Management of phyllodes breast tumors.

Phyllodes tumors are a rare distinctive fibroepithelial tumors of the breast and their management continues to be questioned. The aim of our study was to examine the treatment and outcome of 165 patients with phyllodes tumors and to review the options for surgical management. This is a retrospective study of 165 patients who presented to the Institut Curie between January 1994 and November 2008 for benign, borderline or malignant phyllodes tumors. The median follow-up was 12.65 months [range 0-149.8]. The median age at diagnosis was 44 years [range 17-79]. One hundred and sixty patients (97%) had breast-conserving treatment, of whom 3 patients (1.8%) had oncoplastic breast surgery. Younger women had a significantly higher chance of having a benign phyllodes tumor (p = 0.0001) or a tumor of small size (p < 0.0001). Histologic examination showed 114 benign (69%), 37 borderline (22%) and 14 malignant tumors (9%). The median tumor size was 30 mm [range 5-150]. The tumor margins were considered incomplete (< 10 mm) in 46 out of 165 cases (28%) with 52% revision surgery. Only the tumor grade was a significant risk factor for incomplete tumor margins (p = 0.005). Fifteen patients developed local recurrence (10%) and two, metastases. In univariate analysis, the histologic grade (p = 0.008), and tumor size (p = 0.02) were significative risk factors for local recurrence with an accentuated risk for "borderline" tumors and tumors of large size.).Similar results were obtained using multivariate analysis (p = 0.07). The mainstay of treatment for phyllodes tumors remains excision with a safe surgical margin, taking advantage breast conserving surgery where amenable. For borderline or malignant phyllodes tumors or in cases of local tumor recurrence, mastectomy, and immediate breast reconstruction may become the preferred option. Genetic analysis will potentially supplement classical histologic examination in order to improve our management of these tumors. The role of adjuvant treatments is unproven and must be considered on a case-by-case basis.