[Primary breast sarcomas: About 30 cases treated at Salah-Azaiez institute in Tunisia]. / Sarcomes mammaires primitifs : à propos de 30 cas traités à l'institut Salah-Azaiez de Tunis.

PURPOSE: To identify retrospectively prognostic factors of primary breast sarcoma and review its treatment modalities. MATERIALS AND METHODS: This is a descriptive study on 30 cases of primary breast sarcoma. We carried out a univariate and multivariate analysis correlating clinical, pathological and therapeutic parameters with disease-free survival and overall survival. RESULTS: The mean age was 46.8 years. The mean tumour size was 10cm. The 30 cases were 18 phyllodes sarcomas, eight angiosarcomas, three liposarcomas and a case of granulocytic sarcoma. Sixteen patients had adjuvant radiotherapy and only seven patients received adjuvant chemotherapy. The median follow-up was 64 months. Overall survival rates at 3 and 5 years were 49.1% and 33.7%. Disease-free survival rates at 3 and 5 years were 22.8% and 15.2% respectively. The analytical study of the following parameters: tumour size and presence or absence of node or distant metastases, showed no correlation with overall survival nor with disease-free survival. Furthermore, adjuvant radiotherapy did not improve overall survival (P=0.298; hazard ratio [HR]=1 [0.982-1.04]) nor disease-free survival (P=0.61; HR=0.942 [0.862-1.029]). By univariate analyses, we identified a correlation between overall survival, surgical margins (>1cm) (P=0005; HR=3.4 [1.217-9.919]) and tumour necrosis (P=0.028; HR=0.099 [0.014-0.682]). We did not find any independent prognostic factor by multivariate analysis. CONCLUSION: The prognosis of primary breast sarcoma seems to depend essentially on optimal surgical excision (margin over 1cm). The only potential histological parameter correlated with the prognosis is the presence of tumour necrosis. The histological subtype should not be considered as a prognostic marker for overall or disease-free survival in patients with primary breast sarcoma.