Synovial sarcoma is a high-grade tumor that is associated with poor prognosis. Previous studies analyzing prognostic factors are limited because of inclusion of heterogeneous cohorts of patients with nonextremity and recurrent tumors. The objective of this study was to determine independent prognostic factors of primary synovial sarcoma localized to the extremity.

Between July 1, 1982, and June 30, 1996, 112 patients underwent surgical resection for cure at our institution and then were followed-up prospectively. Clinical and pathologic factors examined for prognostic value included age, sex, tumor site and location, depth, size, microscopic status of surgical margins, invasion of bone or neurovascular structures, and monophasic or biphasic histology. The end points analyzed were the time to first local recurrence that was not preceded by a distant recurrence, time to any distant recurrence, and time to disease-related mortality. These end points were modeled using the method of Kaplan and Meier and analyzed by the log-rank test and Cox regression.

The median duration of follow-up among survivors in this cohort of 112 patients was 72 months. The 5-year local-recurrence, distant-recurrence, and mortality rates were 12%, 39%, and 25%, respectively. Tumor size > or = 5 cm (P =.001; relative risk [RR] = 2. 7; 95% confidence interval [CI], 1.5 to 5.2) and the presence of bone or neurovascular invasion (P =.04; RR = 2.3; 95% CI, 1.0 to 5. 3) were independent adverse predictors of distant recurrence. Tumor size > or= 5 cm (P =.003; RR = 2.3; 95% CI, 1.4 to 6.3) and the presence of bone or neurovascular invasion (P =.03; RR = 2.7; 95% CI, 1.0 to 6.5) were also independent adverse predictors of mortality.

The natural history of primary synovial sarcoma of the extremity is related to tumor size and invasion of bone and neurovascular structures.