The development of local recurrence after multimodal treatment of osteosarcoma is associated with a very poor prognosis. The importance of clear surgical margins has been demonstrated in multiple studies, however up to date there are no studies defining which margin width is safe from an oncological perspective. The purpose of this retrospective analysis was to evaluate whether margin width or other surgical and tumour-related factors influence the development of local recurrence in osteosarcoma patients.

The files of 1867 consecutive patients with high-grade central osteosarcoma of the extremities, the pelvic bones and the shoulder girdle, who had achieved a complete surgical remission during combined-modality therapy on neoadjuvant Cooperative Osteosarcoma Study Group (COSS) protocols between 1986 and 2005, were reviewed. Of those, the data required were available for 1369 patients, who were the subject of this analysis. Eighty of these patients developed a local recurrence during the course of their illness.

The median surgical margin width amounted to 45 mm (range, 0 to 140 mm) in the local recurrence (LR) group and 50 mm (range, 0 to 350 mm) in the non-local recurrence (NLR) group (p=0.106). No statistically significant difference between the two groups was found regarding tumour size (mean, 10.38 cm and 9.53 cm respectively, p=0.169), T-status (p=0.225) and presence of pathological fracture (p=0.231). However infiltration of the soft tissue beyond the periosteum was documented in 58.8% of the patients with local recurrence and only in 36.9% of the rest (p=0.003). Furthermore, in 50% of the LR group the biopsy had been performed in a centre other than the one performing the definitive tumour resection, compared to 30.2% of the NRL group (p=0.001).

In conclusion, the absolute metric width of surgical margins does not define oncological safety. Local recurrence is more likely to develop in patients with soft tissue infiltration beyond the periosteum or those biopsied in a centre other than the one performing the tumour resection.

Frequent imaging after a completed multimodal therapy of osteosarcoma is recommended by therapy optimization studies to detect local or systemic tumor recurrence. Considering the low rates of local recurrence, regular local imaging has to be questioned.

150 patients with osteosarcoma were treated in our department between 1991 and 2005. The median age of patients with osteosarcoma was 17 years with a range of 4 – 79 years and a female:male ratio of 1:1.1. The primary tumors of 147 patients were treated surgically, while 3 patients refused to be operated. After a wide resection, a tumor endoprosthesis was implanted in 103 (70.1%) of the 147 patients, 16 (10.9%) patients underwent a Borggreve rotationplasty, a resection and biological reconstruction was implemented in 10 (6.8%) patients, while further 18 (12.2%) patients were amputated. The median follow up was 95 months.

Local recurrences appeared in 2 (1.4%) patients which had been treated with a hemipelvectomy. After implantation of a tumor endoprosthesis, local recurrences were not observed. Postoperative complications observed after the implantation of a tumor endoprosthesis included infections (n=14; 13.6%), loosening, fractures and wearing of endoprotheses (n=7; 4.8%), luxation (n=1; 0.7%) as well as traumatic shaft fractures of involved bones (n=5; 3.4%). All complications included specific symptoms and were diagnosed outside the routine follow up.

In conclusion, local radiological imaging after resection of an osteosarcoma and reconstruction with a tumor endoprosthesis as a routine examination should be questioned, however it is definitely indicated in patients with specific symptoms.