Epithelioid sarcoma represents a rare and aggressive soft tissue malignancy characterized by SMARCB1/INI1 inactivation and propensity for local recurrence and distant metastasis. Comprehensive population-based analyses of prognostic determinants remain limited. We evaluated survival outcomes and identified independent prognostic factors using the Surveillance, Epidemiology, and End Results (SEER) database. We retrospectively analyzed 1123 patients diagnosed with epithelioid sarcoma between 2001 and 2022 from the SEER 17 registries database. Cox proportional hazards regression models were employed to identify prognostic factors for cancer-specific survival. Kaplan–Meier analysis with log-rank testing was performed for survival comparisons. The cohort comprised 634 males (56.5%) and 489 females (43.5%) with a mean age of 47.1 ± 21.3 years. Stage distribution revealed localized disease in 45.3%, regional involvement in 26.5%, and distant metastases in 28.2% of evaluable cases. Surgical resection was performed in 70.8% of patients. The median overall survival was 84 months, with a 5-year cancer-specific survival of 53.7%. Multivariate analysis identified distant stage (adjusted hazard ratio aHR = 5.379, 95% confidence interval CI: 3.652–7.921, P < .001), regional stage (aHR = 2.474, 95% CI: 1.711–3.578, P < .001), tumor size per centimeter (aHR = 1.078, 95% CI: 1.052–1.104, P < .001), and age per year (aHR = 1.011, 95% CI: 1.004–1.018, P = .002) as adverse prognostic factors. Surgical intervention demonstrated substantial survival benefit (aHR = 0.397, 95% CI: 0.280–0.563, P < .001). This contemporary population-based analysis represents the largest evaluation of epithelioid sarcoma outcomes to date, encompassing more than twice the sample size of prior SEER analyses. Stage at diagnosis emerges as the predominant prognostic determinant, with surgical resection conferring consistent survival benefit across disease stages. Compared with historical cohorts, observed improvements in 5-year survival likely reflect advances in diagnostic recognition and multimodal management.
Akef Obeidat (Fri,) studied this question.