in turn by longer-term bone loss (e.g., see (22,23)).

Stress shielding refers to the reduction in mechanical stimulus to peri-implant bone because the implant transfers the joint reaction force rather than the host bone (Fig. 48.2). Femoral hip stems are loaded in bending and therefore the bending stiffness of the implant relative to the bone is a key factor (24). Bending stiffness depends upon the modulus of elasticity of the material (E) and the area moment of inertia (I) of the structure. Femoral stems are made from materials that have an elastic modulus that is roughly five times (titanium and its alloys) or 10 times (cobalt-chrome) greater than the elastic modulus of cortical bone. Even though the bone may have a larger moment of inertia, the bending stiffness (E × I) of the implant is large enough, relative to the bone, to lead to a significant part of the load being transmitted by the implant (25,26). According to Wolff’s Law, bone adapts its shape to best accommodate the mechanical stresses experienced by the tissue (27). Thus, changes in the bone’s mechanical environment because of stress shielding have been correlated with the pattern of bone loss and bone gain following THA, much like the reduction in mechanical stress caused by weightlessness in astronauts causes bone loss (28). In the context of THA, bone
loss is particularly noticeable in regions that experience the greatest reduction in mechanical stimulation such as the calcar region of the proximal femur. Various design parameters such as the relative bending stiffness of the implant compared to the host bone are determinants of the severity of stress shielding and the amount of subsequent bone loss. Although stress shielding is thought to primarily affect cementless implant designs, cemented implants can also be affected (29). Most attention has been paid to loss of bone mass, but there is evidence that bone shape (30) and bone quality (31) can also be affected. In addition to stress shielding, other factors can contribute to the altered mechanical environment in the peri-implant tissue following THA, including altered loading caused by changes in gait (32), subject activity level (33), and the use of muscle splitting as opposed to muscle-sparing surgical approaches (34).

Particle-induced osteolysis is widely accepted as a key factor in aseptic loosening (35,36,37), although lack of initial mechanical stability and other factors may also play a role (15,38,39). Wear debris as well as other particles and ions released from implants are thought to invoke an immune system response through toll-like receptors 2 and 4 and the NALP3 inflammasome (40,41,42). Endotoxins (such as lipopolysaccharide) can potentiate the ability of particles to induce an inflammatory response (43). The potential role of adaptive immunity is not clear (44,45). Collectively, the inflammatory responses increase osteoclastogenesis and bone resorption and may also induce osteoblast apoptosis (42). In addition to these inflammatory pathways, particles can directly downregulate collagen type I production by osteoblasts (46), negatively affect osteogenic lineage cell proliferation and differentiation (47), and can induce a catabolic phenotype in osteoblasts and osteocytes (48,49). More detailed consideration of the cascade of events that occur during particle-induced osteolysis is beyond the scope of this chapter, but can be found in several review papers (42,50,51,52) as well as in Chapter 21 in this book.

Osteolysis, stress shielding, and osseointegration are all associated with changes in the bone remodeling process that occurs following THA and probably do not act independently (Fig. 48.1). Stress shielding, for instance, has been theorized to facilitate particle-induced osteolysis (53,54,55), by opening the bone–implant interface, thereby allowing for migration of particles within the intramedullarly canal to trigger osteolysis and peri-implant remodeling (56,57). Therefore, as the formation of a strong bone–implant interface can limit the migration of particles along the implant surface, maintenance of the interface through osseointegration is important to prevent failure of implant fixation (58,59,60,61). Perhaps, equally important is maintenance of peri-implant bone mass by reducing stress shielding and the accompanying loss of bone.

Radiography is the mostly commonly used method of evaluating the skeleton noninvasively and the technique has been commonly used to monitor changes in the bone following THA (62,63). Radiographs are simple to perform, inexpensive and relatively safe to the patient, and therefore are routinely used today. Radiographic evaluation is primarily qualitative and large amounts of bone must be lost before the presence of depressed bone mass can be inferred, therefore making a consensus definition of implant failure difficult (64,65). Radiographic evaluation of THA is a two-dimensional technique in which regions of lost bone are identified by the presence of relatively radiolucent areas, which in the extreme can indicate the presence of osteolytic lesions, that is, areas largely devoid of bone surrounding the implant (Fig. 48.3).

Dual-energy X-ray absorptiometry (DEXA) is used to investigate bone mineral density (BMD) in patients noninvasively. BMD is actually the bone mineral content (BMC) divided by the projected area of interest. Thus, this is not a true density measurement as the units are g/cm². DEXA scans are the gold standard for the diagnosis of osteoporosis and low BMD scores have been shown to be positively correlated with an increased risk of fracture (66,67). The use of DEXA to measure periprosthetic BMD has been validated as an accurate and reproducible technique (68,69). Seven periprosthetic “Gruen zones” were first defined in a radiographic study of the bone–implant interface to provide anatomic context (62) and these zones have become the standard for reporting DEXA results (Fig. 48.4). Following THA, BMD is not lost homogeneously, and therefore, the Gruen zones are useful to provide an anatomically based coordinate reference system for the femur. The Gruen zones most commonly associated with stress shielding are the proximal most zones, particularly zone 7, or the calcar region and zone 1, which encompasses the greater trochanter. These regions have been shown to be the primary sites of net bone loss following implant placement and simulation studies have validated the mechanical unloading, or stress shielding, of these regions, as detailed below. In THA patients, decreased BMD can be caused by both stress shielding and osteolysis, but nearly all DEXA studies list the presence of osteolytic lesions as an exclusion criterion, so DEXA is almost always used to examine adaptive remodeling.

Prospective DEXA studies in which periprosthetic bone loss was directly inferred by comparing early postoperative scans to subsequent scans have shown that most of the bone loss occurs within the first 12 months, followed by a leveling off for 5 to 6 years, with evidence of continued bone loss in
subsequent years in Gruen zone 7 (Fig. 48.5) (22,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96). Although the early impression was that measurable bone loss only occurred within the first 12 months (76,84,97) and was stable after 1 year (87,92,98), it is now clear that longer-term periprosthetic bone loss does occur. For instance, a 14-year longitudinal study (95) showed that BMD in Gruen zone 7 appeared to be stable from 6 through 24 months and then linearly decreased over the next 150 months. Another long-term study has shown rapid bone loss in the first year, with relative stability through 6 years, followed by continued bone loss up to 10 years (93). Between 12 and 17 years following cementless THA, a recent study has shown slight loss in Gruen zone 7 in men and losses in Gruen zones 1, 4, 6, and 7 in women (99). In some cases, BMD has been reported to increase after an early decline (75). Studies of bone changes adjacent to the acetabular component by DEXA are relatively rare along with reports of considerable bone loss (89) and small or no changes (100).

DEXA measurements have limitations, as the technique is highly dependent on the placement of the patients within the scanner. Slight changes in the hip angle can have a dramatic effect on the resulting BMD outcome (68). In addition, BMD can be highly variable between individuals within a population, and therefore it has become a standard practice to use early postoperative BMD as the baseline for accurate prospective determination of implant effects (92). Because BMD is the ratio of BMC to the projected bone area, in cases where BMD has changed it is always wise to examine the BMC and area measurements, as well, particularly because the area measurement tends to change in Gruen zone 7 through “rounding off” of the calcar. Unfortunately, these data are rarely reported. Finally, having DEXA measurements from elsewhere in the body (e.g., spine or even the femur, well distal to the implant) are helpful in interpreting the BMD data from the bone adjacent to the implant and its relationship to age-related bone loss, but again, are rarely included in the literature.

There is limited experience with measuring bone mass around implants with computed tomography (CT). One of the primary concerns is the metal-induced artifact which occurs in the images (101) Nevertheless, there are a few reports showing changes in bone density in the vicinity of implants (102,103).

Bone turnover and perfusion have been examined through imaging of the bone-seeking radiotracers, ⁹⁹mTc-labeled diphosphonates and ¹⁸F-NaF, for diagnosis of various skeletal pathologies (104) and may prove to be a valuable research tool for studying peri-implant bone remodeling. Recently, the combination of PET with CT has shown that remodeling activity can be anatomically localized through ¹⁸F-NaF imaging. For instance, studies using this imaging modality have suggested that bone metabolism returned to normal levels adjacent to cementless acetabular components 1 year after surgery (105), while activity adjacent to the femoral component tended to be higher at 4 months post surgery than at either the immediate postsurgical period or at 1 year (106). ¹⁸F-NaF PET imaging combined with DEXA has been used to infer that bone resorption led to continued loss of peri-implant bone mass even in the presence of elevated bone formation following revision THA (107). These methods have also been used to discriminate between aseptic and septic loosening (108).

Bone remodeling is a cell-mediated process and monitoring the biochemical activity of the cells involved in this process presents an opportunity to assess the dynamic status of periprosthetic bone. The concentration of biochemical markers in body fluids has been shown to change as early as 3 to 6 months postoperatively, while radiographic evidence of bone loss can take up to 36 months to become apparent (109). The ability to diagnose excessive remodeling activity prior to the loss of considerable bone mass is particularly attractive as it may help to identify patients in need of pharmacologic intervention (see below) or allow for early surgical intervention. Therefore, the use of biochemical markers presents an opportunity for improved diagnostic evaluation of periprosthetic bone remodeling and subsequent implant loosening (110). The studies that have evaluated the concentration of biomarkers commonly do so by determining the utility of the marker of interest for the diagnosis of implant loosening and as such the biomarker concentrations are often correlated to radiographic evidence of osteolysis (Table 48.1).

Overall, biomarkers of bone formation, such as osteocalcin (OC), bone-specific alkaline phosphatase (b-ALP), and collagen propeptides, PINP and PICP, have had limited utility in predicting implant failure (Table 48.1). Although, the concentrations of these markers tend to change following THA, the direction, magnitude and significance of the changes are variable. Biomarkers of osteoclast-mediated bone resorption including NTX (74,111,112,113) and TRAP 5b (110,114,115), in particular, have successfully demonstrated correlation with clinical loosening.

Longitudinal studies of biomarkers have demonstrated the transient nature of the cellular activity following THA (109,118,119,120). In general, resorption markers tend to peak shortly after implant placement, and formation markers seem to initially decrease with increasing concentration developing thereafter. Therefore, it is important that the time of sampling be chosen to best monitor the biomarker of interest. Biomarker concentrations have also been shown to have high interindividual variations related to age, gender, disease, and menopausal status. Therefore, it is important to determine baseline marker concentrations prior to implant surgery (121,122,123). In addition, preoperative biomarker concentrations themselves may have predictive value for the determination of patients at risk to develop loosening (74,124).

Implants that were retrieved because of failure at the time of revision surgery (129

Stay updated, free articles. Join our Telegram channel

Join