ABSTRACT Phyllodes tumors (PTs) are uncommon biphasic breast tumors that usually occur in adult females. They are composed of a benign epithelial component and a cellular, spindle cell stroma forming a leaf-like structure. No one morphologic finding is reliable in predicting the clinical behavior of the tumor. The purpose of this study was to explore the clinicopathologic factors associated with outcome and metastasis.
We retrospectively reviewed the records of 172 patients seen at Mackay Memorial Hospital from January 1985 to December 2003. Clinical data analyzed included age, presenting symptoms and signs, tumor size, location, type of surgery, time to recurrence, and metastasis. The clinicopathologic factors associated with outcome and metastasis were analyzed statistically using the chi-square test with Yate correction.
The mean follow-up was 71 months (range 7-237). The mean age was 37 years (range 11-73). The majority of tumors were found in the upper outer quadrant (46.0%), with an equal propensity to occur in either breast (48.8% vs. 50.0%). The pathologic diagnoses included 131 benign, 12 borderline, and 29 malignant lesions. Nineteen patients (11%) had a recurrence and three (1.7%) had metastases. The initial diagnosis of all 19 recurrent tumors were benign. Age, surgical approach, mitotic activity, and surgical margin were significantly correlated with recurrence (P = 0.029, 0.020, 0.048, and 0.00018, respectively). Stromal cellularity, stromal overgrowth, stromal atypia, mitotic activity, tumor margin, and heterologous stromal elements were significantly correlated with metastases (P = 0.032, 0.00008, 0.000002, 0.004, 0.005, and 0.046, respectively). Mammography and breast echo were not reliable for differentiating PTs from fibroadenomas (6.9% vs. 37.9% and 3.3% vs. 45%, respectively). Frozen section was of limited value (41.6%). The role of adjuvant radiotherapy and chemotherapy remains to be defined. Local excision, wide excision, or mastectomy with negative surgical margins yielded high local control rates (88.7%, 88.2%, and 100%, respectively), but local excision was associated with a relatively high percentage of positive surgical margins (18.3%). A total of 42 modified radical mastectomies were performed. The reasons for these procedures included a diagnosis of malignancy on frozen section or because the tumors were so large, they were assumed to be carcinomas. No axillary lymph node metastases were found. Fifteen patients in our series had tumors with infiltrating tumor margin, severe stromal overgrowth, atypia, and cellularity. All three patients with metastases (3/15) were in this group. Presence of metastases was significantly correlated with this group (P = 0.0000038).
Wide excision with a clear margin may be the preferable initial therapy, even for malignant PTs. Routine axillary lymph node dissection is not recommended. Patients have tumors with infiltrating tumor margin, severe stromal overgrowth, atypia, and cellularity are at high risk for metastases.