The resurgence of metal-on-metal (MoM) bearings about 10 years ago was due in part to efforts to avoid the wear-associated complications of metal-on-polyethylene (MoP) bearings. MoM bearings also allowed the use of larger diameter femoral head components, which could theoretically decrease dislocation rates.

MoM bearings gained approval from the U.S. Food and Drug Administration (FDA) through the 510(k) premarket notification pathway that does not require testing prior to marketing of the device. In 2006, 35 percent of U.S. hip arthroplasties used a MoM bearing; approximately 1 million patients received MoM bearings during the past 20 years.

However, after the United Kingdom joint registry reported revision rates of greater than 5 percent at 4 years from surgery and greater than 10 percent by 7 years after surgery, the use of MoM bearings has been controversial. In fact, three MoM bearing systems have been recalled: the Zimmer Durom (in 2008), DePuy ASR (in 2010), and the Smith & Nephew R3 systems (in 2012). The FDA has requested manufacturers to conduct postmarket studies of the devices and required premarket approval for MoM systems beginning in 2013.

Despite the high revision rate with MoM bearings, the pathophysiology has not been fully elucidated. This article discusses the biomechanical and biologic factors in MoM bearing failure.

Biomechanical etiologies
MoM implants were designed to have the midpoint of the femoral head centered within the acetabular cup. The load is carried both through the synovial fluid, which serves as a lubricant, and via contact between the metal bearing surfaces. Over time, proteins in the synovial fluid can create a boundary layer 250 nm to 400 nm thick directly on the metal surface. This fluid-film lubrication mechanism has low wear rates—less than 1 mm3 per million cycles, which is a hundred-fold less than with MoP.

Larger bearings (diameters up to 36 mm) had a lower steady-state wear rate compared to 28-mm bearings. Reports indicate that 85 percent of patients with MoM have low wear rate conditions and acceptable metal ion levels.

However, when this optimal bearing setting does not occur, the wear rates increase dramatically. The wear particles from MoM bearings occur due to a combination of mechanical wear and surface corrosion that results in the release of metal ions and insoluble particles. These released particles range in size from 51 nm to 116 nm (compared to the average size of polyethylene wear debris of 660 nm) and can also result in abrasive wear, worsening the situation.

Edge loading is one of the leading explanations for the breakdown of the fluid film boundary. Edge loading occurs when the acetabular edge experiences high-contact loads, resulting from a variety of biomechanical situations. Rotational malpositioning of the acetabular component, with inappropriate inclination (> 55 degrees or < 30 degrees) or version (> 25 degrees or < 5 degrees), can lead to wear rates of 1 mm3 to 5 mm3 per million cycles (Fig. 1).

A finite-element model analysis of MoM acetabular cups designed to evaluate cup position on wear found inclination angle to be the most important factor; increased version angle was the second most important factor. The highest wear occurred at 55 degrees inclination, and the least wear occurred with an inclination angle of 40 degrees to 45 degrees. Lower inclination angles of 35 degrees were associated with greater wear, but not to the extent of the highest angle. Edge loading in the setting of a poorly positioned acetabular component is exacerbated with the use of a small femoral head.

Malpositioning of the components that lead to instability can also result in edge loading. Medial or superior translation of the acetabular center, offset deficiency, stem subsidence, or head-neck impingement can lead to the femoral head’s sliding inappropriately in and out of the center of the socket and result in high contact stresses at the edge of the acetabulum. This can lead to wear rates of 1 mm3 to 10 mm3 per million cycles.

When the fluid-film boundary layer is disrupted or not allowed to form, wear can occur through abrasion and adhesion mechanisms. This roughens the articulating surfaces and further aggravates wear.

The trunnion is also an important source of metal wear products. The trunnion is the cylindrical device that sits angled on the femoral component to hold the prosthetic head. With large head MoM bearings, friction at the bearing surface acts through a longer moment arm to create greater torque on the Morse taper head/neck junction. Furthermore, if the Morse taper junction is not located at the center of the hip ball, the larger head can result in the joint reaction force acting through a larger moment arm to produce additional torque on the Morse taper head/neck junction.

To compound matters, some implants evolved to use smaller and shorter trunnions and thus had less surface area to distribute the loads being transferred at the Morse taper junction. All of these mechanisms can contribute to greater stresses at the Morse taper junction. This can lead to motion-induced crevice corrosion and fretting corrosion as well as wear at the Morse taper.

Implant retrievals have found that corrosion and wear are generally greater on the head side of the trunnion interface and can range from 0.6 mm3 to 4.9 mm3 per year. The risk of wear at the trunnion is greater with threaded and with shorter trunnions. Registry data show that larger femoral heads have been associated with higher revision rates for total hip arthroplasty (THA). However, for hip resurfacing, larger femoral heads had a smaller rate of revision. This difference may be due to the presence of trunnion wear with THA.

Other sources of wear include acetabular loosening, microseparation, patient activities, and gait patterns. The acetabular component may loosen due to the greater frictional torque with larger femoral heads and with edge loading. The loose component can act as a generator for wear products. Microseparation of the acetabular component from the pelvis may lead to initial instability and subsequent wear generation. Some studies suggest that complications that develop a little over a year after surgery may be associated with microseparation, while failures from other causes may begin to occur 2 or more years after surgery.

Females have been reported to have higher complication rates than males. One study found that females tended to have factors associated with greater wear, including acetabular components with increased inclination and version, as well as a smaller joint. The increased femoral anteversion in females can lead to posterior impingement and microseparation.

Biologic etiologies
The metal wear products that are produced from MoM implants can be deposited locally or spread systemically. The chromium ions form a stable oxide, chromium orthophosphate, that has been found concentrated in the periprosthetic region. On the other hand, cobalt does not form a stable oxide in physiologic conditions, remains as an ion that tends to spread, and leads to much higher levels in the blood than chromium. In vitro studies suggest that toxic effects are mainly mediated by cobalt ions.

An adverse tissue reaction to the metal can develop when the patient’s immune system is stimulated by complexes of metal ions bound to plasma proteins. These reactions can lead to the formation of pseudotumors, cystic and solid masses around the prosthesis. Pseudotumors have been reported to occur in 1 percent of patients within 5 years of a MoM resurfacing (Fig. 2).

This response is classified as a delayed-type hypersensitivity reaction, aseptic lymphocyte-dominated vasculitis-associated lesion (ALVAL). Pseudotumors are more frequently seen in females than in males. In one series, patients with a metal reaction had normal white blood counts, C-reactive protein (CRP), and erythrocyte sedimentation rates (ESR). However, other studies report elevated levels of CRP and ESR, making ALVAL more difficult to distinguish from infection. Symptomatic patients tend to have smaller bearing surfaces and malposition of the acetabular component.

Another impact of metal wear can be metallosis, which is the fibrosis, necrosis, and/or prosthesis loosening due to metal debris. Intraoperatively, the periprosthetic tissue has a gross dark or cloudy staining (Fig. 3). Hip resurfacing patients have a metallosis rate of 2 percent to 5 percent. Revisions in patients with metallosis have higher complication rates, due to tissue damage resulting in a severely weakened capsule, abductor muscles, and bone. Metallosis is associated with elevated metal ion levels. However, the presence of metallosis is not associated with ALVAL.

Other biologic effects of metal wear products include inflammatory cytokine release by macrophages, histiocytosis, fibrosis, and necrosis. Bony ingrowth into arthroplasty components can be hindered as well. Metal ions can damage DNA and reduce tumor suppressive CD8 T-cells, but studies have not found increased cancer rates with MoM implants.

High levels of metal ions have been shown to be bacteriostatic and decrease biofilm formation. On the other hand, they hinder neutrophil function, and biofilms can develop on wear products. Further research is needed to determine if metal wear products affect infection rates. The metal wear products can spread systemically and have been found in liver and spleen macrophages. The impact of this spread and accumulation is not known.

Impact on different patient populations
Retrieval studies have shown that patients with both high and low implant wear levels have needed revision. The patients with high wear generally had elevated serum metal ion levels and often had malpositioned acetabular components (frequently an increased inclination). These patients generally comprise more than half of revisions.

However, 33 percent to 45 percent of revisions have been found to have components with acceptable position and a low wear rate. These patients also have an acceptable level of metal ions in the blood. This second group is thought to have a metal hypersensitivity resulting in a greater chance of failure despite lower metal ion levels.

Patients with pseudotumors tend to have greater rates of wear on both the femoral and acetabular components. In particular, the wear on the acetabulum tends to be located at the edge of the implant surface, suggesting edge-loading. Nonetheless, pseudotumors have developed in both high- and low-wear patient populations.

Conclusion
The resurgence in MoM popularity was due to the hope for reduced bearing surface wear and improved hip stability. However, in practice, higher-than-expected wear and corrosion has led to a concerning rate of adverse tissue reactions in patients with MoM bearings.

Up to half the patients who need a revision of a MoM bearing surface do not have significant wear and may represent a population of patients more sensitive to metal wear products. An improved understanding of MoM failure mechanisms can improve both follow-up and management and possibly suggest new treatments for the more than 1 million patients who have received a MoM bearing surface arthroplasty.

Links to additional information and references for the studies cited in this article can be found in the online version, available at www.aaosnow.org

Karthikeyan E. Ponnusamy, MD, and Nicholas J. Giori, MD, are members of the AAOS Biomedical Engineering Committee.

References:

1. Hosman AH, van der Mei HC, Bulstra SK, Busscher HJ, Neut D. Effects of metal-on-metal wear on the host immune system and infection in hip arthroplasty. Acta Orthop. 2010;81(5):526–534.
2. Hart AJ, Matthies A, Henckel J, Ilo K, Skinner J, Noble PC. Understanding why metal-on-metal hip arthroplasties fail: A comparison between patients with well-functioning and revised birmingham hip resurfacing arthroplasties. AAOS exhibit selection. J Bone Joint Surg Am. 2012;94(4):e22
3. National Joint Registry. MHRA medical device alerts for metal-on-metal hip implants. Updated 2012. Accessed April 16, 2014
4. Food and Drug Administration. Metal-on-metal hip implants recalls. Updated 2013. Accessed April 16, 2014
5. Fisher J. Bioengineering reasons for the failure of metal-on-metal hip prostheses: An engineer's perspective. J Bone Joint Surg Br. 2011;93(8):1001–1004.
6. Kwon YM, Glyn-Jones S, Simpson DJ, et al. Analysis of wear of retrieved metal-on-metal hip resurfacing implants revised due to pseudotumours. J Bone Joint Surg Br. 2010;92(3):356–361.
7. Hart AJ, Muirhead-Allwood S, Porter M, et al. Which factors determine the wear rate of large-diameter metal-on-metal hip replacements? multivariate analysis of two hundred and seventy-six components. J Bone Joint Surg Am. 2013;95(8):678–685.
8. Clarke SG, Phillips AT, Bull AM, Cobb JP. A hierarchy of computationally derived surgical and patient influences on metal on metal press-fit acetabular cup failure. J Biomech. 2012;45(9):1698–1704.
9. Bishop N, Witt F, Pourzal R, et al. Wear patterns of taper connections in retrieved large diameter metal-on-metal bearings. J Orthop Res. 2013;31(7):1116–112.
10. Bozic KJ, Browne J, Dangles CJ, et al. Modern metal-on-metal hip implants. J Am Acad Orthop Surg. 2012;20(6):402–406.
11. Langton DJ, Jameson SS, Joyce TJ, Hallab NJ, Natu S, Nargol AV. Early failure of metal-on-metal bearings in hip resurfacing and large-diameter total hip replacement: A consequence of excess wear. J Bone Joint Surg Br. 2010;92(1):38–46.
12. Pelt CE, Erickson J, Clarke I, Donaldson T, Layfield L, Peters CL. Histologic, serologic, and tribologic findings in failed metal-on-metal total hip arthroplasty: AAOS exhibit selection. J Bone Joint Surg Am. 2013;95(21):e163.
13. Chang EY, McAnally JL, Van Horne JR, et al. Metal-on-metal total hip arthroplasty: Do symptoms correlate with MR imaging findings? Radiology. 2012;265(3):848–857.