D.N. Burner1, K. Crowell2,3, A. Khattab1, C.C. Huang4, M.M. Diab5, J.E. Grilley-Olson2,6, J.K. Plichta2,5, D.G. Kirsch7,8,9, L.H. Rosenberger2,5 1Duke University School of Medicine, Durham, NC, USA 2Duke Cancer Institute, Durham, NC, USA 3Duke University, Biostatistics And Bioinformatics, Durham, NC, USA 4Duke University Medical Center, Radiation Oncology, Durham, NC, USA 5Duke University Medical Center, Surgery, Durham, NC, USA 6Duke University Medical Center, Medicine, Durham, NC, USA 7University of Toronto, Medical Biophysics, Toronto, Ontario, Canada 8University Health Network, Radiation Medicine Program, Princess Margaret Cancer Centre, Toronto, Ontario, Canada 9University of Toronto, Radiation Oncology, Toronto, Ontario, Canada
Introduction: Angiosarcoma of the breast occurs sporadically (primary angiosarcoma, AS) or after radiation therapy (secondary, radiation-associated angiosarcoma, RAAS). Although studies show that surgery and radiation therapy (RT) are associated with improved outcomes, optimal treatment is unknown. We aimed to evaluate the association between types of therapies and treatment order with outcomes.
Methods: We identified adult female cases of breast angiosarcoma (AS, RAAS) from our center (1999-2024). Patient, tumor, and treatment factors were summarized and compared. The Kaplan Meier (KM) method was used to estimate unadjusted overall survival (OS), and log rank tests for differences in OS. Cox proportional hazards models were used to explore associations of factors with OS, after adjustment.
Results: We identified 55 cases; n=11 AS, n=44 RAAS. Median age differed by type (42 vs. 69 years, AS vs. RAAS, p=<0.001). AS frequently presented as a mass (90.9% vs. 18.2% in RAAS), while RAAS frequently presented with skin changes (77.3% vs. 9.1% in AS), both p=<0.001. All cases of RAAS had previous breast cancer treated with adjuvant RT (median time to RAAS, 7.8mo).
All cases of AS had surgery: 81.8% (n=9/11) as mastectomy and 18.2% (n=2/11) as breast conservation. A third of women (36.4%) with AS received chemotherapy and the majority (81.8%, n=9/11) received RT; 9.1% (N=1/11) neoadjuvant RT and 72.7% (n=8/11) adjuvant RT. Local recurrence (LR) occurred less frequently than distant metastases [9.1% (n=1/11) vs. 45.5% (n=5/11)]. 27.2% (n=3/11) were deceased.
A majority of RAAS had surgery: 79.5% (n=35/44) as mastectomy and 9.1% (n=4/44) as breast conservation. A quarter of women (27.3%) with RAAS received chemotherapy and half (54.5%) received additional RT; 15.9% (n=7/44) as neoadjuvant RT and 38.6% (n=17/44) as adjuvant. LR was more frequent than distant metastases [38.6% (n=17/44) vs. 25.0% (n=11/44)]. 25.0% (n=11/44) were deceased.
Unadjusted KM analysis showed no difference in PFS (p=0.29) or OS (p=0.54) between types. (Figure 1) For patients with RAAS, the multivariable analysis showed no association with the relative combination of treatments, nor sequencing, with OS.
Conclusion: Though AS and RAAS cohorts had similar rates of surgery ± chemotherapy with similar PFS and OS, there was a divergence in receipt of RT and LR. While nearly all AS patients received RT, only half of RAAS did. Omission of RT in RAAS treatment may be driven by lack of guidelines, concerns by physicians for toxicity from re-irradiation, and concern by patients about further RT in the setting of a radiation-associated malignancy. Significantly higher LR rates in RAAS may be due to this lower uptake of RT. Ongoing education may improve these outcomes.