A nationwide analysis of desmoplastic small round cell tumor.

This study aim is to enhance the understanding, diagnosis and treatment of desmoplastic small round cell tumor (DSRCT) and to determine what factors can affect survival of the disease in China.We report here 8 patients with DSRCT in our center who received a variety of treatment methods. By reviewing the literature published from Chinese database (CNKI, WANGFAN, VIP, CBM, CMCC) in 2000 to 2015 with the terms of "dsrct", "desmoplastic" and "small round-cell tumor",104 eligible cases of DSRCT(including 8 cases in our hospital) were retrospectively analyzed.Among the 104 patients, Median age was 24 years with a range of 15 to 54 years. The main primary tumor site was the abdomen and/or pelvis in 92/104 patients (88.5%). Only 25% of patients had localized disease. Most of the patients had received adjuvant chemotherapy (87.5%) and 76.9% patients had not experienced adjuvant radiotherapy. One-fourth of the patients underwent grossly complete surgical resection, and 33.7% and 41.3% patients received no surgery and incomplete surgical resection, respectively. Median overall survival for all patients was 26 months (95% CI: 20.29-31.71). Multivariate analysis revealed that Metastatic status (HR: 2.327, 95% CI: 1.136-4.768, P = .021), Surgical patterns (HR: 0.673, 95% CI: 0.487-0.928, P = .016), and Adjuvant chemotherapy (HR: 0.337, 95% CI: 0.167-0.678, P = .002) were significant independent prognostic factors for longer overall survival. It was noteworthy that CD99 were significantly associated with OS (P = .002).Here, we identified the prognostic factors which may facilitate risk-adapted treatments for this rare DSRCT group, which should be further investigated.

Introduction

Desmoplastic small round cell tumor (DSRCT) is a rare and highly aggressive neoplasm described as a distinct clinicopathological entity in 1989 by Gerald and Rosai.[ The literature regarding this tumor is limited and generally comprises case report. And there is no large sample of national studies, especially in China. DSRCT has a predilection for children and adolescents, and primarily involves the abdomen and pelvis. The tumor is more common among men than among women. The extent and location of the tumor correlate with symptoms seen at presentation. Abdominal fullness, constipation, and abdominal distension are the most common symptoms.

The histopathology and immunocytochemistry of the disease have been well presented.[ The diagnosis is very difficult and correct diagnosis depends on pathological immunohistochemical characteristics, further ectopic EWS-WT1 fusion gene detection needed if possible.[

Owing to diversity in treatment options, there is no proper consensus on the best treatment plan of DSRCT. Current treatment options for the international mainstream is alkylating agent-based chemotherapy, the maximum cytoreductive surgery (>90%) and the whole abdominal radiotherapy.

The prognosis of DSRCT patients is very poor, overall survival is approximately 30% to 55% despite chemotherapy, radiotherapy, and aggressive surgical resection.[ The objective of this study was to identify the prognostic risk factors by retrospectively analyzing the 104 mreported cases and review systematically the current knowledge of this unusual tumor to present clinical characteristics, pathologic features, treatment approaches, and prognostic factors of this tumor.

Materials and methods

Patients

The records of 8 cases of DSRCT diagnosed between 2000 and 2015 were retrieved from the consultation files and database of the First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou, China. The clinical characteristics, immunohistochemical result, treatment and outcome were collected and summarized as follow (Tables 1 and 2). Written informed consent for study participation was obtained from all patients.

Table 1

Clinical feature and follow up data of 8 DSRCT cases.

Table 2

Immunohistochemical result of 8 DSRCT cases.

The literature search used the Chinese databases between 2000 and 2015, including CNKI, WANGFAN, VIP, CBM and CMCC. By retrieving the key words of “dsrct”, “desmoplastic tumor” and “small round cell tumor”, we obtained the available clinical data for 96 patients with DSRCT. The database contains patient age, gender, presentation, primary tumor site, metastatic status, pathologic findings, therapy and follow-up data. In addition, the incomplete reports were removed. These references were selected because they presented patient information independent of each other, and all references were reviewed and evaluated by three independent members.

Data collection

The prognostic factors were analyzed: sex, age, tumor primary site, metastatic status, surgery, adjuvant chemotherapy, adjuvant radiation and immunohistochemical indicators. In the present study, the surgery were divided into three groups (1) those who received no surgery, (2) those who received complete surgical resection, and (3) those who received incomplete surgical resection. The complete surgical resection was defined as definitive removal of at least 90% of tumor; otherwise surgery was considered to be incomplete surgical resection. And tumor biopsy at diagnosis was not considered to constitute surgery.

Statistical analysis

All statistical analysis was performed by SPSS16.0 (SPSS Inc, Chicago, IL, USA), including the following variables: sex (male/female), age as a categorical variable using a cutoff of 24 years, tumor primary site (abdomen/pelvis vs other sites), metastatic status (localized vs disseminated), surgery (no/complete/incomplete), Adjuvant chemotherapy, Adjuvant radiation and immunohistochemical indicators. Overall survival (OS) was estimated using the Kaplan-Meier method. Overall survival was calculated from the date of diagnosis to the date of death from the disease or the last follow-up. Multivariate survival analyses using Cox's proportional-hazards method were performed in order to define the prognostic factors for OS. Hazard ratio and 95% CI (confidence interval) were also estimated. A probability value of less than .05 (P < .05) was considered to be statistically significant.

Results

Baseline characteristics

A total of 104 patients were identified from the database who were diagnosed with DSRCT and met the inclusion criteria. In these cases, DSRCT showed an apparent male predominance (89/104, 85.6%) to female (15/104, 14.4%) and at a mean age of 24 years (range from 15 to 54 years). The clinical manifestations were nonspecific and variable. The predominant clinical manifestations at presentation include abdominal pain or discomfort (49.1%), palpable abdominal mass or distension (43.9%), weight loss (19.3%). Other symptoms or signs, like anorexia, umbilical hernia, Bowel habit change, ascites, constipation, urinary complaint and so on, were also reported. This might be associated with tumor sites.

The tumor primary site, metastatic status and other patient characteristics were summarized in Table 3. The main primary tumor site was the abdomen and/or pelvis in 92/104 patients (88.5%). Primary tumor sites in the other patients were lung, testis, cerebellopontine angle, mandible, or heart. Only 25% of patients had localized disease. And 75% patients underwent metastatic disease at the time of diagnosis and therapy, including direct spread to adjacent areas, distant metastasis, lymph node metastasis. Most of the patients had received adjuvant chemotherapy (87.5%), and 12.5% patients had no adjuvant chemotherapy. On the contrary, a few patients received adjuvant radiotherapy (23.1%), and 76.9% patients had not experienced adjuvant radiotherapy. One-fourth of the patients underwent grossly complete surgical resection, and 33.7% and 41.3% patients received no surgery and incomplete surgical resection, respectively.

Table 3

Characteristics of all 104 patients.

Histopathology

In our 8 patients, the neoplastic masses were solid, firm, and multilobulated with a gray-white cut surface, sometimes distorted by cystic change and areas of necrosis, ranging from 2 to 17.5 cm. Microscopically, the tumors were composed of sharply demarcated nests of varying size and irregular shape deposited in an abundant desmoplastic stroma (Fig. 1A). The central necrosis was often seen in large tumor nests, the tumor cells appear small and undifferentiated with inconspicuous nucleoli and abundant mitosis (Fig. 1B). Typically, the cellular aggregates are surrounded and separated by abundant fibrous connective tissue with only a scattering of spindle-sharped fibroblasts and myofibroblasts. Occasionally, the tumor cells have more abundant cleared-out or vacuolated cytoplasm of a single ring-like appearance. A relatively common finding is the presence of rhabdoid-like foci in which the tumor cells have paranuclear intracytoplasmic hyaline inclusions composed of aggregates of intermediate filaments. Other rare features include Homer Wright-like rosettes, papillary areas, zones that resemble traditional cell carcinoma.

Figure 1

Immunohistochemical staining in DSRCT tissue. (A) The tumors were composed of sharply demarcated nests of varying size and irregular shape deposited in an abundant desmoplastic stroma (HE, original magnification ×100). (B) The central necrosis was often seen in large tumor nests, the tumor cells appear small and undifferentiated with inconspicuous nucleoli and abundant mitosis (HE, original magnification ×100). (C) The tumor cells showed diffuse and strong staining for AE1/AE3 (original magnification ×400). (D) Immunostaining for desmin was noted with a distinctive dot-like pattern (original magnification ×400).

Table 2 shows the immunohistochemical analysis of our 8 patients. All 8 cases showed diffuse and strong staining for AE1/AE3 (Fig. 1C), VIM and Desmin, some of them also expressed EMA, CD56, CGA, SYN, S-100 and PLAP and none for LCA and MyoD1 a distinctive dot-like intracytoplasmic localization is seen with Desmin (Fig. 1D). And presented in Table 4 is the immunohistochemical profile of DSRCT from the total of 104 patients. Similarly, in the literature review, the Desmin, VIM, EMA, WT1, CAM 5.2, Keratin and NSE were positive in a large percentage of patients.

Table 4

Results of immunohistochemical studies.

Follow-up and treatment outcome

Among 104 patients, 65 (62.5%) patients had died at the time of analysis and the median duration from diagnosis to death was 26 months (range, 3–120 months). In other words, estimated median OS was 26 months (95% CI: 20.29–31.71). The overall survival rates 1, 2, and 3 years were 80.2%, 54.2%, and 33.2%, respectively (Fig. 2).

Figure 2

Overall survival of all the 104 patients.

Analysis of prognostic factors

Results of univariate and multivariate analyses for analysis of prognostic factors are summarized in Table 5. In univariable analyses, No metastases, surgery (complete and incomplete) and Adjuvant chemotherapy were significantly associated with better OS (Figs. 3–5).

Table 5

Univariable and multivariable analysis of prognostic factors.

Figure 3

Overall survival compared between patients with metastases vs no metastases. The survival rate for those patients with metastases was significantly lower than those with no metastases.

Figure 5

Overall survival compared between patients treated with adjuvant chemotherapy compared with no chemotherapy.

For the patients without metastases, the survival rates 1, 2, and 3 years were 84.6%, 71.1%, and 48.0%, respectively; for metastases, the survival rates were 78.6%, 48.3%, and 28.4%, respectively. The survival rate for those patients with metastases were significantly lower than those with no metastases (P = .005) (Fig. 3).

For the patients with no surgery, the survival rates 1, 2, and 3 years were 67.4%, 47.0%, and 13.7%, respectively. For incomplete surgery, the survival rates 1, 2, and 3 years after were 76.0%, 51.9%, and 31.1%, respectively. For complete surgery, the survival rates were 92.8%, 61.1%, and 49.6%, respectively. Obviously, the patients with complete surgery had the highest survival rate (P = .009) (Fig. 4).

Figure 4

Overall survival in patients treated with no surgery vs incomplete surgery vs complete surgery. Complete surgical resection is defined as greater than 90% tumor resection.

For the patients with receiving adjuvant chemotherapy, the survival rates 1, 2, and 3 years were 83.0%, 56.7%, and 35.1%, respectively. For no chemotherapy, the survival rates 1, 2, and 3 years were 61.5%, 38.5%, and 19.2%, respectively. The survival rate for those patients with receiving adjuvant chemotherapy were significantly higher than those with no chemotherapy (P = .026) (Fig. 5).

Univariable modeling in DSRCT patients demonstrated that sex, age, tumor primary site and Adjuvant radiotherapy did not influence OS. Subsequently, three remaining variables (Metastatic status, Surgery and Adjuvant chemotherapy) showed statistical differences in univariate analysis were introduced into the Cox regression model. Results indicated that metastatic status (HR: 2.327, 95% CI: 1.136–4.768, P = .021), surgery (HR: 0.673, 95% CI: 0.487–0.928, P = .016), and adjuvant chemotherapy (HR: 0.337, 95% CI: 0.167–0.678, P = .002) were also independent predictor for DSRCT patients (Table 5).

No data regarding immunohistochemical findings assessment of prognosis was found in other literature about DSRCT. We specially evaluated the immunohistochemical indicators (including EMA, NSE, CD99, CGA and CAM5.2) on prognosis of DSRCT by using univariate analyses. Only CD99 were significantly associated with OS (P = .002) (Fig. 6).

Figure 6

The patients with CD99 negative expression have longer survival.

Discussion

DSRCT is an orphan disease and is still poorly understood. It is characterized by neste of small tumor cells surrounded by a cellular and vascular collangenous stroma hence the name desmoplastic small round cell tumor. This tumor is a rare and highly aggressive neoplasm, which usually originates from the abdomen/peritoneum and has been often seen in male adolescents.[ Most of the literatures are limited to small, single-center case reports, and their results lack universality. There is no large sample of national studies, especially in China. Then we collected several research reports of DSRCT originating in multiple overseas centers, which was similar to our results (Table 6).[

Table 6

Desmoplastic small round cell tumors reported in the recent literatures.

In our study, DSRCT showed an apparent male predominance to female and the gender ratio was 89:15 (85.6%). And in the study by Lal et al the ratio was 60:6 (91%).[ In our series, the main primary tumor site was the abdomen and/or pelvis in 92/104 patients (88.5%). And the predominant clinical manifestations were abdominal pain or discomfort (49.1%), palpable abdominal mass or distension (43.9%) and Weight loss (19.3%), with a small number of patients complaining about Anorexia, Umbilical hernia, Bowel habit change, ascites, constipation, Urinary complaint and ect. According to the study by Gil et al,[ pain (52.1%) and increased abdominal girth (8.4%) were the predominant initial symptom or sign. In a word, DSRCT do not have specific clinical presentation. This might be associated with tumor sites. On this regard, doctors need to take DSRCT into consideration when the above symptoms are present. And the relevant clinical examination is necessary, such as imaging, cytology, and laboratory blood test.

The diagnosis of DSRCT is very difficult, usually based on histologic and immunohistochemical examination. Molecular biology tests looking for the EWS-WT1 fusion gene transcripts are therefore mandatory to confirm the diagnosis.[ By investigating 8 patients of our institution, we can see the histological characteristic of DSRCT. Histologically, the tumors were composed of sharply demarcated nests of varying size and irregular shape deposited in an abundant desmoplastic stroma. The central necrosis was often seen in large tumor nests, the tumor cells appear small and undifferentiated with inconspicuous nucleoli and abundant mitosis (Fig. 1A, B). these characteristic features of DSRCT distinguish it from other small round cell tumors and tumors with dense stroma. The tumors were found to be the coexpression of neural, mesenchymal and epithelial markers by immunohistochemical examination.[Table 4 describes detailedly immunohistochemical results in our group. The Desmin, Keratin (AE1/AE3), Vimentin, NSE, EMA, WT1 and CAM 5.2 were positive in a large percentage of patients. And Vimentin even reached 100%. The result is similar to the previous literature reports.[ The presence of these is uncommon in other tumors. In general, desmoplastic small round cell tumor is a vexing disease with poor long-term survival. As reported earlier by Lal et al, the survival rate of DSRCT in a 3-year period was 44%.[ But a large sample analysis of 491 patients from Gani et al reported a 3-year survival of 32.3%.[ Our results showed that the three-year overall survival rate is 33.2%, which is somewhat closer to the recent large sample studies.

Multivariate factors analyses showed that metastatic status, surgical patterns and adjuvant chemotherapy had an impact on overall survival independently (Table 5). Metastatic status has a very important impact on prognosis. Our study results also demonstrated that patients with metastatic diseases had significantly reduced survival rates. The patients with no metastases were associated with better OS (3-year OS 48% vs 28.4%, P = .005). This suggests that it is desirable to touch the patients with no metastases in order to improve the patient survival rate. Nevertheless, the reality is that DSRCT patients often have tumor metastases when they go to the hospital. Therefore, disease early detection is very important.

Lal et al had reported that complete surgical resection was highly significant in prolonging overall survival. Our study results also demonstrated that the patients experienced complete surgical resection were significantly associated with best OS (3-year OS 49.6% vs 31.1% vs 13.7%, P = .009). Similarly, the survival rate for those patients received adjuvant chemotherapy were significantly higher than those with no chemotherapy (3-year OS 35.1% vs 19.2%, P = .026). The most representative chemotherapy regimen is the P6.[ Intraperitoneal chemotherapy (either HIPEC or EPIC) after resection in DSRCT is controversial. The only randomized trial comparing the administration or not of intraperitoneal chemotherapy after resection tumors showed that intraperitoneal chemotherapy has no significant effect on OS.[ Therefore, we suggest that DSRCT patients should accept the treatment protocol with Complete Surgery combined with chemotherapy.

To the best of our knowledge, no studies have reported the influence of immunohistochemical indicators on the prognosis of DSRCT. Here, we specially evaluated the immunohistochemical indicators (including EMA, NSE, CD99, CGA and CAM5.2) on prognosis of DSRCT. In the results of Kaplan-Meier survival analysis, we found that only CD99 was a significant predictor for OS (P = .002).

CD99 is a transmembrane glycoprotein encoded by the MIC2 gene, located on the short arm of the X and Y chromosomes.[ The protein has recently been involved in cell adhesion, apoptosis, differentiation of T cells and thymocytes, the migration of monocytes and the intercellular adhesion between lymphocytes and endothelial cells, which may play an important role in the development of tumor.[ In previous studies, loss of CD99 was found to be associated to poor prognosis in patients with osteosarcoma, pancreatic endocrine tumor and gastric adenocarcinoma.[ However, our results suggest that patients with CD99 negative expression have longer survival, which should be further investigated.

In conclusion, the surgical patterns, metastatic status, and adjuvant chemotherapy are independent factors affecting the desmoplastic small round cell tumor prognosis. The survival rates are lower in patients with metastases or no surgery or incomplete surgery or no chemotherapy, suggesting the importance of early diagnosis, early treatment, complete surgery and adjuvant chemotherapy. In addition, we find that CD99 has also a very important impact on prognosis in DSRCT patients.

Author contributions

Conceptualization: Tao Xiang, Hui Li.

Data curation: Tao Xiang, Shuyang Zhang, Saisai Wang.

Formal analysis: Tao Xing.

Funding acquisition: Tao Xing.

Investigation: Tao Xing.

Methodology: Tao Xing.

Project administration: Tao Xing.

Resources: Tao Xing.

Software: Tao Xing.

Validation: Rushan Fei.

Writing – original draft: Tao Xing.

Writing – review & editing: Tao Xing.