We analysed the outcome of patients with primary non-metastatic diaphyseal sarcomas who had en bloc resection with preservation of the adjoining joints and reconstruction with re-implantation of sterilised tumour bone after extracorporeal radiation (50 Gy). Between March 2005 and September 2009, 32 patients (16 Ewing’s sarcoma and 16 osteogenic sarcoma) with a mean age of 15 years (2 to 35) underwent this procedure. The femur was the most common site in 17 patients, followed by the tibia in 11, humerus in three and ulna in one. The mean resected length of bone was 19 cm (10 to 26). A total of 31 patients were available at a mean follow-up of 34 months (12 to 74). The mean time to union for all osteotomy sites was 7.3 months (3 to 28): metaphyseal osteotomy sites united quicker than diaphyseal osteotomy sites (5.8 months (3 to 10) versus 9.5 months (4 to 28)). There were three local recurrences, all in soft-tissue away from irradiated graft. At the time of final follow-up, 19 patients were free of disease, one was alive with disease and 11 had died of disease. The mean Musculoskeletal Tumor Society Score for 29 patients evaluated at the last follow-up was 26 (9 to 30).

Extracorporeal irradiation is an oncologically safe and inexpensive technique for limb salvage in diaphyseal sarcomas and has good functional results.

We retrospectively reviewed 101 consecutive patients with 114 femoral tumours treated by massive bone allograft at our institution between 1986 and 2005. There were 49 females and 52 males with a mean age of 20 years (4 to 74). At a median follow-up of 9.3 years (2 to 19.8), 36 reconstructions (31.5%) had failed. The allograft itself failed in 27 reconstructions (24%).

Mechanical complications such as delayed union, fracture and failure of fixation were studied. The most adverse factor on the outcome was the use of intramedullary nails, followed by post-operative chemotherapy, resection length > 17 cm and age > 18 years at the time of intervention. The simultaneous use of a vascularised fibular graft to protect the allograft from mechanical complications improved the outcome, but the use of intramedullary cementing was not as successful.

In order to improve the strength of the reconstruction and to advance the biology of host–graft integration, we suggest avoiding the use of intramedullary nails and titanium plates, but instead using stainless steel plates, as these gave better results. The use of a supplementary vascularised fibular graft should be strongly considered in adult patients with resection > 17 cm and in those who require post-operative chemotherapy.