Desmoplastic small round cell tumor (DSRCT), which was first described as a specific disease by Gerald and Rosai [8, 9], is a rare and aggressive soft-tissue sarcoma. Generally, DSRCT originates from the serosal surface of the abdominal cavity [30–33], but it can also be found in the lung, eye and salivary gland [3, 43–45, 49–53]. DSRCT has no specific clinical symptoms. Most patients have the initial symptoms of abdominal mass, constipation, ascites, and vomiting [18, 32, 33, 35–38]. It can be accompanied by the manifestations of cachexia, such as fatigue and emaciation. Patients may develop intestinal obstruction, hydronephrosis, and urinary tract irritation owing to tumor compression [15]. In our patient, a right submandibular mass without apparent clinical manifestations was detected incidentally. DSRCT is often widely disseminated throughout the peritoneal cavity, and some patients may present with metastasis to the lymph nodes, liver, and occasionally the lungs [16, 18, 30]. Hence, its prognosis is exceedingly poor. The clinical features of previously published DSRCT cases in the last 5 years are summarized in Supplementary Table 2, Additional File 1.
Imaging examinations of DSRCT lack characteristic features. Ultrasound examination usually shows a lobulated soft tissue mass with an uneven internal echo [17]. Computed tomography (CT) usually reveals single or multiple lobular nodules or lumpy soft tissue masses, with an uneven density of the tumor body and multiple spotted calcifications [32]. The lesions tend to crowd out, surround and invade the surrounding tissues [43]. DSRCT is usually accompanied by flakes of low intensity when there is a necrotic area in the tumor. Enhanced CT presents mild uneven enhancement, and edge enhancement may be observed in some larger masses. Moreover, positron emission tomography (PET)-CT have the potential to monitor residual disease and detect relapse or tumor progression at the early stages [14]. Imaging findings are non-specific, but they can indicate the location, size, and the number of tumors, thereby contributing to biopsy, surgery, and radiotherapy.
The definitive diagnosis of DSRCT is based on typical morphological and immunohistochemical features, especially distinctive molecular characteristics. The pathological and molecular features of previously published cases of DSRCT in the last 5 years are summarized in Supplementary Table 3, Additional File 2. Histologically, the tumor tissue consists of small round cells and peripheral desmoplastic stroma, which can be accompanied by cystic degeneration and hemorrhagic necrosis. The tumor had a variegated histology revealing pseudopapillary architecture, rhabdoid, clear or pleomorphic cells in addition to typical small round cell morphology. Moreover, the tumor can have intermittent areas of primitive tubules or rosette-like structures [18]. Immunohistochemically, tumor cells show a pattern of multi-phenotypic differentiation [30, 39–41]. This multiple antigen expression profile is a characteristic of DSRCT and can be used to distinguish DSRCT from the other histologically related small round cell tumors. Further, para nuclear dot-like desmin positivity has important diagnostic significance. However, in our case, immunohistochemical staining showed a diffuse perinuclear staining pattern with desmin, but characteristic dot positivity was not prominent. Almost all cases of DSRCT are positive for WT1 and show cytoplasmic and paranuclear staining. Although the immunohistochemical analysis of classic cases of DSRCT tend reveal WT1 positivity, N- and C-terminals may be useful as a form of “molecular immunohistochemistry” to identify the EWS–WT1 transcript as the immunostaining pattern may be altered by variant transcripts and WT1 immunostaining may be negative (as in our case) [46–48]. To establish a DSRCT diagnosis, the interpretation of WT1 immunostaining requires knowledge of antibody target epitopes and correlations with clinical, morphological, and molecular findings.
In the present case, the tumor was composed of nests of small to medium-sized cells, which might be misdiagnosed as small cell carcinoma. Small cell carcinoma can also show the immunoreactivity for epithelial and neuroendocrine markers. But in our case, immunohistochemical analysis of the co-expression of EMA, vimentin and desmin by tumor cells strongly support a diagnosis of DSRCT. DSRCT should also be distinguished from other carcinomas such as malignant melanoma, malignant lymphoma and metastatic neuroblastoma. However, the current case was negative for Melan-A and S100, which made malignant melanoma unlikely. And the positivity for epithelial markers helped to rule out the possibility of malignant lymphoma, which often involves lymph nodes, bone marrow and peripheral blood. In addition, stroma of massive nerve fiber network is a characteristic feature in neuroblastoma, which might be a diagnostic clue. Due to the histological features of small round cells in the present tumor, it must be distinguished from other small round cell tumors such as rhabdomyosarcoma, primitive neuroectodermal tumor (PNET) and Ewing sarcoma (EWS). Rhabdomyosarcoma is more common in children; the tumor cells are commonly positive for myogenic markers (such as MYOD1, myogenin), but negative for epithelial and neuroendocrine markers. Morphologically, DSRCT and PNET can revealed chrysanthemum-like structure, and both of them have positive expression of CD99 and NSE. In our case, immunohistochemical results show the positivity for desmin and epithelial markers, which favors a diagnosis of DSRCT over that of PNET. EWS shares histological and immunophenotypic similarities with DSRCT. However, the survival rate of patients with DSRCT is significantly lower than that of EWS patients, which indicates that DSRCT and EWS have different biological backgrounds. EWS mainly occurs in children and is common in bones, EWS involving soft tissue is rare. EWS can also be positive for cytokeratin, desmin, CD99, FLI-1 and neuroendocrine markers, which may cause confusion. The diffuse membranous positivity of CD99 is typical of EWS, but in our case, nonspecific cytoplasmic positivity for CD99 is one of the features of DSRCT. Furthermore, in our case, negative of NKX2.2 is an important clue to distinguish EWS. Both EWS and DSRCT harbor EWSR1 rearrangements, the break-apart FISH assay for EWSR1 will not be helpful in the differential diagnosis between them. But characteristic translocation of EWS involves EWSR1 and the ETS family of transcription factors, not WT1. Convincingly, documentation of EWSR1-WT1 fusion is the “gold standard” for the diagnosis of DSRCT [35, 39, 40]. It was not performed in our case due to limited conditions. Taken together, combined with the tumor location and morphological features, as well as a distinctive pattern of multi-phenotypic differentiation on immunohistochemistry, we prefer the diagnosis of DSRCT. Like DSRCT, both myoepithelial carcinoma and synovial sarcoma are also multiphenotypic and expresses multilineage markers. Myoepithelial carcinoma expresses cytokeratin and myogenic markers such as myogenin, smooth muscle actin (SMA) and HHF35. The case in this study is positive for desmin and negative for all other myogenic markers, P63 and CK5/6, strongly favors a diagnosis of DSRCT. Synovial sarcoma mainly occurs in the extremities. It expresses epithelial and mesenchymal markers, but desmin positivity is uncommon, which can help in the differentiation.
DSRCT is distinguished by the t (11;22) (p13; q12) chromosomal translocation involving a fusion between the transcriptional activating domain of EWSR1 and the WT1 gene [19–21]. Studies have also suggested that the EWSR1-WT1 fusion protein can induce the expression of PDGFA. PDGFA can induce the growth and proliferation of fibroblasts and the production of collagenous stroma, which may explain the characteristic reactive fibrosis of DSRCT [22]. Downstream activation of EWSR1-WT1 gene fusion includes signaling pathways of vascular endothelial growth factor (VEGF), IL2RB, and insulin growth factor (IGF)-1 [23–25]. A better understanding of the effects of these target genes will provide avenues for future treatment.
Despite multimodality treatment, DSRCT is highly aggressive and has a poor prognosis. The overall survival in patients is < 3–5 years after diagnosis, and the 5-year survival rate is < 20% [18, 26, 39, 48]. There is no standardized approach for the treatment of this malignant disease. Effective cytoreduction combined with comprehensive therapies, as the best treatment strategy presented in most studies, may prolong patient survival [27, 37–41]. With the in-depth analysis of molecular genetics of DSRCT, targeted therapy, immunotherapy and other methods have been used for the treatment of DSRCT in recent years [28, 29].