Low-grade fibromyxoid sarcomas (LGFMS) bear either the t(7,16) (q32-34;p11) or t(11,16) (p11;p11) translocations, resulting in FUS-CREB3L2 or FUS-CREB3L1 fusions, respectively. Heretofore, fusion transcripts were mainly detected in frozen tissues, using reverse transcription-polymerase chain reaction. In this study, we aimed to develop a reliable method to detect these in paraffin-embedded tissues, and to examine the clinicopathologic characteristics of a series of translocation-positive LGFMS. Sixty-three neoplasms with typical morphologic features of LGFMS and 66 non-LGFMS tumors selected for their resemblance to LGFMS (LGFMS-like tumors) were examined. RNA of sufficient quality could be extracted from 111/129 (86%) cases (59 LGFMS, 52 non-LGFMS). Of all, 48/59 (sensitivity, 81%) LGFMS contained detectable transcripts (45 FUS-CREB3L2, 3 FUS-CREB3L1). Most relevant clinicopathologic features of fusion-positive LGFMS included predominance in lower extremities (22/48; thigh: 13/48), deep situation (46/48), and occasional presence of unusual histologic features, for example, hypercellular areas (16/48), foci of epithelioid cells (13/48), and giant rosettes (6/48). Most tumors expressed EMA (41/45), at least focally, CD99 (38/41) and bcl-2 (36/41) while being essentially negative for CD34 (2/45), mdm2 (1/41), smooth muscle actin (1/45), S100 protein (0/46), desmin (0/44), h-caldesmon (0/42), keratins (0/44), and CD117 (0/40). Eleven presumed LGFMS were fusion negative. Of all, 7/52 non-LGMFS neoplasms contained FUS-CREB3L2 transcripts, of which 4 had been diagnosed as sclerosing epithelioid fibrosarcoma. In conclusion, FUS-CREB3L1/L2 fusion transcripts can be detected in paraffin-embedded LGFMS in a sensitive manner, using reverse transcription-polymerase chain reaction. Most fusion-positive LGFMS are EMA-positive and CD34/S100/smooth muscle actin negative. The presence of epithelioid cells and fusion transcripts in both LGFMS and a subset of sclerosing epithelioid fibrosarcoma suggest that these neoplasms might be related.

The clinicopathologic and immunohistochemical features of 63 pleomorphic liposarcomas are presented. There were 35 men and 28 women (median age 63 years; range 18-93 years). Tumor size ranged from 2 to 23 cm (median 10 cm). Tumor locations included lower extremity (36.5%), especially the thigh (28.5%), limb girdles (17.5%), upper extremity (16%), thoracoabdominal wall (9.5%), and internal trunk (20.5%). A total of 75% were deep seated and/or extracompartmental. Histologically, lesions show a varying combination of lipogenic and nonlipogenic areas characterized by malignant fibrous histiocytoma-like, round cell liposarcoma-like, and/or epithelioid/carcinoma-like features. A pericytic pattern was focally present in 15 (24%) tumors. Eighteen (29%) lesions were grade 2, and 45 (71%) were grade 3 sarcomas. Tumor necrosis was observed in 51 (81%) cases, vascular invasion in three, and mitotic counts ranged from 3 to 124 per 10 high power fields (median 25). Lipogenic areas were S-100 protein immunoreactive, at least focally, in 20 of 42 (48%) cases. Nonlipogenic areas showed focal reactivity for smooth muscle actin (24 of 49; 49%), desmin (9 of 48; 19%), CD34 (18 of 45; 40%), S-100 protein (5 of 49, 10%), CD68 (6 of 46, 13%), and epithelial membrane antigen (13 of 49, 26.5%). Epithelioid areas showed epithelial membrane antigen (4 of 11; 36%) but not cytokeratin (0 of 11) reactivity. Treatment procedures in 51 patients consisted of simple tumorectomy (16) and wide excision (33). Five and 31 patients received neoadjuvant and adjuvant chemotherapy and/or radiation therapy, respectively. Follow-up (48 patients, range 7-276 months; median 38 months) showed a 45% local recurrence rate and a 42.5% metastasis rate, metastases occurring mostly in lungs and pleura. Seventeen patients (35%) died of disease, of whom none was metastatic at diagnosis. Five-year overall, metastasis-free, and local recurrence-free survivals were 57%, 50%, and 48%, respectively. Patient age > or =60 years, truncal tumor location, deep situation, tumor size >5 cm, vascular invasion, and incomplete tumor excision were significant adverse prognostic factors. Tumor grade and histology did not affect patient outcome. In conclusion, pleomorphic liposarcoma is a rare, often deep-seated and limb-based aggressive and metastasizing neoplasm of late adulthood. It shows a wide range of morphologic appearances, but tumor grade and histology have no effect on patient outcome.