Hip resurfacing: expectations and limitations

The first hip resurfacing arthroplasty (HRA) was performed by Smith-Petersen (1939) in 1923. He temporarily interposed a thin hemispherical shell between the femoral head and acetabulum in patients with arthrosis. He expected new cartilage to grow on the articular surfaces and called his technique two-stage mold arthroplasty. Originally, the shells were made of glass, bakelite, or celluloid. In 1938, he changed to metal shells made of a recently developed cobalt-chrome-molybdenum alloy. He discovered that it was not cartilage that grew beneath the shell but fibrous tissue. He also discovered that leaving the shell in situ gave better clinical results than removing the shell. These shells were used up until the late 1950s, by which time the first intramedullary prosthetic stems had entered the market. (See also Northover and Maqsood (2008) in this issue of Acta).

In the 1970s, the second generation of resurfacings were developed. They were total resurfacings with a metal-on-polyethylene bearing. The most well-known design was the Wagner double cup. It consisted of a cemented metal or ceramic femoral shell and a cemented all-polyethylene acetabular component. The clinical results were disappointing. Howie et al. (1990) reported only 40% survival at 8-year follow-up for the Wagner double cup and Amstutz et al. (1986) found 53% survival at 7 years for the THARIES resurfacing prosthesis. These implants failed from osteolysis and aseptic loosening. The cause of failure was thought to be either polyethylene wear, stress shielding, or avascular necrosis. In retrospect, these failures can be mainly attributed to cytokine-mediated osteolysis induced by wear particles from the thin-walled polyethylene acetabular component (Harris 1994).

Recognizing this problem, McMinn and Treacy in Birmingham in the 1990s developed the third generation of resurfacings with a metal-on-metal bearing. After clinical trial and error, the best fixation was obtained with a cemented, stemmed femoral hemispherical shell and a conventional HA-coated Co-Cr-Mo cup. In 1994, implantation of this McMinn prosthesis was started, to be followed in 1996 by the Conserve Plus designed by Amstutz. When favorable clinical results were reported (Daniel et al. 2004), many other companies followed with designs that all had the same basic features as the McMinn prosthesis. By now, at least 15 types of resurfacing prosthesis are available, and the number is still increasing. These prostheses differ in many details, such as shape, sizing, head coverage, clearance, metal alloy used, heat treatment, instrumentation, and so on. Just as with conventional hips, they exhibit differences in clinical outcome. The Swedish and Australian registries have listed a two- to threefold difference in revision rate between different makes (Swedish Hip Arthroplasty Register 2006, Australian Orthopaedic Association 2007).

Advocates of resurfacing claim that there are a number of advantages of HRA over conventional total hip arthroplasty (THA), e.g. less bone removal, restoration of anatomical hip center, physiological loading, higher postoperative activity levels, excellent survival rates, and ease of revision. These advantages appeal strongly to the imagination of both surgeons and patients. Over a short period of time, resurfacings became very popular and the number of implantations rose to about 10% of all primary hip replacements in countries such as the UK, Australia, and the Netherlands (National Joint Registry for England and Wales 2006, Australian Orthopaedic Association 2007). In Sweden, where the choice of implant is based more on clinical evidence from the hip register, the number of resurfacings is only 1% (Swedish Hip Arthroplasty Register 2006). In 2004, the designers of the McMinn and BHR resurfacing prosthesis published excellent medium-term results (Daniel et al. 2004) with 1 revision in 446 resurfacings with a mean follow-up of 3 (0–8) years. Publications from designers, however, often present results that are better than those from multicenter studies. Survival rates in multicenter studies have shown worse results for HRA than for THA. The National Joint Registry for England and Wales (2006) has reported revision rates of 2.4% for HRA and 0.9% for cemented THA at 3 years; the Australian Orthopaedic Association (2007) has reported 4.4% revision for HRA and 3.1% revision for cemented THA at 5 years; and the Oswestry registry has reported a revision rate of 4.6% for HRA at 7 years (Kahn et al. 2008). In the Australian registry, the revision risk ratio, adjusted for age and sex, for HRA and THA is 1.4. Main reasons for revision are neck fractures (occurring in 1–2% of resurfacings), loosening, and pain.