Introduction: Deep infection following distal femoral endoprosthetic replacement remains an uncommon, (< 7%), but serious complication; we present the results all three-phase revisions performed at our unit.

Method: Using the endoprosthesis-survivorship database we identified and analysed 15 consecutive cases, (including MSTS functional assessment of all available patients), performed between 1993 and 2002. The primary replacement had been performed for trauma and fourteen for limb reconstruction following excision of tumour. All cases underwent a three-phase revision. The first stage involved debridement and exchange of prosthesis for a custom-made antibiotic-impregnated spacer. Following at least six weeks of intravenous antibiotics, a further endoprosthesis was inserted.

Results: Eight patients had complete clinical, radiological & biochemical resolution of infection, (mean follow-up 60 months). Mean MSTS score for this group was 83% (range 60–97%). The remaining seven had recurrence of infection, all within 18 months. Of this group, two underwent a successful second revision procedure with conversion to a total femoral replacement. Two cases are satisfactorily managed with antibiotic suppression therapy and three have required amputation. Two of these cases underwent above-knee amputation following a failed second revision, whilst the third was given a femoral stump endoprosthesis to avoid disarticulation. Revision was generally more successful in younger patients. Neither the original pathology nor the timing of revision surgery appeared to affect outcome. Negative tissue cultures from the first stage were associated with a successful result. Very high levels of inflammatory markers were associated with failure of revision

Conclusion: We recommend two-stage revision of distal femoral replacement as an effective treatment for infection, allowing limb salvage with excellent functional outcome in the majority of patients. The antibiotic phase may need to exceed six weeks in certain cases, and levels of inflammatory markers appear to be critical. If this revision fails, conversion to a total femoral replacement should be considered.

Introduction: We present a review of the long-term results of custom-made massive unicondylar femoral replacement for reconstruction following tumour excision, and compare the functional outcome of this procedure with prosthetic distal femoral replacement.

Method: Using our centre’s endoprosthetic database we identified and analysed all cases of massive unicondylar femoral replacement performed at our unit (group 1). Patients were evaluated for function, (Musculoskeletal Tumour Society System), and for stability (adapted from Oxford Knee Score). An age and sex-matched cohort of patients who had undergone distal femoral replacements for similar pathologies, and in who the follow-up was of a comparable time period (group 2) was evaluated in an identical manner. Statistical analysis was performed on the results.

Results: Twelve cases of massive unicondylar replacement have been performed between 1990 and 2001, for a variety of malignant and benign tumours. There have been no incidences of infection, aseptic loosening or tumour recurrence. One patient has died of metastatic disease and another has undergone revision to distal femoral replacement for osteoarthritis. Of the remaining ten patients, nine were available for assessment Each of the two groups consisted of 5 males and 4 females, with mean age 48 years in group 1 and 49 years in group 2. The average follow-up since surgery in both groups was 10 years. The mean MSTS and stability scores of group 1 were 83% and 3.9 respectively, and 71% and 3.2 for group 2; the difference in scores between groups was statistically significant, (p< 0.02).

Conclusion: With stringent case selection criteria, the custom-made massive unicondylar femoral replacement generally produces a good outcome, with functional results significantly better than distal femoral replacement. This may be because a substantial proportion of the knee joint with at least one cruciate and one collateral ligament are kept intact, thus facilitating enhanced proprioception.