Answer: Sinonasal hemangiopericytoma

Histology: The subepithelial connective tissue is infiltrated by a highly cellular proliferation of oval-to-spindled cells. The cells form tightly packed swirls around central compressed thin-walled vascular channels. Despite the high cellularity, the tumor cells are very bland and uniform. The cells are immunoreactive for CD34; and they are not immunoreactive for actin, S-100, cytokeratin or EMA.

Discussion: Hemangiopericytomas usually arise in the deep soft tissues, but they can also involve the sinonasal tract. In these instances, they typically arise in the paranasal sinuses, secondarily extend into the nasal cavity, and present with non-specific nasal obstruction and epistaxis. On clinical grounds they are usually mistaken for an inflammatory/allergic sinonasal polyp. On histologic grounds they may be confused with various spindle cell mesenchymal neoplasms such as cellular schwannomas and synovial sarcomas. There is no specific marker for the pericyte such that the major role of immunohistochemistry is to exclude other entities. In this case, the absence of S100 staining was useful in excluding a schwannoma, and the absence of EMA staining was useful in eliminating the possibility of a synovial sarcoma. As some sinonasal hemangiopericytomas label with CD34, the distinction from solitary fibrous tumor is best made on histologic grounds. Granted that the two lesions may also demonstrate considerable overlap at the microscopic level as well, the classic solitary fibrous tumor tends to demonstrate more significant stromal collagen with more variation is lesion cellularity (i.e. alternating zones of hypo- and hyper-cellularity).

There is much confusion regarding the appropriate nomenclature for hemangiopericytomas of the sinonasal cavity. Despite their similarity to their soft tissue counterparts, hemangiopericytomas of the sinonasal tract more consistently behave in a benign fashion – they may recur if inadequately excised, but they do not metastasize. Consequently, it has been suggested that sinonasal hemangiopericytomas may be biologically distinct from their soft tissue counterpart, and hence the term hemangiopericytoma-like tumor of the sinonasal cavity. This distinction is probably erroneous. Many now believe that this more innocent behavior may simply reflect earlier presentation and removal of low-stage tumors, and that given the opportunity for tumor progression they could behave in a more aggressive fashion. Indeed, examples of sinonasal hemangiopericytomas that have metastasized are now well-documented. Malignant behavior is predicted by tumor necrosis and a mitotic rate exceeding 1 mitosis per 10 high-power fields.

Reference(s):