Aseptic loosening as a result of particulate induced osteolysis is the most common cause of implant failure after total hip arthroplasty.1 (Figure 1) The process of aseptic loosening occurs as a result of bone loss that ultimately effects the integrity of the host-implant bond. There are at least three possible causes for this bone loss: 1) osteolysis (particulate induced bone resorption); 2) adaptive bone remodeling/ stress shielding due to material and structural properties of the prosthesis; and 3) bone loss due to aging (osteoporosis).2 Such bone loss may be stable or progressive over time and may or may not result in clinical instability of the prosthesis. In the early stages, osteolysis is generally asymptomatic and treatment is determined by the stability of the prosthesis and the nature of the lesion (i.e., progressive or non-progressive).3 Attempts to limit osteolysis have focused on the design of the prosthesis, the insertion technique, the search for alternative bearing materials that produce fewer wear particles, the use of smaller heads in order to decrease volumetric wear, and the development of improved cement techniques intended to limit migration of particles. A more recent approach has focused on understanding and manipulating osteolysis at a molecular level through pharmacological agents and other strategies.

Figure 1: Osteolysis is the most common cause of implant failure after total hip arthroplasty. Anteroposterior (A) and lateral (B) radiographs demonstrate osteolysis around both the femoral and acetabular implants (arrows).		Figure 2: This polarized light micrograph of the pseudomembrane-bone interface demonstrates bone resorption and an intracellular UHMWPE particle (arrow 3).

Background

         When surgeons first began identifying bone loss around total hip prostheses, it wasnÕt clear if there was an infectious or neoplastic process at work.4 By 1977, Dr. Willert5 and Dr. Harris(4) provided important early insights regarding the role of particles and macrophages in this process. A more detailed understanding of the basic cellular mechanisms responsible for this bone resorption caused by a particulate induced inflammatory responseÑnow termed periprosthetic osteolysisÑwas made by Goldring and colleagues at Massachusetts General Hospital in 1983.6 A number of research groups have taken an interest in the molecular biology of this process and some novel means of inhibiting the process have been suggested.

         The particulate induced foreign body reaction of osteolysis results in the generation of a pseudomembrane composed of granulomatous tissues including macrophages, fibroblasts, giant cells, and osteoclasts.6 (Figure 2) The extent of this response is driven by the number, size, composition, surface area, and types of particles present.7, 8 These particles include polymethylmethacrylate (PMMA), ultra high molecular weight polyethylene (UHMWPE), titanium, cobalt chrome, and ceramic debris. Although there are differences in the relative local toxicity of each of these particles, the end result is the same. Each class of particles is capable of stimulating macrophages to coordinate a foreign body response that ultimately results in osteolysis.

         Although the biology of aseptic loosening is the same after total hip (THA) and total knee arthroplasty (TKA),9, 10 osteolysis is much less common at the knee. Observations such as this illustrate the importance of differences at the molecular level of the process. Recent research has shown that the larger particles associated with TKA (10-100 microns) generate membranes that have less biologic activity than membranes generated from the smaller particles found after THA (< 5 microns). These differences in biologic activity have been quantified using in vitro techniques to measure the production of mediators known to be important in the process of bone resorption including stromolysin, prostaglandin E2, interleukin 1a (IL1-a), interleukin 1b (IL-1-b), and Tumor Necrosis Factor alpha (TNF-a).11

         Regardless of particle type, the periprosthetic membrane contains similar numbers of macrophages, fibroblasts, giant cells, and osteoclasts. These cells are capable of secreting a variety of cytokines and proteases that modulate bone resorption.6, 9, 11-13 Macrophages in particular are abundant in the pseudomembrane and are capable of differentiating into mature osteoclasts.9, 10, 14, 15 Macrophages secrete a number of cytokines that activate resident osteoclasts to resorb bone12, 16, but they are capable of limited independent bone resorption.14 Thus, therapeutic modalities aimed at inhibiting osteoclastic bone resorption offer the most direct means of inhibiting the process of osteolysis. Several steps are required for the debris stimulated osteoclasts to resorb bone. Each of these steps represents a potential point at which the osteolytic process could be inhibited.16, 17
(Table 1)

Inhibition of Osteolysis by Bisphosphonates

         The bisphosphonates are a class of medication that directly inhibits osteoclast bone resorbtion. Bisphosphonates bind to the mineral portion of the bony substrate and are released locally in high concentrations as osteoclasts resorb bone. The bisphosphonate then acts to inhibit the osteoclast via an unknown and possibly diverse set of cellular metabolic reactions.18

         One of the newer bisphosphonates, alendronate(Merck; Rathaway, NJ), was the first medication shown to inhibit osteolysis in vivo. Drs. Harry E. Rubash and Arun ShanbhagÑnow part of the MGH Adult Reconstruction Unit--were presented the John Charnley Award by the Hip Society in 1997 for their study demonstrating that alendronate significantly reduced osteolysis in a canine model.12 Their group performed uncemented total hip arthroplasty in adult dogs and introduced particulate debris containing ultra high molecular weight polyethylene particles of cobalt chrome and titanium alloy in a reproducible fashion at the time of surgery. Three groups were studied: group I were controls into which no particulate debris was added; group II received particulate debris; and group III received particulate debris and were treated with oral alendronate therapy at a dosage of five milligrams per day. The results were impressive. The rates of loosening were markedly decreased in alendonate treated animals. Although periprosthetic membranes producing inflammatory cytokines were generated in group III animals, little bone resorption or loosening was demonstrated.12 Thus, alendronate had little effect on the animalsÕ inflammatory response to the particulate debris, but blocked the particulate induced osteolysis presumably via a direct inhibition of the osteoclast. This study demonstrated the potential for novel interventions tailored to the molecular biology of a disease to transform the management of the disease.

The Search for Other Inhibitors of Osteolysis

         The Orthopaedic Biomechanics Laboratory at the Massachusetts General Hospital was recently awarded grants from The National Orthopaedic Fellows Foundation and the Orthopaedic Research and Education Foundation (OREF) to fund the development of an in vitro model to test clinically relevant osteoclast inhibitors. We believe that there are several means of inhibiting osteoclast-mediated bone resorption (Table 1). The development of an in vitro model of osteolysis will facilitate testing of a series of pharmacological agents capable of inhibiting osteolysis at various steps.

         Periprosthetic membranes will be obtained from patients at the time of revision total hip arthroplasty and prepared for study using organ culture techniques. Potential osteoclast inhibitors to be tested include several bisphosphonates, proton pump inhibitors, carbonic anhydrase inhibitors, and prostaglandin synthesis inhibitors. Bone resorption will be quantified by measuring the elution of radioactive calcium and N-Telopeptide from radiolabeled cortical bone placed in the tissue cultures; by testing the supernatant from each tissue culture using and enzyme-linked immunosorbent assays (ELISA) for PGE-2, IL1-a, TNF-a, and TGF-b; and by immunoflourescent histological techniques for localizing and quantifying osteoclasts using antibodies against osteoclast-specific cell surface proteins such as calcitonin receptors and tartrate resistant acid phosphatase (TRAP).

          The agents with the greatest percentage inhibition in vitro will then be studied using an in vivo model of particulate induced osteolysis. In vivo models have been established by researchers at the Brigham and WomenÕs Hospital19 among others.12 Ultimately, the agents with the greatest documented inhibition in vivo will be candidates for randomized trials in humans. ItÕs possible that some combination of agents acting synergistically via separate pathways may provide the greatest degree of inhibition.

Conclusion

         The efforts of our laboratory and others to better understand and manipulate the process of osteolysis on a molecular basis will hopefully generate effective therapeutic options for the management of osteolysis. Ultimately, our goal is to provide an improved clinical outcome for patients with total hip and knee arthroplasties.

Richard Illgen II, MD is a Resident in the Harvard Combined Orthopaedic Residency Program

Harry E. Rubash, MD is Chief of the Department of Orthopaedic Surgery at Massachusetts General Hospital and Professor of Orthopaedic Surgery at Harvard Medical School

Arun S. Shanbhag, PhD is Assistant Professor of Orthopaedic Surgery at Harvard Medical School

William H. Harris, MD is Chief of the Adult Reconstructive Unit at Massachusetts General Hospital, and Allen Gerry Clinical Professor of Orthopaedic Surgery At Harvard Medical School

Address correspondence to:
William H. Harris, MD; Massachusetts General Hospital; Jackson 1126; 55 Fruit St.; Boston, MA 02114