(P 068) CLINICO-DEMOGRAPHIC CHARACTERISTICS AND OUTCOMES OF RADIATION-INDUCED SARCOMAS: A CANADIAN SARCOMA RESEARCH AND CLINICAL COLLABORATION (CANSARCC) STUDY

Objective: To characterize the clinical and demographic profile of radiation-induced sarcomas (RIS) in a cohort of patients (pts) available in the Canadian Sarcoma Research and Clinical Collaboration (CanSaRCC) database.

Methods: We defined RIS as sarcomas arising within a previously irradiated site following a latency period, whose histology was different from that of the index tumor. After ethics board approval, we retrospectively abstracted clinicopathologic data available in the CanSaRCC database comprising pts diagnosed with RIS between 1996-2021 and treated in three institutions, including: age at radiation therapy (RT) for the first malignancy, age at RIS diagnosis, time between RT and development of RIS, dose, fractionation/site of RT, as well as survival outcomes. Overall survival (OS) was defined as time from RIS diagnosis to death from any cause. Univariate (UVA) and multivariate (MVA) Cox regression models were computed for variables correlating to outcomes. OS and RFS were estimated using the Kaplan-Meier method and log-rank test. T-test, ANOVA and Wilcoxon rank were used to compare means and medians between groups. Chi-square test and Fisher’s test were used to compare demographic features. A p-value< 0.05 was considered statistically significant.

Results: 105 pts met criteria for RIS: breast angiosarcoma (BAS, n=54), osteosarcoma (OST, n=12), and other STS (n=39). The most common histologic subtypes among STS were undifferentiated pleomorphic sarcoma (n=9, 23%) and liposarcoma (n=6, 15%); 3.7% (2/56) of the angiosarcomas developed outside the breast. The most frequent RIS site was breast (n=50, 47.6%), followed by abdomen/pelvis/retroperitoneum (n=16, 15.2%, [41% of other STS])(p< 0.0001). Median age of RIS diagnosis was 75.5 (47-89.6), 45.5 (18.5-74), and 62 (19.1-85.3) years, respectively (p< 0.0001). Regarding the time (years) between RT and RIS, the shortest interval was for BAS (median: 8.58, 3.8-27), followed by OST (median: 17.5, 5.2-29.3), and other STS (median: 18.5, 1.3-57)(p=0.0003). The median dose (Gy) and the number of fractions prescribed to the initial malignancy are described in Table 1. Breast carcinoma was the most common primary cancer (n=63, 60%), followed by other non-sarcoma cancer subtypes (n=34, 32%)(Table 1). Eighty-three percent of RIS pts underwent surgery and 79.5% achieved R0 resections, with 44% receiving RT. Among 47 pts who had systemic treatment, 27 (57.4%) received neoadjuvant chemotherapy; 33% non-BAS (n=3/9) had a tumor shrinkage compared with 66% BAS (n=12/18)(p=0.10)(Table 2). The median OS and RFS was 51 (95%CI 34-101) and 21 (95%CI 14-39) months, respectively. Median RFS for BAS, OST, and STS was 18 (95%CI 9.3-26.7), 29 (95%CI 1.65-56.34), and 28 months (95%CI 2.18-53.82), respectively (p=0.76)(Figure 1); 5-year RFS was 36%, 25%, and 36.2%. The median OS for BAS, OST, and STS was 51 (95%CI 0-108.5), 37 (95%CI 17-57.1), and 57 (95%CI 21-93.2) months (p=0.32)(Figure 2); 5-year OS was 49.3%, 27.5%, and 41.5%. By MVA, neither RIS subtype nor RIS RT field influenced RFS; however, worse OS outcomes were observed in RIS arising in extremities/axillae (p=0.035).

Conclusion: RIS are highly heterogenous entities with overall reserved prognosis. In our cohort, BAS (followed by other STS) was the most common RIS, with the shortest time between RT to RIS development. Surgery remained the main treatment modality for RIS, with a possible role of neoadjuvant modalities in large lesions. Despite inherent limitations associated with the heterogeneity of these malignancies, collaborative efforts such as CanSaRCC can be useful to better understand their clinico-demographic characteristics and treatment outcomes. This study will be further expanded to include data from other institutions under the CanSaRCC consortium agreement.

Figure 1.docx

Figure 2.docx

Table 1.docx

Table 2.docx