A Population-Based Examination of the Surgical Outcomes for Patients with Esophageal Sarcoma.

Background: Esophageal sarcoma (ES) is a rare malignancy. The literature is limited to small case series and reports. This study used a population data set to study the characteristics, treatments, surgical outcomes, and prognostic factors for survival among ES patients.

Methods: The study identified 178 ES cases (0.3 %) and 63,548 esophageal carcinoma (EC) cases (99.7 %) including adenocarcinoma and squamous cell carcinoma using the Surveillance, Epidemiology, and End Results (SEER) Registry (1973-2011). Characteristics and therapeutics were compared between ES and EC. Survival data were analyzed using Kaplan-Meier estimation. Uni- and multivariate Cox proportional hazard models determined predictors of 5-year overall survival (OS).

Results: Compared with the EC patients, the ES patients were more likely to be women, to have localized tumors, and to undergo surgery but less likely to receive radiation (p < 0.001). The most common histologies were carcinosarcoma, leiomyosarcoma, and gastrointestinal stromal tumor (GIST). The 5-year OS for the ES patients was 38 % compared with 17 % for the EC patients (p < 0.001). The median survival period for the ES and EC surgical patients with locoregional disease was respectively 50 and 24 months. The ES patients with nonmetastatic disease who received surgery had better OS than those who did not (37 vs. 14 %; p < 0.0001). In the multivariate analysis, age and advanced stage conferred worse OS, whereas GIST histology and surgery were favorable predictors for OS.

Conclusion: The ES patients were more likely to have localized disease, to be treated with surgery, and to have better OS than the EC patients. The survival benefit of surgery suggests that surgery should be the primary treatment for ES patients with resectable disease, particularly those with GIST.