BACKGROUND: Metastatic dissemination in osteosarcoma occurs haematogenously, though regional lymph node involvement is rarely reported. We investigated incidence, patient characteristics and survival for patients with osteosarcoma and regional lymph node involvement at diagnosis. METHODS: We identified 2748 cases of high-grade osteosarcoma with available information regarding regional lymph node involvement in the Surveillance Epidemiology and End Results database from 1973 to 2009. Demographics were compared using chi-squared tests or t-tests. Overall survival was estimated using Kaplan-Meier method and compared with log-rank tests. Multivariate analysis of overall survival was performed using Cox proportional hazards methods. RESULTS: There were 74 patients (2.7%) with regional lymph node involvement at diagnosis of whom 19 (0.7%) were pathologically confirmed. Patients with regional node involvement were more likely to have extraskeletal tumours, distant metastases, tumours arising outside the lower extremity (p<0.0001 for all comparisons) and larger tumours (p=0.033). Five-year overall survival in those with and without regional node involvement was 10.9% (95% confidence interval (CI) 4.6-20.4) and 54.3% (95% CI 52.2-56.4; p<0.0001). In multivariate analysis, regional node involvement remained predictive of inferior survival after controlling for differences in metastatic status, age, tumour site and extraskeletal origin (hazard ratio 2.05, 95% CI 1.57-2.67; p<0.0001). Similar survival results were found when the analysis was restricted to patients with pathologically confirmed positive or negative regional lymph nodes. CONCLUSION: This analysis confirms that regional node involvement is a significant adverse prognostic factor that is independent of metastatic status, extraskeletal origin, age and tumour site.
BACKGROUND: Metastatic dissemination in osteosarcoma occurs haematogenously, though regional lymph node involvement is rarely reported. We investigated incidence, patient characteristics and survival for patients with osteosarcoma and regional lymph node involvement at diagnosis. METHODS: We identified 2748 cases of high-grade osteosarcoma with available information regarding regional lymph node involvement in the Surveillance Epidemiology and End Results database from 1973 to 2009. Demographics were compared using chi-squared tests or t-tests. Overall survival was estimated using Kaplan-Meier method and compared with log-rank tests. Multivariate analysis of overall survival was performed using Cox proportional hazards methods. RESULTS: There were 74 patients (2.7%) with regional lymph node involvement at diagnosis of whom 19 (0.7%) were pathologically confirmed. Patients with regional node involvement were more likely to have extraskeletal tumours, distant metastases, tumours arising outside the lower extremity (p<0.0001 for all comparisons) and larger tumours (p=0.033). Five-year overall survival in those with and without regional node involvement was 10.9% (95% confidence interval (CI) 4.6-20.4) and 54.3% (95% CI 52.2-56.4; p<0.0001). In multivariate analysis, regional node involvement remained predictive of inferior survival after controlling for differences in metastatic status, age, tumour site and extraskeletal origin (hazard ratio 2.05, 95% CI 1.57-2.67; p<0.0001). Similar survival results were found when the analysis was restricted to patients with pathologically confirmed positive or negative regional lymph nodes. CONCLUSION: This analysis confirms that regional node involvement is a significant adverse prognostic factor that is independent of metastatic status, extraskeletal origin, age and tumour site.