PurposeThe present study is aimed at investigating whether (1) primary tumour surgery confers an improved survival on patients with metastatic osteosarcoma and (2) primary tumour surgery influences survival of patients with metastatic osteosarcoma differently according to primary tumour site.MethodsWe retrospectively identified 517 patients with high-grade, metastatic osteosarcoma in the Surveillance, Epidemiology, and End Results (SEER) database between 1994 and 2013. The effect of primary tumour surgery on survival was assessed using Kaplan-Meier analyses, log-rank tests, and multivariate Cox proportional hazard regression modeling.ResultsOf those 517 patients with metastatic osteosarcoma in the cohort, 351 patients (68%) underwent primary surgery, and 166 patients (32%) did not undergo surgery. Primary tumour surgery was associated with increased overall survival (hazard ratio (HR)=0.457, 95% CI 0.354-0.590, p<0.001) and cancer-specific survival (HR=0.422, 95% CI 0.325-0.550, p<0.001). When we focused on different primary tumour sites, receipt of primary tumour surgery significantly prolonged the survival of patients with extremity osteosarcoma (p<0.05 for overall and cancer-specific survival). However, for patients with pelvis/spine osteosarcoma, both univariate and multivariate analyses indicated that primary tumour surgery might not be associated with improved survival (p>0.05 for overall and cancer-specific survival).ConclusionsOur study is the first population-based analysis to provide evidence of a favourable prognostic impact of primary tumour surgery on metastatic extremity osteosarcoma patients but not metastatic axial (pelvis/spine) osteosarcoma patients. Moreover, we found that surgery type (resection of the primary tumor without amputation vs. amputation) did not influence survival in patients with metastatic osteosarcoma.