Pencil Beam Scanning Proton Radiation Therapy Reduces Dose to the Planned Surgical Skin Flap and Uninvolved Bone in Preoperative Radiation Therapy for Soft Tissue Sarcomas of the Lower Extremity (original) (raw)

2017, International Journal of Radiation Oncology*Biology*Physics

Materials/Methods: From our institution's pathology database, 143 nonmetastatic non-breast angiosarcoma RAS diagnosed between 1994-2015 were identified. Pts were excluded for <2 year (yr) latency to RAS (nZ3) and <6 month follow-up (nZ57) to form a study cohort of 83 pts. Kaplan-Meier analyses were used to calculate overall survival (OS), progressionfree survival (PFS) and local recurrence-free rate (LRFR). The log rank test was used to compare time to event distributions. Results: Median age at diagnosis of primary malignancy was 43 yrs (1-74 yrs). The most common primary histologies were lymphoma (28%), breast cancer (27%) and prostate cancer (11%). The most common sites of primary RT were non-mantle chest/breast (33%), abdomen/pelvis (27%) and mantle (18%). RT was administered to the primary malignancy between 1938-2012. Median dose was 50.4 Gy (10.8 e 88.6 Gy). Chemotherapy (CT) was delivered to 61% (nZ51). RAS diagnoses occurred at a median age of 57 yrs (27-87 yrs). Median latency between primary malignancy and RAS was 17 yrs (2-60 yrs). The most common RAS histologies were unclassified spindle cell sarcoma (39%), undifferentiated pleomorphic sarcoma (22%) and leiomyosarcoma (13%). Median tumor size was 6 cm (range 1-22 cm). RAS treatment consisted of: surgery (S) alone (33%), S+RT (27%), S+CT (22%), S+CT+RT (13%); CT+RT (4%), and CT alone (2%). No pts received RT alone. Among the 77 who underwent surgery, margins were positive in 27%, negative for 65% and unknown in 8%. Median follow-up from RAS diagnosis was 2.5 yrs (0.6-16.6 yrs). Seven pts had locally progressive disease, 13 had local recurrence (LR), 15 had distance metastasis (DM), and 13 had both. There were 15 deaths, 14 due to RAS. Two-year OS, PFS and LRFR were 85%, 47% and 65%. Median survival and time to LR were 12.9 yrs (95% CI 7.9-not reached) and 9.5 yrs (95% CI 9.5-not reached). On univariate analysis, RAS > 5 cm was associated with worse OS (96% vs. 78%, pZ0.02). There was no relationship between RAS location and outcome. In terms of treatment: patients who underwent S had better 2-yr OS than those who did not (89% vs. 25%, p<0.0001). RT had no significant association with outcomes. Use of CT was associated with worse OS (72% vs. 95%, pZ0.0004) and PFS (28% vs. 60%, pZ0.0001), however, these groups were not balanced for prognostic factors. Conclusion: In this large, single institution cohort of RAS, excluding breast angiosarcoma, outcomes were generally poor with half of patients progressing within 2 years. Definitive resection was associated with better outcome, but favorable effects of RT or CT were not apparent. Further study is needed into improved therapeutic options for this heterogeneous patient population.