Postrelapse survival in osteosarcoma of the extremities: prognostic factors for long-term survival.

PURPOSE: To identify factors that influence postrelapse survival (PRS) in patients with nonmetastatic osteosarcoma of the extremity. PATIENTS AND METHODS: One hundred sixty-two patients with recurrent osteosarcoma of the extremity were retrospectively reviewed. The first-line treatment included surgery of the primary lesion and chemotherapy with methotrexate, doxorubicin, cisplatin, and ifosfamide. RESULTS: The projected 5-year PRS rate was 28%. Patients who had complete surgery of recurrence had a 5-year PRS of 39%, whereas for those who did not have complete surgery, PRS was 0% at 3 years (P <.0001). In the latter group, PRS was not influenced by site of recurrence and relapse-free interval (RFI), although it was influenced (P =.006) by the use of second-line chemotherapy (PRS, 53% at 12 months for patients who received chemotherapy v 12% for those who did not). In patients who had complete surgery, PRS was influenced by site of relapse (5-year PRS, lung 44%, other 19%; P <.06), RFI (5-year PRS at < or = 24 months, 20%; at > 24 months, 60%; P <.0001), and number of lung metastases (5-year PRS, two or fewer nodules, 59%; more than two nodules, 14%; P <.0001) but not by the use of a second-line chemotherapy treatment. CONCLUSION: RFI, site of metastases, and number of pulmonary nodules are the main prognostic factors for PRS in osteosarcoma. Complete surgery of recurrence is pivotal in the strategy of treatment. Patients with unresectable recurrence benefit from second-line chemotherapy, whereas our data do not support a generalized use of chemotherapy after complete surgery of first recurrence.