Ph Hernigou (Créteil)
Although the link between numerous host-related and environmental factors and post-operative joint infection is better understood, the link between the use of different prosthetic biomaterials and post-operative joint infection has not been clearly defined. Post-operative joint infection is caused by the attachment of infecting organisms to the prosthesis surface and the formation of biofilm, as a result of this one would expect that the “affinity” of organisms to attach themselves to the different biomaterials surfaces would vary. This issue has not been explored to a great degree as there is little clinical data studying the potential influence of different biomaterials on post-operative joint infection. Based on the available medical literature, the incidence of post-operative joint infection does not vary regardless of whether cemented arthroplasty components or un-cemented arthroplasty components are used, and that the presence of hydroxyapatite on the un-cemented surfaces does not seem to influence the incidence of post-operative joint infection. The literature however confirms a higher incidence of post-operative joint infection following the use of a metal-on-metal bearing surface.
There are a number of potential reasons as to why the incidence of post-operative joint infection may be higher after the use of a MoM bearing surface. For example, the failure of a MoM bearing surface can result in adverse local tissue reactions (ALVA) and extensive soft tissue destruction, which could then provide a favorable environment for bacterial proliferation. It has been demonstrated that metal particles generated by the MoM bearing surface increased the potential risk of post-operative joint infection because of the ability of metal particles to modulate the immune system and bacterial growth. The question that remains is whether other bearing surfaces influence the incidence of post-operative joint infection as well.
We selected therefore a population of patient with high risk for infection. Most previous studies of total THA in sickle cell disease reported high risks of complications including infections and a higher incidence of failure with infection than after THA for other conditions. Based on our experience (procedures during the last 35 years), we questioned these conclusions and retrospectively reviewed arthroplasties performed in 435 patients (636 hips) with sickle cell disease from the year1981 to the year 2010. The mean age of the patients at the time of surgery was 32 years. The bearing surfaces were Metal on PE (Me/PE) in 120 hips, Ceramic on PE (Ce/PE) in 356 hips and Ceramic on Ceramic (Ce/Ce) in 160 hips. The minimum followup was 5 years (mean, 13 years; range, 5–35 years).
The frequency of preoperative infection in this population was evaluated according to the number of patients with a previous infection in the site of the prosthesis documented as previous osteomyelitis in the proximal part of the femur, as a positive culture observed on the excised specimens, or as evidence of infection on the histologic specimens. The ratio of post- operative infection was defined as the percentage of patients with early infection to the percentage of patients with late infection requiring revision arthroplasty. We also checked the number of recurrent infections after revision arthroplasties for infection or loosening.
A previous preoperative infection had occurred in 45 of 636 hips (7%). This previous infection was documented by patient histories or by the cultures obtained during surgery. Late infection occurred in 13 (3.6%) of the 356 Ce/PE hips, in 10 (8.4%) of the 120 Me/PE hips and in 2 (1.2%) of the 160 Ce/Ce hips. Only 4 infections occurred among the 45 hips with a previous infection, but all occurred with Me/PE bearing.
We revised therefore 25 hips (3.9%) for infection and during the same period 65 hips (10.2%) for aseptic loosening in this series of patients. After revision, 43 of these 88 revised hips had Me/PE bearing surface, 31 had Ce/PE, and 14 had Ce/Ce. The rate of recurrent infection was 16% (4 hips) on the 25 revisions for infection, and all the recurrent infections occurred among the 15 hips with Me/PE bearings as compared with none among the 10 Ce/PE bearings. The rate of infection among the 65 hips revised for aseptic loosening was 10.7% (3 hips) for the 28 hips with Me/PE bearings, 4.7% (1 hip) for the 21 hips with Ce/PE, and 0% for the 14 Ce/Ce bearings.
Our results indicate that implants with articulations involving a metal component are more prone to becoming infected than those involving ceramic-on- ceramic or ceramic-on-polyethylene bearings, where no metal ion release occurs from the bearing and only minimal release occurs from the taper junction both for primary and revision arthroplasties in this series of patients with SCD who are at high risk for infection. Accepting this theory means that infection in arthroplasty occurs at the joint (bearing surface or taper) and not at the stem interface that was “protected” in all our patients with cement with antibiotics for the stems and PE cemented cups but not for the patients with Ce insert in a metal-back for Ce/Ce hips. This is concordance with the reported organisms of our series that show no salmonella (usual pathogen of diaphysis osteomyelitis in SCD) as comparison with Staphylococcus (30 hips among 44) that has been reported as a pathogen of septic arthritis in SCD.
It is thus possible that the bearing type influences local and, eventually, also systemic, host defenses. It has been shown that different particles have different biological activities and subsequently propensities for macrophage activation and osteolysis formation. According to some studies, ceramic particles are the most bio-tolerant. On the other hand, the corrosion products of metal particles can induce profound derangements of local tissue, resulting in pseudo-infections or pseudotumors in some patients. The relative bio-tolerance of polyethylene Toll-like receptors may also be involved in the pathogenesis of a decreased local immune response to metal ions. Innate and adaptive immune responses, in which TLR plays an important role, are consequently decreased. Low-grade infections that would otherwise remain permanently under control are prone to actuate in an immunosuppressed milieu. Metal ions activate antigene presenting cells (APC), which lead to an enhanced expression of the MHC-peptide and costimulatory molecules. The fate of the response, however, depends on which type of T-cell receptor the costimulatory molecules act on.
The whole spectrum of immunological changes in the local and systemic environment caused by the release of particles from the bearings is not known in detail. However, growing evidence demonstrates that important derangements do occur, which alter local and systemic immunologic mechanisms and induce a status of relative immunodeficiency, resulting in higher infection rates. The profound influence of metal ions on periprosthetic tissues has been summarized by Konttinen.38
It seems that metal ion release influences the incidence of clinically manifested PJI. Articulation is probably the main source of metal ions (particularly in MoM implants), and the importance of taper junctions has also recently been acknowledged. Tapers are prone to crevice corrosion, particularly in high torque conditions that occur with larger heads.
Our results indicate that implants with articulations involving a metal component are more prone to becoming infected than those involving ceramic-on- ceramic or ceramic-on-polyethylene bearings, where no metal ion release occurs from the bearing and only minimal release occurs from the taper junction.