The US population-based cancer registry Surveillance, Epidemiology, and End Results (SEER) database provides an opportunity to evaluate the incidence and survival rates of Ewing sarcoma (ES) for the past 3 decades. This analysis reflects trends for the diagnosis of localized versus metastatic disease and changes in ES survival in a setting of wide-ranging cancer care institutions across the United States, which is expected to be different from clinical trials published to date.Materials/Methods
Data from the SEER public-access database were reviewed for the diagnosis of ES of the bone among patients of 1 to 19 years of age between 1973 and 2004. Age-adjusted incidence was analyzed for the entire group and for localized and metastatic disease separately over the past 3 decades. Actuarial survival rates were examined for 3 intervals: 1973 to 1982, 1983 to 1992, and 1993 to 2004.Results
The overall incidence of ES seemed to remain unchanged with an average of 2.93 cases/1,000,000 reported annually between 1973 and 2004. The proportion of patients with distant metastasis among all ES cases remained in the 26% to 28% range, whereas the percent of localized cases slightly increased from 57% in 1973 to 1982 to 67% in 1993 to 2004 and the proportion of unstaged cases decreased from 17% to 5%. The 5-year survival of localized disease increased from 44% in the first study decade to 68% in the period after 1993, whereas 5-year survival of metastatic disease increased from 16% to 39%. The corresponding 10-year survival increased from 39% to 63% for localized disease and from 16% to 32% for metastatic ES.Conclusions
The incidence of ES has not increased appreciably over the last 30 years. A marked decrease in the proportion of unstaged cases may be reflective of diagnostic improvement or changes in reporting. There is a clear improvement in survival for both localized and metastatic disease. Poorer outcome of metastatic patients still warrants intensification of therapy, which is currently being tested in several ongoing trials.