Purpose: Hinged knee prostheses are mainly used for reconstruction after major tumour resection. Aseptic loosening is the main problem with these implants. One of the solutions proposed to reduce the rate of loosening is to add a hydroxyapatite collar on the shaft stems. This work was conducted to study bone ingrowth with a new hinged implant with a hydroxyapatite collar at the junction between the zone of resection and the shaft.

Material and methods: Twenty-nine massive prostheses with a hydroxyapatite collar were implanted between 1998 and 2001. Nine patients were excluded from the analysis because follow-up was less than two years. This retrospective analysis thus compared twenty massive prostheses with twenty matched hinged GUEPAR prostheses without a collar. Bony ingrowth was measured on plain x-rays (two orthogonal views) at 6, 12, 24, and 36 months. Filling of the gap between the bone and the implant was also assessed. Signs of loosening were noted.

Results: Mean bony ingrowth in implants with a hydroxyapatite collar was 6.58 mm at 6 months 9.84 mm at 12 months, 12.3 mm at 24 months and 13.25 mm at 36 months. Mean bony ingrowth in the implants without a hydroxyapatite collar was 1.65 mm at 6 months, 3.31 mm at 12 months, 4.8 mm at 24 months and 4.35 mm at 36 months. In the implants with a collar, gap filling was partial in five cases and total in 15. In implants without a collar, there was no gap filling in eight cases, partial filling in two cases and total filling in fifteen cases.

Discussion: Prostheses with a hydroxyapatite collar enable better radiological bony ingrowth than observed in implants without a hydroxyapatite collar. Gap filling is better for prostheses with a collar. There was no case of loosening at last follow-up for implants with a hydroxyapatite collar.

Conclusion: In light of these results, shaft anchorage appears to be better with implants with a hydrosyapatite collar. Confirmation of improvement in clinical outcome and lower rate of aseptic loosening will require longer follow-up.

Purpose: Infection after resection and total knee arthoplasty for malignant bone tumours in children is a serious complication which may compromise limb salvage. The purpose of this work was to study the aetiology, treatment and prognosis of this event.

Material and methods: Among the 169 total knee arthroplasties performed for malignant bone tumours between 1981 and 1999, we selected 17 patients meeting the following criteria: proven infection with identified germ on deep samples or presence of a fistula more than two years after surgery. All of the patients had osteogenic sarcomas (excluding Ewing sarcomas which account for 30% of the tumours in this localisation). The bone tumour involved the femur (n=11) or the tibia (n=6) and required extraarticular (n=14) or transarticular (n=2) resection. Infections were primary (n=9) developing after the first operation, or secondary (n=8) to surgical revision in six, joint wound in one, or haematogeneous dissemination in one patient. The causal germ was identified in thirteen patients (76%) and was a staphylococcus in all cases. Treatment included systemic antibiotics and lavage (n=10), one-procedure change in prosthesis (n=3), removal of the implant with replacement by a spacer (n=2), surgical abstention (n=2), or amputation (n=1).

Results: Mean follow-up was eight years (2 – 16 years). On the average, treatment of infection lasted 51 months and required a 3.9 surgical interventions. At last follow-up, infection was considered cured in 70% of the patients who were free of clinical or laboratory signs of infection without antibiotics for at least one year. The arthroplasty could be preserved in one-third of the cases (22% of the primary infections and 50% of the secondary infections). Another treatment, arthrodesis (n=6), Borggreve procedure (n=1), or amputation (n=4), was given in the other two-thirds.

Discussion: The 10% complication rate observed here is in agreement with data in the literature. Development of primary infection is influenced more by the histology of the tumour and the presence of skin wounds (methotrexate) than by tumour site or type of resection. The diagnosis of primary infection is made late, often at the end of the postoperative chemotherapy protocol. Changing the implant is the ideal treatment. Secondary infection is characteristically less difficult to diagnose; infection is recognised earlier and the chances of preserving the implant are better.