Multifactorial analysis of the survival of patients with distant metastasis arising from primary extremity sarcoma

BACKGROUND. Despite optimal multimodality limb-sparing therapy for extremit y soft tissue sarcoma (STS), a significant number of patients develop dista nt metastasis. The objective of this study was to analyze patterns of metas tatic disease and define prognostic factors for survival in a large group o f patients followed prospectively at a single institution. METHODS, Between July 1, 1982, and June 30, 1996, all adult patients admitt ed to the Memorial Sloan-Kettering Cancer Center with primary extremity sar coma were treated and prospectively followed. Patients who developed distan t metastases constituted the study group. Prognostic factors were analyzed for postmetastasis survival. These included both factors related to the pri mary tumor and factors related to the pattern of metastasis. Postmetastasis survival was modeled using the Kaplan-Meier method. Statistical significan ce was evaluated using the log rank test for univariate analysis and the Co x proportional hazards model for multivariate analysis. RESULTS. During the study period, the authors admitted and treated 994 pati ents with primary extremity STS. The median follow-up was 33 months. Distan t metastasis developed in 230 patients (23%). Median survival after distant metastasis was 11.6 months. The lungs were the first metastatic site in 16 9 patients (73%). Other first sites of metastasis included the skin and sof t tissues of the head and neck, trunk, and extremities. There was no statis tically significant difference in survival between patients with pulmonary and those with nonpulmonary metastatic disease. In multivariate analysis, r esection of metastatic disease, the length of the disease free interval, th e presence of a preceding local recurrence, and patient age > 50 years all were significant predictors of postmetastasis survival. Other factors that defined the primary tumor, including histologic grade, depth, and microscop ic margins, were not associated with postmetastasis survival. CONCLUSIONS. Despite optimal multimodality therapy, 23% of the patients in this series with primary extremity sarcoma developed distant metastasis. Me dian survival after metastasis was approximately 1 year. After metastasis, the independent favorable factors that are associated with patient survival include resection of the metastases, a long disease free interval, the abs ence of preceding local recurrence, and patient age < 50 years. Although a definitive conclusion regarding the benefit of resection can be made only w ith a randomized clinical trial, these data suggest that resection of metas tatic STS may contribute to patient survival, which in some cases may be lo ng term. (C) 1999 American Cancer Society.