Image of the Month—Diagnosis FREE

Originally described by Johannes Muller in 1838, phyllodes tumors comprise less than 1% of all breast tumors.¹Although they closely resemble fibroadenomas, they are distinguished by larger size, rapid growth, and late occurrence. The overlying skin may show dilated veins, bluish discoloration, and, rarely, pressure necrosis, although frank ulceration is uncommon.²Rarely, pressure necrosis of the overlying skin can occur with a huge benign phyllodes tumor,³more so in a small breast. This, coupled with the growth stimulus of pregnancy, can cause extrusion of the tumor through the skin, as was seen in the present case. When it extrudes out of the skin, the exophytic lump can be mistaken for an aggressive malignant tumor. About 40% manifest estrogen receptors, and almost 100% are progesterone receptor positive,⁴which explains their rapid growth during pregnancy. Mammographic features resemble those of a fibroadenoma. On fine-needle aspiration cytological examination, cohesive stromal cells, mesenchymal cells, clusters of hyperplastic ductal cells, foreign body giant cells, bipolar naked nuclei, and absence of apocrine metaplasia are highly suggestive of phyllodes tumor.⁵On histological examination, hypercellular stroma is seen to project into epithelial-lined cystic spaces. The stroma is the neoplastic component and determines the behavior of the tumor. Based on the margin, infiltration, tumor necrosis, stromal cellularity, atypia, and number of mitoses per high-power field, these can be classified as benign, borderline, or malignant.⁶

It is yet unsettled whether a phyllodes tumor arises from fibroadenoma or de novo. It has been proposed that a somatic mutation in a fibroadenoma may result in monoclonal selection and propagation to a phyllodes tumor, conferring the ability to recur locally and metastasize.⁷Stromal growth may be induced by growth factors released from the breast epithelium. One of these, endothelin-1, a breast fibroblast stimulator, is elevated in phyllodes tumor, suggesting paracrine control.⁸Studies indicate that p53 expression, high DNA content and ploidy, a high proliferation index, and a poor histological grade are associated with a bad prognosis.⁹Telomerase, a ribonucleoprotein enzyme, also has been proposed as a useful prognostic marker.¹⁰

The mainstay of treatment is surgery. Mastectomy is no longer justified because there is no survival benefit. Excision with a 1- to 2-cm margin, which is histologically negative for disease, is sufficient for cure, even for the malignant tumors. Simple intracapsular enucleation increases local recurrence and is not recommended. It is unclear whether a surveillance policy or a re-excision is required for a benign phyllodes tumor with margins positive for disease.¹¹Because lymph node spread occurs in 10% of malignant tumors, and distant metastasis is mainly blood borne, routine lymphadenectomy is not recommended. The role of adjuvant radiotherapy is questionable and is not practiced widely. The 5-year disease-free survival rates for benign, borderline, and malignant phyllodes tumors are 90%, 74%, and 66%, respectively.¹²Local recurrence occurs in 15% of cases, and the risk is highest with incomplete excision.¹³A re-excision is adequate for a benign recurrence. In case of a malignant recurrence, mastectomy should be considered. Recurrent disease has no survival disadvantage or relation with distant metastasis. Of malignant tumors, 20% show metastatic disease.¹⁴Tumor size, stromal overgrowth, necrosis, and histological grade are directly proportional to the metastatic potential.¹⁵The most common sites are the lung, bone, and abdominal viscera. The prognosis is poor, and no long-term survival is reported. Combination chemotherapy has been used, with varying results. The role of hormone therapy, an exciting approach in view of the hormonal sensitivity of these tumors, remains to be defined.

Correspondence:Somprakas Basu, MS, Department of General Surgery, Institute of Medical Sciences, Banaras Hindu University, Varanasi-221005, India (sombasu@hotmail.com).

Accepted for Publication:November 8, 2006.

Author Contributions:Study concept and design: Basu and Sharma. Acquisition of data: Basu, Saxena, Sharma, and Kumar. Analysis and interpretation of data: Basu, Sharma, and Kumar. Drafting of the manuscript: Basu and Saxena. Critical revision of the manuscript for important intellectual content: Basu, Sharma, and Kumar. Administrative, technical, and material support: Basu, Saxena, Sharma, and Kumar. Study supervision: Basu, Sharma, and Kumar.

Financial Disclosure:None reported.