FDA approval of metal-on-metal (MOM: 28, 32mm) bearings has provided 10 years of clinical experiences in USA. However there has been no detailed mapping of wear phenomena in retrieval cases. We present an analysis of 28 cases, MOM retrievals with 1 to 10 years follow-up, radiographic reviews and metal ion studies. Ball diameters ranged from 28mm to 42mm. Two balls were the early design with skirts. Main indicators for revision were the progressive radiographic changes indicative of osteolysis, with associated hip pain. Approximately 54% of patients were males and ages ranged from 36 to 76 years of age. Only 7 femoral stems were recovered but all had impingement marks. Only three cases lacked any evidence of stripe wear and these were in very elderly patients. Approximately 85% of these cases showed some evidence of stripe wear and multiple stripes were clearly visible on 50% of the femoral balls. The medial ball stripes were twice as common as lateral. Stripe wear was identified in 25% of CoCr liners.

In the hip simulator studies generally show ‘run-in’ wear rates of 1–7mm3 per million cycles (Mc). We noted that above the 5mm3/Mc threshold, the serum generally appeared black. In contrast, the ‘steady-state’ wear rates of 0.1–1.6 mm3/Mc showed the true potential of MOM bearings. However there were often examples of higher wear (7–20 mm3/Mc), which gave confounding trends in published studies. Our studies of metal ions in the simulator lubricant provided a very accurate representation of MOM wear.

There are many limitations in comparing in-vitro to in-vivo wear performance. Our retrieval data are biased to cases that failed due to hip pain, had radiographic signs of progressive osteolysis and some showed high levels of metal ions. There was also the bias of having predominantly a CoCr sandwich design (polyethylene adaptor). Use of the small ball added the well-known risks of impingement, subluxation and dislocation with rigid cups. Using the ‘damage modes’ from McKellop, we found only normal Mode-1 wear to be rare in these cases, whereas Modes# 2–4 had an incidence approaching 30% each. Signs of impingement were evident in 85% of our cases. Thus summarizing these MOM wear phenomena in retrieved 28mm sandwich cups, the evidence implicated impingement and 3rd-body wear modes (#2–4) as the clinical risk for adverse wear effects at 10 years follow-up. The in-vitro wear studies have not yet simulated such adverse clinical effects.

From 1985 metal-on-metal (MOM) designs of resurfacing (RSA) and total hip arthroplasties (THR) have been available over a large diameter range (28–60mm). In-vitro studies indicated satisfactory low wear performance for all designs and diameters tested (wear = 0.1 to 7 mm3). While reports from many centers have been encouraging, some have reported adverse effects. We reviewed clinical and metal ion studies in large diameter retrievals and compared these to 28mm MOM cases. Patients with the latter THR ranged 36–76 years of age and were followed 9–11 years. Main finding in our revisions was osteolysis and pain. The 28mm ball was represented 86% of cases; 71% balls had stripe wear. For liners, 25% had circumferential stripe wear and impingement was evident in 64% cases. Seven cemented stems were recovered with impingement marks; 26 stems were undamaged and therefore not revised. Using the concept of ‘damage modes’ from McKellop, normal wear mode #1 was evident in only 14% of 28mm retrievals whereas incidence of ‘abnormal’ modes #2-4 approached 30% each. Thus the 28mm MOM appeared susceptible to impingement risks with CoCr liners. Summarizing MOM retrievals, damage modes 2–4 were most likely implicated in revisions. The performance of such ‘small diameter’ THRs will be contrasted to our large diameter THR and RSA experience. The questions to be reviewed include, how much of the reported MOM adversity was predictable and how much risk was due to

wear of small diameter MOM,