A case of pure sclerosing epithelioid fibrosarcoma with a cytogenetic and fluorescence in situ hybridization study

Hiroaki Suzuki,H. Takeda,K. Yamashiro,T. Soma,T. Osanai,H. Hiraga,K. Isu,M. Tamakawa,T. Nojima

DOI: https://doi.org/10.1111/pin.12223

2015-01-01

Abstract:To the Editor: Sclerosing epithelioid fibrosarcoma (SEF) is a variant of fibrosarcoma characterized by epithelioid fibroblasts arranged in cords and nests in a sclerotic stroma. It is known that a subset of SEF appears to be related to lowgrade fibromyxoid sarcoma (LGFMS).SEF rarely contains the morphology of LGFMS and LGFMS also rarely contains the morphology of SEF. Doyle et al. applied the term ‘hybrid SEF-LGFMS’ for the tumor with hybrid features. MUC4 which was firstly found as a sensitive and specific marker for LGFMS was also shown as a useful marker for SEF later. Thus, the potential relationship between SEF and LGFMS is very interesting. On the other hand, the cytogenetic findings of SEF were not the same as the findings of LGFMS. The translocation, t(7;16)(q33;p11), resulting in the FUS-CREB3L2 fusion gene has been found in approximately two thirds of LGFMS cases. But, the SEF cases which were karyotyped did not show the cytogenetic changes observed in LGFMS although the cytogenetic information of SEF has been limited. Wang et al. showed that FUS rearrangements are rare (2 out of 22 cases) in pure SEF by fluorescence in situ hybridization (FISH). Arbajian et al. recently reported recurrent EWSR1CREB3L1 gene fusions in SEF. EWSR1 gene locates on chromosome 22q12. However, chromosomal abnomalities involving 22q12 have not been observed by the G-banded karyotypic analysis in SEF. We report a case of pure SEF with the karyotypic findings and FISH results. The patient was an 16-year-old Japanese female. She noticed an abdominal mass. Magnetic resonance imaging demonstrated a tumor in the right abdominal external oblique muscle. Biopsy and resection was performed. The patient was alive and free of disease for 18 months after surgery. Macroscopically, the tumor measuring approximately 3.5 × 3.2 × 3.0 cm was observed in the right abdominal external oblique muscle. The tumor was well circumscribed. The cut surface was yellowish-white (Fig. 1a). Necrosis was not observed. On microscopic examination, the tumor cells were embedded in dense fibrous stroma The cellularity was variable. Small to medium sized epithelioid cells were arranged in cords and nests embedded in fibrous stroma (Fig. 1b). The epithelioid cells had relatively bland round and ovoid nuclei. Bone formation and cartilage formation were focally observed (Fig. 1c). Giant cells or highly pleomorphic cells were not seen. The morphology of LGFMS was not found in the tumor. Immunohistochemically, the tumor cells were diffusely positive for MUC4 (clone 8G7; Santa Cruz Biotechnology, Santa Cruz, CA, USA) (Fig. 1d) and vimentin (clone SP20; Nichirei, Tokyo, Japan). The tumor cells were focally positive for EMA (clone E29; DAKO Cyotomation, Glostrup, Denmark) in cytoplasm and negative for desmin (clone D33: Nichirei), beta catenin (clone 14: BD Transduction Laboratories, San Jose, CA, USA), S100 (polyclonal, DAKO Cytomation), cytokeratin (AE1/AE3, Nichirei), bcl-2 and alpha smooth muscle actin (clone 1A4, DAKO Cytomation). Karyotype analysis was performed with G-banding in Sapporo clinical laboratory (Sapporo, Japan) using fresh tissue sample from the surgical specimen. Twenty four metaphases were analyzed. All 24 metaphases revealed numerical and structural abnoromalities as follows: 45, XX, add(1)(p13)x2, + add(1)(p36), der(8)t(8;12)(q22; q13), add(11)(p11), del(11)(q?), −12, −15, −17, add(19) (p11), add(20)(q11), add(22)(q13)x2, + mar (Fig. 2a). FISH analysis was performed on paraffin sections following manufacturer’s protocol, using FUS and EWSR1 dual-color break-apart probes (Vysis, Abbot, IL, USA). Approximately 40% of 200 tumor cells showed one fused (red/green) signal and one or two red split signals per nucleus for EWSR1 break-apart probes (Fig 2b). The cutoff level for scoring aberration was 30% abnormal nuclei. The red split signals corresponded to the 5′ side of EWSR1. The green split signal corresponding to the 3′ side of EWSR1 was absent. The FISH results suggested presence of nonreciprocal translocation involving the EWSR1 gene. On the other hand, tumor cells showed 2 fused (red/green) signals per nucleus for FUS break-apart probes (Fig 2c). Pathologic diagnosis of the present case was pure SEF. FISH analysis suggested the presence of nonreciprocal translocation involving the EWSR1 gene. EWSR1 gene locates on chromosome 22q12. It is also uncertain whether the lack of green signal by FISH represents non-reciprocal translocation of EWSR1 gene or potential deletion involving the telomeric part of 22q12. However, recent literature suggested the presence of nonreciprocal translocation of EWSR1 gene in our ESF case. Recurrent EWSR1CREB3L1 gene fusion were recently shown in SEF by Arbajian et al. They showed that the split signals (2 cases) or the loss of the part flanking the 3′-part (3 cases) of the gene, indicative of involvement in a gene fusion, were seen in 5/10 informative SEF cases for the EWSR1 gene by FISH. It was also shown that RT-PCR for EWSR1-CREB3L1 fusion transcripts detected in-frame fusion in one SEF case with a loss of the part flanking the 3′-part of EWS gene by FISH. The loss of the part flanking the 3′-part of the EWS gene suggests Pathology International 2015; 65: 48–50 doi:10.1111/pin.12223 bs_bs_banner