BACKGROUND: Sarcomas are a rare complication of radiotherapy for breast carcinoma and patients have a poor prognosis. The incidence, histology, and management of patients with sarcomas were reviewed in the current study.METHODS: The authors reviewed the records of 16,705 patients with breast carcinoma. Of these, 13,472 (81%) were treated with megavoltage radiotherapy and 3233 were treated without at the Institute Curie (Paris, France) between 1981 and 1997. Median doses of 50–55 grays (Gy) in 25–27 fractions were delivered to the whole breast over a period of 5–5.5 weeks (2 Gy/day, 5 weekly fractions) followed, when indicated, by a 16–26‐Gy boost to the tumor or tumor bed. Treatment of radiation‐induced sarcomas (RIS) consisted mostly of radical surgery and chemotherapy.RESULTS: Overall, 35 patients developed sarcomas. Of these, 27 fulfilled the Cahan criteria. The median follow‐up was 9.3 years (range, 1–22.4 years). The latency period ranged from 3 years to 20.3 years. Thirteen sarcomas were located in the breast, 5– in the chest wall, 3 in the sternum, 2 in the supraclavicle, 1 in the scapula, and 3 in the axilla. Histologic evaluation identified 13 angiosarcomas, 3 osteosarcomas, 5 undifferentiated sarcomas, 1 malignant fibrous histiocytoma, 2 leiomyosarcomas, 1 fibrosarcoma, 1 rhabdomyosarcoma, and 1 myosarcoma. The cumulative RIS incidence was 0.07% (± 0.02) at 5 years, 0.27% (± 0.05) at 10 years, and 0.48% (± 0.11) at 15 years. Standardized incidence ratios were 10.2 (95% confidence interval, 9.03–11.59) for irradiated patients and 1.3 (0.3–3.6) for nonirradiated patients. Of the 27 patients, 15 died of sarcoma within 1 month to 14.5 years (mean, 34.2 ± 0.7 months). The 5‐year actuarial survival rate after diagnosis of RIS was 36% (± 0.11).CONCLUSIONS: The current study confirmed the rarity of RIS. However, it showed that the risk increased with time. Therefore, careful, long‐term follow‐up of patients treated with radiotherapy is needed for early detection and efficacious treatment of these malignancies. Cancer 2005. © 2005 American Cancer Society.

Treatment for non-metastatic breast cancer (BC) may be the cause of second malignancies in long-term survivors. Our aim was to investigate whether survivors present a higher risk of malignancy than the general population according to treatment received. We analysed data for 16 705 BC survivors treated at the Curie Institute (1981–1997) by either chemotherapy (various regimens), radiotherapy (high-energy photons from a ⁶⁰Co unit or linear accelerator) and/or hormone therapy (2–5 years of tamoxifen). We calculated age-standardized incidence ratios (SIRs) for each malignancy, using data for the general French population from five regional registries. At a median follow-up 10.5 years, 709 patients had developed a second malignancy. The greatest increases in risk were for leukaemia (SIR: 2.07 (1.52–2.75)), ovarian cancer (SIR: 1.6 (1.27–2.04)) and gynaecological (cervical/endometrial) cancer (SIR: 1.6 (1.34–1.89); P<0.0001). The SIR for gastrointestinal cancer, the most common malignancy, was 0.82 (0.70–0.95; P<0.007). The increase in leukaemia was most strongly related to chemotherapy and that in gynaecological cancers to hormone therapy. Radiotherapy alone also had a significant, although lesser, effect on leukaemia and gynaecological cancer incidence. The increased risk of sarcomas and lung cancer was attributed to radiotherapy. No increased risk was observed for malignant melanoma, lymphoma, genitourinary, thyroid or head and neck cancer. There is a significantly increased risk of several kinds of second malignancy in women treated for BC, compared with the general population. This increase may be related to adjuvant treatment in some cases. However, the absolute risk is small.