Breast
Brenna M. Murphy
Medical Student
Mayo Clinic
Rochester, Minnesota, United States
Disclosure: Disclosure information not submitted.
Borderline/malignant phyllodes tumors are rare breast neoplasms about which little is known regarding prognostic tumor features or the value of adjuvant therapies beyond surgical wide local excision with 1 cm margins. Our aim was to evaluate the association of patient and tumor factors with recurrence to better risk stratify these patients.
Methods:
With IRB approval, we identified 73 patients with borderline/malignant phyllodes tumors operated on at our institution 08/1971-02/2015. Patients with benign phyllodes or metastatic disease at presentation were excluded. All histopathology slides were reviewed by a specialist breast pathologist. Recurrence-free survival (RFS) and breast-cancer-specific survival (BCSS) were analyzed using the Kaplan-Meier method and Cox proportional hazards regression.
Results:
Median patient age was 46 years. Of the 73 tumors, 19 were classified as borderline and 54 as malignant. 36 tumors (49%) were ≥5 cm. 36 patients had breast-conserving surgery and 37 mastectomy; resection margin width was ≥1 cm in 63/73 (86%). After 96 months median follow-up, 24 recurrences were observed: 10 local only, 0 regional, 6 distant only and 8 local and distant. 15 patients died of disease. 5- and 10- year RFS were 70% (95% CI: 60-82%) and 62% (95% CI: 51-76%), respectively, while BCSS was 83% (95% CI: 75-93%) and 78% (95% CI: 68-89%). Of key tumor and histologic features including mitotic rate, atypia, stromal cellularity and stromal overgrowth, the latter (present in 26% of tumors) was most strongly associated with BCSS (HR 14.73, p< 0.001). In multivariable analysis, tumor size ≥5 cm and stromal overgrowth were most predictive of recurrence. Patients with either a tumor ≥5 cm or with stromal overgrowth had a greater than 5-fold increased risk of recurrence (HR 5.60, 95% CI: 1.66-18.83, p=0.005) and patients with tumors having both features had nearly a 15-fold increased risk (HR 14.85, 95% CI: 3.98-55.36, p< 0.001) compared to patients with neither tumor feature (Figure).
Conclusions: These data suggest that patients with larger tumors, particularly those exhibiting stromal overgrowth, may benefit from more aggressive treatment and close surveillance.