Purpose: The impact of different local therapy approaches on local control, event-free survival, and secondary malignancies in the CESS 81, CESS 86, and EICESS 92 trials was investigated.
Methods and materials: The data of 1058 patients with localized Ewing tumors were analyzed. Wherever feasible, a surgical local therapy approach was used. In patients with a poor histologic response or with intralesional and marginal resections, this was to be followed by radiotherapy (RT). In EICESS 92, preoperative RT was introduced for patients with expected close resection margins. Definitive RT was used in cases in which surgical resection seemed impossible. In CESS 81, vincristine, adriamycin, cyclophosphamide, and actinomycin D was used. In CESS 86, vincristine, adriamycin, ifosfamide, and actinomycin D was introduced for patients with central tumors or primaries >100 cm(3). In CESS 92, etoposide, vincristine, adriamycin, ifosfamide, and actinomycin D was randomized against vincristine, adriamycin, ifosfamide, and actinomycin D in patients with primaries >100 cm(3).
Results: The rate of local failure was 7.5% after surgery with or without postoperative RT, and was 5.3% after preoperative and 26.3% after definitive RT (p = 0.001). Event-free survival was reduced after definitive RT (p = 0.0001). Irradiated patients represented a negatively selected population with unfavorable tumor sites. Definitive RT showed comparable local control to that of postoperative RT after intralesional resections. Patients with postoperative RT had improved local control after intralesional resections and in tumors with wide resection and poor histologic response compared with patients receiving surgery alone. Patients with marginal resections with or without postoperative radiotherapy showed comparable local control, yet the number of patients with good histologic response was higher in the latter treatment group (72.2% vs. 38.5%).
Conclusion: Patients with resectable tumors after initial chemotherapy had a low local failure rate. With preoperative RT, local control was comparable. RT is indicated to avoid intralesional resections. After intralesional or marginal resections and after a poor histologic response and wide resection, postoperative RT may improve local control.