Cutaneous leiomyosarcomas are classified as dermal, arising from the arrectores pilorum muscle, or subcutaneous, arising from the blood vessel wall [1
]. Several histologic variants of the cutaneous leiomyosarcomas, such as, epithelioid [2
] and granular cell type [3
] have been reported. However, we were unable to find any reported case of a leiomyosarcoma of the skin where osteoclast-like giant cells were present in the tumor. Leiomyosarcomas with osteoclast-like giant cells arising in non-cutaneous locations, such as, uterus and rectum have been reported in the literature [4
]. It can morphologically be confused with atypical fibroxanthoma (AFX)/cutaneous malignant fibrous histiocytoma (MFH), malignant melanoma, Kaposi sarcoma, spindle cell carcinoma, epithelioid angiosarcoma, and malignant peripheral nerve sheath tumor. The most difficult diagnostic dilemma for such a tumor is to distinguish it from an AFX/MFH. The histologic features and the immunohistochemical profile may be somewhat similar. However, the AFX/MFH typically occurs in the upper dermis of the sun-exposed skin, usually in the head and neck. It is not associated with pilar muscle or blood vessels. Immunohistochemically, the cells in AFX/MFH can be focally positive to smooth muscle actin, however, a strongly positive diffuse pattern is unusual. A subset of leiomyosarcomas is thought to arise from undifferentiated mesenchymal cells which may acquire smooth muscle features. However, the spindle cells and the giant cells of such a tumor are usually positive for CD68.
The tumor in our case clearly arises from muscle wall of an artery (Fig. ) with histologic features of a leiomyosarcoma, including spindle cells with eosinophilic cytoplasm, oval or cigar-shaped nucleoli with paranucleolar vacuoles and immunologic feature of strongly SMA-positive tumor cells. On the other hand, the AFX/MFH is composed of fibroblastic cells with CD68 positivity. The predominant spindle cells of our tumor were negative for CD68. The only CD68-positive cells in the tumor were the osteoclast-like giant cells. We believe that our case represents a cutaneous leiomyosarcoma with reactive osteoclast-like giant cells based on the demonstration of origin from the arterial wall and histologic and immunologic evidence.
In addition to leiomyosarcoma, osteoclast-like giant cells have been noted in carcinomas of pancreas and liver [7
]. The origin and nature of the osteoclast-like giant cells in various malignant tumors has remained controversial. However, most of the authors believe that the giant cells are of histiocytic origin and are reactive in nature. Features suggesting their benign nature include: bland appearance identical to osteoclasts in osteoclastoma, different immunostaining patterns from the malignant spindle cells, and no proliferating evidence, such as non-immunoreactivity to Ki67 [4
The prognostic significance of osteoclast-like giant cells in cutaneous leiomyosarcoma is unknown. Dermal leiomyosarcomas are frequently recurrent, but almost never metastatic [8
]. Conversely, subcutaneous leiomyosarcomas behave similar to those arising within deep soft tissue with frequent local recurrences and as much as 50% distant metastasis [9
]. A complete excision with wide surgical margins should be the preferred treatment.