Misinterpretation of positive staining of antibodies to desmin, smooth muscle actin, and muscle actin as representing smooth muscle differentiation in the context of a spindle cell tumor is not uncommon. Anti-h-caldesmon is a promising novel immunohistochemical reagent for more specific smooth muscle differentiation. We studied 72 tumors (11 leiomyosarcomas, 26 malignant fibrous histiocytomas [MFHs], 11 fibromatoses, 11 cellular cutaneous fibrous histiocytomas [CCFHs], 5 malignant peripheral nerve sheath tumors, 4 synovial sarcomas, and 4 cases of nodular fasciitis), the reactive myofibroblastic response in 5 cases of acute cholecystitis, and the desmoplastic response surrounding 5 invasive breast carcinomas. Tissues were examined for expression of h-caldesmon, desmin, smooth muscle actin, and muscle actin. Diffuse staining for h-caldesmon was present only within the leiomyosarcomas. Focal staining for h-caldesmon involving less than 1% of lesional cells was present in 3 of 26 MFHs and 1 of 11 CCFHs. There was overlap in staining for the other "myoid" markers in all of the lesions that contained myofibroblasts. Anti-h-caldesmon seems to be a reliable marker of smooth muscle differentiation, and its inclusion in a panel of myoid immunohistochemical reagents should allow distinction of smooth muscle and myofibroblastic tumors.

Heavy, multinodular cutaneous lymphoid infiltrates have an excellent prognosis. Multiple lesions at presentation are the best predictor of recurrent multiple lesions confined to the skin. The presence of a clonal B-cell population does not correlate with clinical presentation or histology, nor does it predict development of further lesions or systemic lymphoma.

Although studies suggest that microinvasive cervical adenocarcinoma has an excellent prognosis, none has reported treatment-related complications and many have lacked detailed measurement criteria. Our study looks at the rate of lymph node metastases and outcome, including complications, in patients with FIGO 1A1 and 1A2 adenocarcinomas of the cervix. Invasion was strictly defined, and the method of measurement was standardized. Villoglandular, papillary serous and clear cell carcinomas were excluded, as were tumors in which invasion exceeded 7 mm in width or 5 mm in thickness, with tumor thickness measured from the basement membrane of the overlying endocervical or ectocervical surface to the deepest focus of invasive tumor. A mean follow-up of 54 months (range, 5-159 months) was available for 31 of 32 (97%) patients. A total of 29 of 32 patients underwent hysterectomies, 2 patients had radical trachelectomies, and 1 patient was treated by cone biopsy. One patient received adjuvant radiotherapy. A total of 27 of 32 patients had bilateral pelvic lymph node dissections, and no lymph node metastases were identified. No recurrences have been reported to date. One patient died of metastatic ovarian carcinoma 82 months after her diagnosis of cervical carcinoma. Two of 27 (7%) patients have chronic leg edema secondary to lymph node dissection. Given the excellent prognosis of this tumor, the absence of lymph node metastases and a lymph node dissection complication rate of 7%, less radical surgery should be considered in this low-risk patient population.

Gastrointestinal stromal tumors are CD117 (c-Kit)–positive mesenchymal neoplasms with histologic and ultrastructural features of the interstitial cell of Cajal. While tumors outside of the gastrointestinal tract have been described, to our knowledge the case we present is the first such case in the vagina. We describe a 75-year-old woman with a recurrent vaginal gastrointestinal stromal tumor without apparent rectal involvement. This tumor was characterized by short intersecting fascicles of spindled cells, focal necrosis, and 12 to 15 mitoses per 50 high-power fields. Immunohistochemistry revealed diffuse cytoplasmic positivity for CD117 (c-Kit), CD34, vimentin, and h-caldesmon. Tumor cells were negative for S100, desmin, actin, and CAM 5.2. The differential diagnosis in this case included a vaginal smooth muscle tumor. While histologically similar to a smooth muscle neoplasm, the immunohistochemical profile ruled out smooth muscle differentiation. Gastrointestinal stromal tumor should be considered in the differential diagnosis of vaginal mesenchymal neoplasms.