Abstract Book

S879

ESTRO 37

and prognostic factor were reviewed. Survival outcomes were calculated using Kaplan-Meier method and prognostic factors were evaluated by Cox regression analysis. Results Median follow up was 33 months (mean 51, range, 2.5- 179.2). Median age was 9.1 years (range, 0.4-74.1) and twenty patients were diagnosed with de novo stage IV disease. Fifteen patients had favorble primary site and embryonal type was the most common pathological type with 31 (50%) cases. Thirty eight patients underwent surgery and sixty patients received chemotherapy. The most commonly used chemotherapy regimen was vincristine, Adriamycin, and cyclophosphamide (VAC). Excluding stage IV patients, median radiation dose was 45.0 Gy (Mean 44.6, range, 14.0-60.0). While median overall survival (OS) was not reached, 1-year OS was 93.5 percent and 3-year OS was 63.6 percent. Median progression free survival (PFS) was 18.7 months with 1- year PFS of 65.2 percent and 3-year PFS of 41.8 percent. Univariate analysis revealed age (p=0.025), site (p=0.03), TNM stage (p<0.001), and Intergroup rhabdomyosarcoma study clinical group (IRSG) (p=0.001) as significant prognostic factors. In multivariate analysis, age (p=0.002) and TNM stage (p<0.001) were statistically significant. After excluding stage IV patients who had had worst outcome and categorizing patients into four groups according to age and primary site, patients who were younger than 10 with favorable primary site had significantly better survival outcome. For futher investigation on primary site, we divided patients to five groups according to their primary sites as follows: non- parameningeal head and neck, parameningeal, genitourinary, extremity, and others, defined as the rest part of body such as retroperitoneum. 3-year OS was 90.0 percent in non-parameningeal head and neck patients, 57.1 percent in parameningeal patients, 87.1 percent in genitourinary subgroup, 30.0 percent in extremity and 45.5 percent in “others” subgroup. Respective 3-year PFS was 88.9 percent, 35.7 percent, 37.4 percent, 20.0 percent and 18.2 percent. Conclusion This study showed multimodality treatment including radiotherapy could result in acceptable survival outcomes especially in young patients whose primary sites are non- parameningeal head and neck or genitourinary. However, treatment strategies should be further improved to achieve better survival in unfavorable groups. EP-1633 IOERT in primary retroperitoneal sarcoma: a retrospective single center analysis of 69 cases F. Roeder 1 , I. Alldinger 2 , M. Uhl 3 , L. Saleh-Ebrahimi 1 , S. Schimmack 4 , G. Mechtersheimer 5 , M. Büchler 4 , J. Debus 3 , R. Krempien 6 , A. Ulrich 4 1 University Hospital LMU Munich, Radiation Oncology, Munich, Germany 2 CPZ Düsseldorf, Surgery, Düsseldorf, Germany 3 University of Heidelberg, Radiation Oncology, Heidelberg, Germany 4 University of Heidelberg, Surgery, Heidelberg, Germany 5 University of Heidelberg, Pathology, Heidelberg, Germany 6 Helios Clinic Berlin-Buch, Radiation Oncology, Berlin- Buch, Germany Purpose or Objective To report our experience with surgery and intraoperative electron radiation therapy (IOERT) with or without

external beam radiation therapy (EBRT) in patients with primary retroperitoneal soft-tissue sarcoma (RPS). Material and Methods We conducted a retrospective evaluation of patients with RPS who have been treated with IOERT at our institution since 1991. Patients treated for recurrent disease were excluded, leaving 69 patients for the final analysis. Median tumor size was 12.5 cm, most lesions were high grade (79%). Main histologies were as follows: well- differentiated liposarcoma 15%, leiomyosarcoma 26%, dedifferentiated liposarcoma 38% and others 22%. Surgery resulted in gross complete resection in 90%, but only 36% had microscopic negative margins despite multivisceral resections in 64%. All patients had IORT with a median dose of 12 Gy, 58 patients (84%) received additional perioperative EBRT (45% preop., 55% postop.) with a median dose of 45 Gy. Results Median follow-up for the entire cohort was 39 months and 54 months in surviving patients. Estimated 5-year-LC was 71%. In univariate analysis, LC was significantly associated with resection margin (5-year LC R0: 94%, R1: 52%, R2: 31%). Trends were present for gender, grading, and UICC stage. In multivariate analysis, only resection margin remained significant. Estimated 5-year OS was 63%. In univariate analysis, OS was significantly associated with grading, UICC stage, resection margin and timing of EBRT (5-year OS preop. 91% vs postop. 47% vs none 63%). In multivariate analysis, only timing of EBRT remained significant. Interestingly, neoadjuvant EBRT seemed superior while postoperative EBRT did result in even worse survival than no EBRT. Conclusion Combination of surgery, IOERT and EBRT resulted in very promising local control and survival. Complete resection remains the cornerstone of treatment as margin status was associated with local control and survival. EBRT should be preferably applied preoperatively. EP-1634 IOERT in locally recurrent high grade RPS: a retrospective single center analysis of 83 cases L. Saleh-Ebrahimi 1 , I. Alldinger 2 , M. Uhl 3 , S. Schimmack 4 , G. Mechtersheimer 5 , M. Büchler 4 , J. Debus 3 , R. Krempien 6 , A. Ulrich 4 , F. Roeder 1 1 University Hospital LMU Munich, Radiation Oncology, Munich, Germany 2 CPZ Düsseldorf, Surgery, Düsseldorf, Germany 3 University of Heidelberg, Radiation Oncology, Heidelberg, Germany 4 University of Heidelberg, Surgery, Heidelberg, Germany 5 University of Heidelberg, Pathology, Heidelberg, Germany 6 Helios Clinic Berlin-Buch, Radiation Oncology, Berlin- Buch, Germany Purpose or Objective To report our experience with surgery and intraoperative electron radiation therapy (IOERT) with or without external beam radiation therapy (EBRT) in patients with locally recurrent high-grade retroperitoneal soft-tissue sarcoma (RPS). Material and Methods We conducted a retrospective evaluation of patients with RPS who have been treated with IOERT at our institution since 1991. Patients treated for primary disease, suffering from low grade tumors or with distant metastases were excluded, leaving 83 patients for the final analysis. Median tumor size was 9 cm with undifferentiated liposarcoma being the dominating