Osteosarcoma as a second malignant neoplasm

Purpose: The aim of this study is to compare primary and secondary osteosarcomas using data from the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) program. Methods: Osteosarcoma cases were identified in the SEER Cancer Incidence Public-Use Database, 1973-96 (1511 primary and 133 secondary osteosarcomas). Secondary osteosarcomas were further classified as occurring within or outside the previously irradiated field. Comparisons among groups were performed by nominal logistic regression. Survival analyses were performed using Kaplan-Meir and proportional hazards regression. Results: Secondary osteosarcomas were more likely to have a non-extremity primary site and an older age at diagnosis (>40 years) (<0.0001 for both). Secondary osteosarcoma cases continued to be more likely to have a non-extremity site when excluding those occurring within the radiation field. Five-year overall survival was 50% lower for secondary osteosarcomas for both extremity and non-extremity sites. Primary malignancies associated with secondary osteosarcomas included 14 lymphomas/leukemias, six sarcomas, 54 carcinomas, and seven other cancers. Secondary osteosarcomas occurring within a field of radiation were more likely to occur at a younger age, have a malignancy with a primary morphology other than carcinoma, a non-limb site, and a longer duration between their primary malignant neoplasms and the development of osteosarcoma. No difference in the overall survival was noted between secondary osteosarcomas occurring within an irradiated field and those that did not. Conclusions: Secondary osteosarcomas are associated with carcinomas especially in the elderly. Although a 50% decreased 5-year survival was observed for secondary osteosarcomas in this study, this might not reflect the current outcomes with more aggressive therapy