Prognostic features and outcomes in primary liver sarcoma.
150 Background: Primary liver sarcoma (PLS) is a rare and aggressive hepatic malignancy. Due to the low incidence of PLS, prognostic factors have not been well characterized. The purpose of our study was to evaluate survival outcomes in primary hepatic sarcomas and determine which factors predict survival. Methods: The Surveillance Epidemiology and End Results registry was used to identify patients with PLS from 1988-2009. Patients were evaluated by standard clinical and pathological indices including: age, gender, race, tumor size, tumor grade, histology, and extent of disease. Treatment related factors included surgery and radiation. Overall survival was assessed by Kaplan-Meier method. Univariate and stepwise multivariate Cox proportional hazards analyses were performed to identify prognostic factors. Results: 541 patients with PLS were identified. The mean age was 52 and most patients were male (55%) and white (75%). The most common histology type was blood vessel tumors (50.1%) followed by soft tissue neoplasms (17.7%) and complex mixed-stromal neoplasms (14.6%). Only 33% of patients underwent surgery and most (93%) did not receive radiation. When assessing outcomes, we observed median overall survival (OS) and cancer specific survival (CSS) of 6 months for the entire cohort. When stratified by treatment type, those who received surgery + radiation had the best survival (MS=97mos, p<.001) compared to those who received either radiation alone (MS=5mos) or no treatment (MS=2mos). Stepwise multivariate analysis showed that age, male gender, tumor size, advanced stage, and no surgery were independent predictors of worse survival, all p-values < .005. Conclusions: Primary liver sarcoma is an aggressive hepatic malignancy with low median OS of 6 months. Patients treated with surgery + radiation had the best outcome with a median survival of 97 months. Independent predictors for decreased survival included age, gender, tumor size, advanced stage, and no surgery.