A 59-year-old woman presented with a tender, 2 × 2-cm, firm, mobile lump in the upper outer quadrant of her left breast, which had been present for several months and had recently increased in size. Her medical history was significant for fibrocystic disease of the breast, and several cysts had previously been aspirated. Fine-needle aspiration of the lump yielded benign ductal cells with associated myoepithelial cells, connective tissue, and foam cells. Mammography and ultrasonography demonstrated solid and cystic components to the lesion and infiltrative borders. An ultrasound-guided needle core biopsy showed florid ductal epithelial hyperplasia, stromal fibrosis, chronic inflammation, and fragments of cyst wall. Based on her age and the radiologic findings suggestive of neoplasm, excisional biopsy was performed. A 3 × 3 × 2-cm ill-defined mass with surrounding breast parenchyma harboring multiple cysts was excised. Grossly, the lesion was lobulated, had a firm to rubbery consistency, and had infiltrative borders. The cut surface had a white fibrous appearance. Microscopically, the lesion was composed of a spindle cell proliferation arranged in a fascicular growth pattern with rare nests of relatively bland epithelial cells demonstrating glandular and, focally, squamous differentiation (Figure 1 and inset). The fibroblastic-appearing cells infiltrated between ducts and lobules and extended to the periphery of the lesion. Only mild to focally moderate nuclear atypia was seen and mitoses were rare (Figure 2). The lesion was associated with a radial scar. The surrounding breast tissue exhibited proliferative fibrocystic change including florid ductal epithelial hyperplasia and sclerosing adenosis. Immunohistochemically, the spindle cells were focally to moderately reactive for keratins (AE1/3, CAM 5.2, and cytokeratin 5/6), smooth muscle actin (Figure 3), S100 protein, and CD10. The spindle cells also displayed nuclear expression of p63 (Figure 4). Results of other markers were negative including β-catenin, CD34, HMB-45, Melan A, and estrogen receptor protein.
What is your diagnosis?
Pathologic Diagnosis: Low-Grade (Fibromatosis-like) Spindle Cell Carcinoma of the Breast
Abstract
Low-grade fibromatosis-like spindle cell carcinoma (FLSCC) of the breast is a variant of metaplastic carcinoma that has recently been recognized as a distinct entity because of its resemblance to fibromatosis and similar propensity for local recurrence. The major lesional component of FLSCC consists of spindle cells, which can have a deceptively bland appearance. Less than 5% of the tumor is composed of epithelial elements or carcinoma. FLSCC can pose diagnostic difficulty, and the differential diagnosis includes both benign and malignant spindle cell breast lesions. Identifying areas of focal epithelial differentiation and the application of an immunohistochemical panel including keratins, myoepithelial markers, and β-catenin are integral for establishing the diagnosis.
Low-grade (fibromatosis-like) spindle cell carcinoma (FLSCC) of the breast is a variant of metaplastic carcinoma that has recently been recognized as a distinct entity because of its resemblance to fibromatosis and similar propensity for local recurrence. In fact, some authors have chosen to avoid the term carcinoma, preferring fibromatosis-like tumor to emphasize its dominant phenotype (the major lesional component consists of spindle cells, whereas less than 5% of the tumor is composed of epithelial elements or carcinoma) and its biologic behavior.1 FLSCC has only been reported in women, with a mean age of 64 years (range, 40–85 years) at diagnosis. Tumors have ranged in size from 1 to 7 cm, and infiltrative margins are grossly present in two thirds of cases.2 Histologically, these tumors are composed of deceptively bland spindle cells arranged in intersecting fascicles of variable length and exhibiting low mitotic activity. A characteristic feature of FLSCC is the presence of epithelioid cells arranged in small cohesive aggregates or abortive tubules that are immunoreactive for keratin.2
A recent study demonstrated nuclear p63 immunoreactivity in 13 of 15 metaplastic carcinomas including all 10 examples of the spindle cell type, whereas only 1 of 173 cases of typical invasive mammary carcinomas (not otherwise specified) expressed p63.3 Phyllodes tumors and primary breast sarcomas were also found negative for p63. Recent studies have shown myoepithelial differentiation in spindle cell (metaplastic) carcinomas4; therefore, it is not surprising that the spindled (metaplastic) element of FLSCC would similarly express this pattern of staining.
The clinical behavior of FLSCC further warrants classification as a distinct entity. Neither regional nor distant metastases were demonstrated in the initial series by Gobbi et al1 or on follow-up.5 However, Sneige et al6 observed that these tumors can spread to distant sites. Two patients with larger tumors (4–5 cm) developed lung metastases and died of disease within 2 years of diagnosis. None of the patients in either of these series who underwent axillary lymph node dissection were found to have nodal disease, suggesting that axillary lymph node sampling may not be necessary.
The core biopsy performed in this case contained areas of bland-appearing fibrosis adjacent to a cyst wall and a diagnosis of fibrocystic change was rendered. In retrospect, it is not clear whether lesional cells were present on this small sample. The diagnosis of FLSCC was not established until the excisional biopsy specimen was examined.
It is important to distinguish FLSCC from other predominately spindle cell lesions of the breast. Benign lesions of the breast that mimic FLSCC include mammary fibromatosis, pseudoangiomatous stromal hyperplasia, myofibroblastoma, solitary fibrous tumor, and nodular fasciitis. In general, immunohistochemical staining for keratins should be performed whenever the diagnosis of a presumably benign spindle cell lesion of the breast is in question to exclude the possibility of FLSCC.
Pseudoangiomatous stromal hyperplasia, which can mimic FLSCC by virtue of its bland spindle cell component, is characterized by empty, slitlike, anastomosing spaces lined by fibroblast-like cells set in a dense collagenous stroma. Myofibroblastoma is composed of short haphazardly arranged fascicles of spindle cells with eosinophilic cytoplasm partitioned by collagen bands. Solitary fibrous tumor displays multipatterned architecture including a “patternless” growth pattern of short spindled cells in association with a hemangiopericytomatous vascular element and variable amounts of collagen. Nodular fasciitis has only rarely been described in the breast and early lesions resemble florid granulation tissue with a random growth pattern of reactive-appearing spindled cells. Most lesions become less cellular and more collagenous or cystic over time.2
Primary mammary fibromatosis is the most important differential diagnostic consideration in our case, given its close histologic overlap with FLSCC. Immunohistochemical staining for β-catenin is very helpful in resolving this diagnostic consideration. Most cases of primary mammary fibromatosis show nuclear expression of β-catenin, whereas FLSCC and other spindle cell carcinomas should not exhibit nuclear β-catenin expression.7
Malignant entities in the differential diagnosis include myoepithelial carcinoma, melanoma, and fibrosarcoma. Fibrosarcoma is more cellular and has a characteristic “herringbone” growth pattern of convincingly atypical spindle cells.8 Because immunoreactivity for HMB-45 and Melan A can be negative in spindle cell melanoma, the presence of strong, diffuse immunoreactivity for S100 protein would support that diagnosis over FLSCC, in which S100 protein is expression more focal. The spindle cells of pure myoepithelial carcinoma have more abundant cytoplasm than the cells of FLSCC and lack the characteristic small epithelial aggregates.9
Complex sclerosing lesions can mimic malignancy mammographically and additionally spindle cell carcinomas can be associated with radial scar.10 However, the stroma of FLSCC is more cellular and differs from the paucicellular fibroelastotic stroma that characterizes typical radial scar. The tumor in our case was associated with a radial scar.
In conclusion, FLSCC is a locally aggressive spindle cell lesion of the breast with the potential for distant metastasis. FLSCC should be distinguished from other benign and malignant spindle cell lesions of the breast. Identifying areas of focal epithelial differentiation and application of an immunohistochemical panel including keratins, p63, and β-catenin are integral for establishing the diagnosis of FLSCC.
References
The authors have no relevant financial interest in the products or companies described in this article.