Prognosis of Radiation Associated Bone and Soft-tissue Sarcomas

Abstract

To determine the prognosis of radiation associated soft tissue and bone sarcomas arising within the radiation treatment field of a prior malignancy. A total of 9,069 soft tissue and bone sarcoma patients were seen at our institution from 1938 to 2011. We identified 64 radiation associated soft tissue or bone sarcomas (RAS) arising in the radiation treatment fields of prior malignancies. RAS was defined as having a history of radiation treatment at least 1 year prior to the diagnosis of the sarcoma, development within the prior radiation field, and confirmed histology differing from the prior malignancy. Only patients who received radiation at our institution with radiation fields available for review were included; patients treated elsewhere for the prior malignancy, had simultaneous sarcoma and another malignancy, or had no follow-up were excluded. Clinicopathologic characteristics, treatment of the RAS, and survival outcomes were analyzed and compared to sporadic soft tissue and bone sarcomas. Multivariate analysis was performed using Cox proportional hazards. The median follow-up for this cohort of 64 RAS was 21 months (range 1 to 234 months). The median age at the time of prior radiation therapy was 61 years (range 1 to 85 years). The median interval between prior radiation therapy and the development of RAS was 10 years (range 1 to 37 years). The histological subtypes of RAS included MFH (12), fibrosarcoma or fibromyxosarcoma (11), angiosarcoma (9), osteosarcoma (9), leiomyosarcoma (5), adipocytic sarcoma (4, including liposarcoma and myxoid liposarcoma), and others (14). For cases where grading was available, 26 had grade 3/4, 15 had grade 1, and 2 had grade 1. The prior treatment sites included breast (17), head and neck (12), genitourinary (9), CNS (6), lymph nodes (5), thyroid (4), gynecological (4), connective tissue (3), GI (2) retina (1), and others (6). The histology of prior malignancies included 42 carcinomas, 14 hematological malignancies (4 HL, 9 NHL, 1 leukemia), 5 glial/neural, 1 retinoblastoma, 1 Wilms tumor and 1 neuroblastoma. 43 patients received only radiation for the prior malignancy while 21 received both chemotherapy and radiation. Twenty-four RAS were reirradiated, 13 treated with chemotherapy, and 44 surgically resected. Five-year overall survival was 41% +/- 13% for RAS compared with 65% +/- 1% for non-RAS. Multivariate analysis shows that RAS had worse overall survival (HR = 2.67, p < 0.01) and disease-specific survival (HR = 3.32, p < 0.01) independent of size, grade, margin, and histology. Radiation associated bone and soft tissue sarcomas arising within the radiation treatment field of a prior malignancy appear to have worse prognosis than spontaneous sarcomas and warrant further characterization to identify potential targets to improve outcomes.