ABSTRACT FINAL ID: 204; TITLE: Local Control with Pre-operative vs Post-Operative External Beam Radiotherapy in Extremity Soft-Tissue Sarcoma: 5 Year Results of a Prospective Randomized Phase III Trial AUTHORS (ALL): O'Sullivan, Brian 1 ; Davis, Aileen 2 ; Turcotte, Robert 5, 4 ; Bell, Robert 3, 8 ; Catton

BODY: Purpose/Objective: External beam pre-operative (pre-op) or post-operative (post-op) radiotherapy (RT) use is frequent for extremity soft tissue sarcoma. Although preferential referral of larger complex lesions for pre-op RT has existed, high local control rates are maintained with this approach prompting some authors to consider that it may be more effective. Moreover, both approaches have different dose volume considerations and different toxicities that may influence treatment selection. For these reasons controversy has existed about which method is preferred. We report the 5-year local control results of a Canadian multi- center prospective randomized trial that compared pre- and post-op RT and was originally designed to address early wound complications (reported previously). Materials/Methods: The decision that combined RT and surgery were necessary was made by joint consultation by a surgeon and radiation oncologist prior to trial registration. After stratification by tumor size dichotomized at 10 cm, adult subjects were randomized to pre-op (50 Gy in 25 fractions) or surgery followed by post-op RT (66 Gy in 33 fractions). Local control was prospectively evaluated according to several Host factors (gender and age), Tumor factors (depth, size, grade, upper vs lower extremity, and intra- vs extra-compartmental status), and Environmental/Treatment factors (RT timing: pre-op vs post-op RT; resection margin status; treatment center; and presentation with primary vs recurrent status). Center was assessed by comparing the Princess Margaret Hospital (PMH) in Toronto (where approximately 50% of cases were treated) compared to all other centers. Cox modeling controlled for the influence of these factors on local control. Results: 94 patients were randomized to pre-op and 96 to post-op; 52 in each group were treated at PMH. Resection margins were positive in 16% (pre-op) and 14% (post-op) and other factors were similarly equally distributed including low grade (10%) and recurrent status (16%). The 5-year local control rate was 93% (95% CI: 88-98) for pre-op and 92% (95% CI: 87-98) for post-op RT. Local control by margin status was 96% (clear margins) compared to 77% (positive margins) with pre-op RT and was similar to post-op RT: 95% (clear) vs 73% (positive). Other predictive factors had minimal effect on local control though in post-op RT, age <50 years and intra-compartmental had rates of 98% vs older patients and extra- compartmental status where the rates were 87% for each. For pre-op, these factors provided identical local control of 93%. The only factor of significance on multivariate analysis was the resection margin status (see Table). Conclusions: Although not primarily designed to address this end-point, this trial strongly suggests that identical high rates of local control for pre-op and post-op RT are achieved using the RT protocols on the study. Because of this, the choice of RT approach in soft tissue sarcoma should also consider other factors that include the influence of different RT doses and treatment volumes that may have different normal