Infection of total hip replacement (THR) is a serious complication, usually necessitating complete removal of implants and thorough debridement of the site. Mostly implant removal is followed by several weeks of antibiotic therapy before a new prosthesis is inserted. One stage exchange using antibiotic containing cement did not gain widespread use because of several risks, although the possible clinical and economic advantages are evident. Uncemented revision techniques seem to provide better long term results, however in septic cases its use so far has been restricted to two stage procedures. Allograft bone impregnated with high loads of antibiotics using a special technique (antibiotic bone compound ABC) is likely to create markedly higher concentrations of antibiotics in its surrounding than cement.

Between 1998 and 2004 37 patients with infected THR were treated using a standardized protocol. Patients were 17 male and 20 female, their age at revision was 42–83 yrs with a mean of 68,5yrs. After removal of the implants a radical debridement and intensive pulsed lavage was performed. Bone deficiencies were filled with cancellous bone, impregnated with high loads of Vancomycin or (in cases with gramnegative cultures) a combination with Tobramycin (ABC). After impaction uncemented implants were anchored following the principles of press-fit fixation, all without cement; usually we preferred a rectangular diameter titanium stem and a hemispherical cup. Additional ABC was placed around eventually uncovered parts of the implants and impacted for good stability. Wounds were drained and closed immediately; rehabilitation was performed as after non-septic surgery. Cultures taken intraoperatively revealed growth of coag.neg.staph (19x), s.aureus (11x), MRSA (5x), enterococci (8x) and other grampositive pathogens (6x), respectively. In 8 hips gramnegative germs were found additionally. Patients were evaluated prospectively 2 weeks, 6 weeks, 3 months, 6 months and one year after surgery. After the first year evaluation was retrospective. Follow up included clinical and radiological examination and laboratory data (CRP, ESR, blood count, urea and creatinine).

Three hips required re-revision because of re-infection, the remaining 34 hips (92%) stayed infect free and stable throughout a follow up period between 2 and 8 years (mean 4,4yrs). No adverse side effects could be found. Incorporation of grafted bone followed the same patterns as known from unimpregnated grafts.

Infected THRs may be exchanged within a single procedure using antibiotic impregnated allograft bone, providing biological reconstruction of bone stock, stable insertion of an uncemented implant and control of infection. Since only one intervention is necessary rehabilitation of patients is improved and costs are markedly reduced. Improved long term results may be expected.

Infection of total knee replacement (TKR) is considered a devastating complication, which necessitates complete removal and thorough debridement of the site. Usually long term antibiotic treatment and a multitude of surgical interventions within a period of several months are required until a definitive supply can be achieved. Osseous defects are common in such conditions and need to be addressed during re-implantation. Managing removal, debridement, reconstruction and re-implantation within a single operation is the ideal solution, both for the patient and the treating team, but rarely executed due to the fear of re-infection. Allograft bone may be impregnated with high loads of antibiotics using a special incubation technique. The resulting antibiotic bone compound (ABC) provides high and long lasting antibiotic levels at the site of infection and is likely to restore bone stock. We have investigated the results of one-stage exchange of infected TKR using ABC together with uncemented implants.

Between 1998 and 2004 nineteen exchange procedures of infected TKRs were performed in a single stage, all of them without the use of bone cement. After removal of the implants and radical debridement bone voids were filled with ABC using a modified impaction technique. Consequently, new uncemented implants were inserted. We mainly used the revision type of the LCS knee (DePuy, J& J) as long as ligamentary stability was considered sufficient. Otherwise, we used a custom-made uncemented version of the LINK Rotational Endo Model. Joints were drained and closed immediately; rehabilitation did not differ from uninfected revision.

One knee required re-revision because of persisting infection. The remaining 18 patients stayed infect-free for a period between 2 and 8 years after surgery. In two knees loosening was found after one year, once of the tibial and once of the femoral component. Both were found infect-free at the time of re-revision. All could be successfully revised using the same technique again. No adverse side effects could be found. Incorporation appeared as after grafting with unimpregnated bone grafts.

Using antibiotic-impregnated allografts eradication of pathogens, grafting of defects and re-insertion of an uncemented prosthesis may be accomplished in a one-stage procedure. Since the graft gradually is replaced by healthy own bone, improved long-term results may be expected as well as improved conditions in the case of another revision.

Infection of total hip replacement still is considered a devastating complication. One-stage revision, meaning complete removal of the implant and thorough debridement of the site together with the insertion of a new prosthesis during the same operation, is desirable because of improved rehabilitation of the patient and reduced costs. Although this method is known since more than 30 years it is not used widely yet because of several related risks: known methods rely on the use of antibiotic-loaded cement, which often has shown insufficient release of the added antibiotic. The carrier may even act as a bed for colonisation with selected bacteria. Osseous conditions presented during revisions prevent interconnection of the cement with the sclerotic bone. The toxic monomers of PMMA and heating during polymerisation causes necroses which may be origin of repeated loosening and new growth of bacteria. Filling the defects with cement provides unfavourable conditions in case of another revision, which has to be expected at a high percentage. To overcome these disadvantages uncemented techniques seem to be favourable.

After removal of the implants debridement is performed as in conventional septic surgery. After thorough cleaning and rinsing, bone voids are filled with bone graft using an impaction technique. We use allograft or xenograft bone that is free from antigenic material but intact structures of bone concerning collagen and mineral content. The bone is impregnated with high loads of antibiotic, using a specific incubation technique. There are two options of antibiotic impregnation: vancomycin (“V”) or tobramycin (“T”), the choice being dependent on the causative pathogen isolated. Combinations are possible in cases of mixed infections. The impregnation procedure guarantees high levels of antibiotics at the grafting site for several weeks during which the antibiotic is released into the surroundings. Systemic drug levels are usually undetectable. At the acetabular site we take care that the ground be sufficiently filled with antibiotic graft. Preferably an uncemented cup of hemispherical design is inserted. At the femoral site we prefer implants with a rectangular diameter. This design enables stable press fit contact with the shaft medially and laterally and leaves enough space for graft impaction at the posterior and anterior aspect of the endoprosthesis. Wounds are drained and closed immediately; rehabilitation is performed as after non-septic surgery.

Between 1998 and 2004 thirty-seven patients have been revised because of culture-proven infection of hip endoprosthesis. Causative pathogens were Coag.neg. staph (18x), S.aureus (11x), MRSA (4x), enterococci (9x) and other gram-positive pathogens (3x). In 6 hips gram-negative germs were found additionally. All hips could be followed up with a minimum of 2 years and a maximum of 8 years (mean: 4.1 years).

Wound healing was uneventful in all cases. Mean hospital stay was 16 days (10–32 days). Rehabilitation was in the range of uncomplicated primary THR in cases with short history of infection (up to 3 months) and prolonged in relation to duration of infection and amount of preceding surgery. In three hips there was recurrence of the infection, diagnosed between 6 and 12 weeks after surgery. In one of them the well-fixed stem had not been exchanged, in another one a technical error had occurred during impregnation of the bone graft. This one could be successfully re-operated using the same technique with appropriately impregnated bone graft, the other two were converted to a girdlestone situation. All other 32 hips showed no sign of infection until the last follow-up.

Bone processed in an adequate way represents an excellent carrier for vancomycin and tobramycin. With antibiotic graft compound eradication of pathogens, grafting of bony defects and re-insertion of an uncemented prosthesis may be accomplished in a single operation, making it an ideal tool in one stage non-cemented revision for infected total hip replacement. However, principles of septic surgery need to be observed. We now recommend removing even well-fixed prostheses and taking care, that we implant at least 50cc of well impregnated bone graft. Since the graft gradually is replaced by healthy own bone, improved conditions may be expected even in the case of another revision.