Tumor size and metastases are known risk factors in Ewing tumors. Adequate staging is essential to stratify treatment intensity, but TNM staging is not established.

The validity of TNM staging was tested based on tumor volume (T1 ≤200 ml; T2 > 200 ml ≤500 ml; T3, > 500 ml), the presence or absence of lymph node metastases (N0, N1), and distant metastases (M0, no metastases; M1 lung/pleura metastases; M1a ≤5 nodules; M1b > 5 lesions; M2 bone metastases; M2a, 1 lesion; M2b > 1 lesion and/or microscopic bone marrow contamination; M3, multi-system metastases). 1799 Ewing sarcoma patients of the (EI)CESS/EE99 studies entered into the Muenster database from 1981–2008 were analyzed.

Ten-year event-free survival (EFS) was 0.46. EFS in patients without metastases (T1-3N0M0) was 0.56 compared to 0.22 in metastatic patients (T1-3N0,1M1-3), p< .0001. In non-metastatic patients, tumor volume discriminated EFS: T1:0.62; T2:0.43; T3:0.40, p< .0001. The rare event of lymph node metastases correlated with unfavorable prognosis (N0:0.47, N1:0.12, p< .0001). The difference in EFS between pulmonary, skeletal and multi-system dissemination was significant: M1:0.29, M2:0.23; M3:0.12, p< .0001. The discrimination of M1 subgroups (M1a/M1b) was of prognostic relevance (p=.0050); M2 subgroups (M2a/M2b) discriminated outcome less clearly (p=.0457).

TNM staging is appropriate in Ewing tumors and should be incorporated in future trials.