Desmoplastic small round cell tumor (DRSCT) was first described in 1989 by Gerald and Rosai as a highly aggressive primitive sarcoma characterized by nests of blue-staining tumor cells surrounded by dense stroma. Of note was the fact that these tumor cells were positive for markers of mesenchymal, neural, and epithelial lineages, suggesting that they may arise from undifferentiated progenitor cells (1, 2). A (11:22), (p13:q12) chromosomal translocation is present in all cases, resulting in an EWS-WT1 gene fusion product (3–5). This feature is considered pathognomonic and is required for its definitive diagnosis (6). DSRCT is rare, with an age-adjusted incidence rate of 0.3/million in the United States (7), and primarily arises in adolescent and young adult males, with around 80% of patients being male and an average age of 18-22 years at diagnosis (2, 8). It may disproportionately affect African Americans (7, 9). Overall 5-year survival is a dismal 15-30% (7, 10).

Most patients initially present with abdominal pain and distension, evidence of ascites caused by extensive tumor seeding of the peritoneum (6). The vast majority (>90%) of primary tumors are found on the serosal surfaces of the abdomen, but others have been found in the thoracic cavity, the skull, or even the hand (8). Rarely, early-stage DSRCT is discovered incidentally during surgery or imaging for other indications. More commonly, however, the disease has spread extensively and is considered Stage IV at the time of diagnosis. Because of this, a new staging system for DSRCT has been proposed by Hayes-Jordan and colleagues: Stage I would include patients with one or two abdominal tumors, Stage II would include patients with extensive peritoneal spread, Stage III would include patients with peritoneal disease plus liver metastases, and Stage IV would include patients with disease that has spread outside the abdomen (6).

There is no standard therapy for DSRCT. Treatment typically consists of multimodal neoadjuvant chemotherapy followed by surgery and radiation. Treatment with the P6 protocol – cyclophosphamide, doxorubicin and vincristine alternating with ifosfamide and etoposide – is used in the United States, whereas a slightly different protocol used to treat Ewing Sarcoma patients (vincristine, doxorubicin, and cyclophosphamide alternating with ifosfamide and etoposide) is the most common regimen in other parts of the world (11, 12). Because of the extent of disease spread within the peritoneal cavity, surgery is most effective after neoadjuvant chemotherapy, since these tumors are typically responsive to chemotherapy and either shrink or become less vascularized (6). Complete surgical resection of all visible tumors was found to be essential for survival beyond 3 years, with 58% of patients who underwent complete resection surviving to 3 years compared to 0% of patients whose tumors were not resected and instead received only chemotherapy and radiation (10). While surgical excision of all visible tumors is critical, it may be impossible to remove every tumor cell from the peritoneal cavity. Hyperthermic intraperitoneal chemotherapy (HIPEC) has been proposed as a treatment for microscopic residual disease after complete surgical resection. A recent phase II clinical trial showed that HIPEC using cisplatin is effective at improving survival, with a 3-year overall survival rate of 79% (13). However, this treatment modality does not appear effective for patients whose tumors were not able to be completely removed during surgery (14). Whole abdominopelvic irradiation has also been used to treat residual disease following surgery and appears to reduce the incidence of peritoneal relapse, however, severe gastrointestinal and hematopoietic toxicity is common (15, 16). Intensity modulated radiation therapy reduces grade 2-4 toxicities without compromising efficacy (17). Autologous stem cell transplant has been investigated and does not appear to improve outcomes for DSRCT patients (18); allogenic stem cell transplant could be an alternative (19). Several clinical trials are currently underway to test new combinations of chemotherapy and targeted therapy ( Table 1 ). Progress has been made in extending survival, however, cures are rare.

One of the most well-characterized gene targets of EWS-WT1 is PDGFA, whose role in DSRCT was first described by Haber and colleagues in 1997, just a few years after the identification of DSRCT as a distinct malignancy (23). This group found that PDGFA was upregulated in an osteosarcoma cell line following induced expression of EWS-WT1 (23). Additionally, PDGFA expression was found in 13/14 DSRCT tumor specimens and correlated with expression of EWS-WT1 (23). These results and subsequent publications (24, 25) confirming the role of PDGFA in DSRCT laid the groundwork for clinical trials using imatinib mesylate, a tyrosine kinase inhibitor that targets abl, c-Kit, and PDGF receptor (PDGFR) and is known for its use in the treatment of chronic myelogenous leukemia (26). Unfortunately none of these trials have shown any efficacy in DSRCT (27–29).

Another target of EWS-WT1 is IGF-1R, a tyrosine kinase receptor that is frequently upregulated in cancer cells, leading to dysregulation of the IGF pathway (30). There are several reports of DSRCT patients presenting with severe hypoglycemia as a result of an elevated IGF-II : IGF-I ratio, a consequence of IGF pathway dysregulation (31, 32). Several IGF inhibitors have been tested in DSRCT patients in early phase clinical trials. In a phase II trial of 16 DSRCT patients treated with ganitumab [monoclonal antibody (mAb) IGF-1R inhibitor], 25% (4/16) achieved clinical benefit (PR + SD ≥ 24 weeks) (33). In another study combining cixutumumab (mAb IGF-1R inhibitor) and temsirolimus (mTOR inhibitor) in 20 patients with Ewing’s sarcoma family tumors, two DSRCT patients achieved partial tumor regression and one progressed (34).

Because DSRCT predominantly affects male patients, it has been hypothesized that androgen receptor (AR) could play a role in this disease. In one study of 27 patients with end-stage DSRCT, 10 (37%) had tumors that were positive for AR by IHC (35). Among the 10 AR-positive patients, 6 had been treated with combined androgen blockade (CAB). Three responded, with either stable disease or a reduction in tumor burden, though these responses were not durable, lasting only 3-4 months. The patients who responded to CAB had normal testosterone levels, whereas the patients who did not respond had castrate levels of testosterone, which could explain the lack of efficacy of CAB (35). A subsequent study found that AR-positive tumors were enriched for markers of stemness, which could also partially explain why the response to CAB in AR-positive patients with normal testosterone levels was short-lived (36).

In addition to the targets above, numerous other potential targets have been identified and are in various stages of pre-clinical investigation ( Table 2 ). These include pathways common to many types of cancer (DNA damage-response pathway, epithelial-mesenchymal transition) as well as receptors that have been identified as overexpressed or activated specifically in DSRCT as targets of the EWS-WT1 transcription factor (LRRC15, NTRK3).

A suitable target for immunotherapy must be expressed consistently on tumor cells but be restricted in its normal tissue expression to prevent on-target, off-tumor toxicity. Several appropriate targets have been identified in DSRCT. B7-H3 (targeted by the monoclonal antibody 8H9) is expressed in almost all DSRCT (35/37 tumors in one study (37), 44/46 in another) (38). It is tightly regulated by microRNA-29; despite universal transcription in most normal tissues, protein expression is highly restricted (39). Its hepatic expression explained the liver sequestration after intravenous injection, forcing its clinical development into compartmental administrations (NCT00582608). A clinical trial using intraperitoneal compartmental RIT using ¹³¹I-8H9 in DSRCT is ongoing (NCT01099644). Two studies have found GD2 expression in DSRCT, albeit to varying degrees (32/36 in one study (38), 2/20 in another.) (40) A T-cell engaging bispecific antibody (BsAb) trial is currently open for GD2(+) DSRCT (NCT03860207). EGFR, which is frequently mutated or overexpressed in cancer, was found to be amplified in 2/10 (20%) DSRCT tumors (41), and an EGFR CAR T cell trial for children and young adults with refractory/recurrent solid tumors (including DSRCT) is currently underway (NCT03618381). Another phase I CAR T cell trial was carried out for patients with HER2-positive sarcomas and the one DSRCT patient included achieved stable disease for 14 months (42). In a study assessing expression of various markers in DSRCT, HER2 was found to be expressed in 7/18 tumors, albeit at a low level for the majority of the positive-staining tumors (43). The use of these targets for immunotherapy in DSRCT is discussed in more detail below.

Several types of antibody-based therapies are used in the treatment of cancer, all of which rely on the specificity of an antibody-antigen interaction and require an appropriate tumor antigen to be identified that is expressed highly in tumor tissue and not in normal tissue. The most basic antibodies are monoclonal antibodies directed against a tumor antigen and designed to engage effectors cells through their Fc receptors, to mediate antibody-dependent, cell-mediated cytotoxicity (ADCC) either by natural killer cells (NK-ADCC) or by myeloid cells such as neutrophils and macrophages (myeloid-ADCC). In some tumors (e.g. neuroblastoma), complement-mediated cytotoxicity (CDC) can also play a role in the anti-tumor response. Enoblituzumab specific for B7-H3, is being tested in a phase I clinical trial in children and young adults with B7-H3-expressing relapsed/refractory solid tumors including DSRCT (NCT02982941). Enoblituzumab is also being tested in combination with checkpoint inhibitors pembrolizumab (NCT02475213) and ipilimumab (NCT02381314) among patients with B7-H3-expressing tumors.

There is also an ongoing trial to test the efficacy of a T cell-engaging BsAb in patients with advanced GD2+ tumors including DSRCT (NCT03860207). These BsAb are wide-ranging in structure, with one or more domains designed to engage T cells (usually targeting CD3) and one or more domains specific for a tumor antigen and serve as a sort of link between T cells and tumor cells, turning polyclonal T cells into specific killers. The advantage of BsAb over traditional antibody therapy is that they can initiate a robust T cell response against tumors at comparatively very low antibody concentrations and low target density. So far just one BsAb, blinatumomab, which targets CD3 and CD19, has been approved for use in cancer, in this case for ALL (47). Many others are in phase I-II trials, but only in adults (48).

Additionally, tumor antigen-specific antibodies can be used as vehicles to deliver radionuclide conjugates selectively to tumors in a process called radioimmunotherapy (RIT). This type of therapy has been successful clinically and two drugs have so far been approved for non-Hodgkin lymphoma (49). Currently, two radionuclide-conjugated antibodies are in trials for DSRCT: ¹³¹I-3F8, directed against GD2, which has previously been tested in medulloblastoma (50) and as a diagnostic tool in neuroblastoma (51), and ¹³¹I-8H9, directed against B7-H3, which has been previously tested in neuroblastoma (52). A recent Phase I trial (NCT01099644) of intraperitoneal ¹³¹I-8H9 in 48 DSRCT patients and four patients with other B7-H3-positive sarcomas found the treatment to be safe and well-tolerated with no dose-limiting toxicities (53). A Phase II trial for DSRCT is now accruing patients (NCT04022213).

In addition to using bi-specific antibodies to direct T cells to tumors, it is also possible to redirect T cells by genetically engineering them to express receptors specific for more classic antibody targets, i.e. not T cell receptor (TCR) peptide-MCH targets. These engineered cells are termed chimeric antigen receptor (CAR) T cells. In the past two years, two CAR T cell therapies have been approved by the FDA for treating CD19-expressing cancers based on clinical trials that showed they can induce complete responses in a significant proportion of patients (54, 55). Although current success with CAR T cell therapy has been largely limited to liquid tumors, clinical trials for solid tumors, including DSRCT and other sarcomas, are ongoing (56). HER2-CAR T cells were used in a phase I/II trial of sarcoma patients (including one with DSRCT) because evidence shows that HER2 is expressed at low levels by many types of sarcoma, including DSRCT (43, 57). While this low level of expression precludes immunotherapy with IgG monoclonal antibodies, CAR T cell therapy could be effective (57). This trial showed that HER2-CAR T cells delivered at low doses were safe, tolerable, and effective in maintaining stable disease in a subset of patients, with one DSRCT patient stable for >14 months (42). Phase I trials testing CAR T cells targeting EGFR (NCT03618381) and B7-H3 (NCT04483778, NCT04897321) are currently underway for patients with solid tumors including DSRCT.

It has been proposed that chromosomal translocation products can serve as tumor-specific antigens in sarcomas including DSRCT (58). This occurs when mutant proteins are processed by proteasomes and displayed as peptides on MHC molecules on the surface of tumor cells, as was demonstrated with mutant p53 (59, 60). A 9-amino acid epitope from the EWS-WT1 fusion protein has been described that binds to HLA-A3, providing the potential target for cytotoxic T cells (61). In one study among sarcoma patients, 2 patients with DSRCT were treated with autologous lymphocytes, tumor-pulsed dendritic cells and IL-7; unfortunately, neither had any measurable response (62). In another WT1 peptide vaccine trial among a group of 20 children with glioma, rhabdomyosarcoma, neuroblastoma, osteosarcoma, and clear cell sarcoma of the kidney (63), WT1-specific immune response was demonstrated in 4/18 (22%) evaluable patients (63). Although no DSRCT patients were included in this study, these results may provide a rationale for using peptide vaccines in DSRCT and other tumors with chromosomal translocation products that contain WT1.

Allogenic stem cell transplants (SCT) have been transformative in the treatment of hematological cancers. Although intense pre-transplant chemotherapy drastically reduces tumor burden, it rarely eliminates all tumor cells from the marrow compartment (64). Rather, a graft-versus-tumor effect is probably responsible for long term tumor control (65, 66). In this way, SCT is a form of cancer immunotherapy. Allogenic SCT has been shown to increase survival in advanced DSRCT (11.4 months vs. 1.9 months) (19). Autologous SCT has also been used but does not appear to be effective in increasing survival or reducing recurrence (18, 67). Yet, one DSRCT patient treated with CD34+ selected peripheral blood stem cells (PBSC) following surgical resection and high-dose chemotherapy, has maintained remission for over 10 years (68).

The field of immunotherapy was revolutionized by the advent of immune checkpoint inhibitors, drugs meant to release the “brakes” on T cells to reboot or to recruit T cells in their anti-tumor immune responses. Drugs targeting CTLA4 (ipilimumab) and PD-1 (programmed cell death protein 1) (nivolumab and pembrolizumab) have proven effective in treating melanoma, and, most excitingly, many of the patients who respond are alive years after treatment, suggesting that their remission is durable (69). Unfortunately, DSRCT, like sarcomas in general, do not respond well to checkpoint inhibition (36, 70). Typical among pediatric cancers driven by gene fusion, DSRCT has low tumor mutation burden, and hence few neoepitopes and insufficient anti-tumor T cell clonal frequencies (71). Tumor mutation burden, which often also correlates with a low degree of lymphocyte infiltration, is a predictor of response to checkpoint inhibitors (72). In contrast to melanoma and other tumors that are responsive to checkpoint inhibition, DSRCT appears to be an immunologically “cold” tumor, that is to say it is not heavily populated by tumor infiltrating lymphocytes (TILs) (73). DSCRT has low or no expression of PD-1/PD-L1 (programmed death ligand 1), suggesting that signaling along this axis is uncommon (41, 74). Put simply, removing the “brakes” from T cells is only effective if the host already has T cells capable of recognizing tumor epitopes just waiting to be rebooted, or if the naive T cells can be recruited to go after neoepitopes. Despite their failure as monotherapies in DSRCT, they could still hold promise when combined with other forms of immunotherapy, such as CAR T cells, cancer vaccines, and BsAb (75–77).

Low mutation load is now a well-recognized Achilles heel for most solid tumors in children and adolescents/young adults (71). This partly explained the paucity of TILs, which is made worse by the upfront use of aggressive chemotherapy, adopted as the standard of care from the time of diagnosis (6, 78). While this chemotherapy often succeeds at reducing tumor burden, it also depletes the patient’s immune cells, which may lead to depletion of effector cells, and especially T cells, thereby reducing the efficacy of immunotherapies (78).

Another challenge lies in achieving sufficient concentrations of antibody or large proteins in the tumor. When delivered intravenously, it may be difficult to achieve therapeutic doses of antibodies without on-target, off-tumor toxicity and other adverse effects. One solution for this is to deliver the immunotherapeutic agent regionally or to a specific biological compartment (56). Currently, a clinical trial (NCT00445965) is underway to test the efficacy of intrathecal delivery of the radiolabeled antibody ¹³¹I-3F8 to DSRCT and other tumors that have spread to or originated in the brain. Another clinical trial (NCT04022213) is in progress to determine the efficacy of ¹³¹I-8H9 delivered intraperitoneally for DSRCT and other solid tumors which metastasize within the peritoneum. Regional delivery has also been tested in radioimmunotherapy of neuroblastoma that has spread to the CNS, and even CAR T cells in ovarian cancer (52, 56, 79). These delivery methods aim to allow the treatment to reach sufficient levels in and around the tumor while protecting other normal tissues from the deleterious effects of high concentrations of these immunotherapeutic agents. Given the peripatetic presence across all tissues, T cells have the capability to penetrate deep into tissues unlike antibodies. Hence BsAbs that arm T cells either in vivo or ex vivo may be able to overcome penetration hurdles.

Although speculative, regional or compartmental delivery of BsAb may reduce on-target, off-tumor effects, which can be life-threatening. The most common adverse effect experienced by patients in these T cell-based therapy is cytokine release syndrome, which occurs when large numbers of T cells become activated and release cytokines such as IFN-y, GM-CSF, IL-10, and IL-6 (80). This occurs because of immediate contact between active T cells with targets in the hematogenous compartment (e.g. CD19), or direct activation of T cells even before arriving at the tumor site. The effects of this “cytokine storm” range from mild to severe, and can cause fever, fatigue, pain, nausea, hypotension, and organ failure (80). One patient treated with 1 × 10¹⁰ HER2 CAR T cells experienced respiratory distress within minutes of the infusion and died 5 days later (81). It is thought that the CAR T cells localized to her lungs and became active upon recognition of low levels of HER2 expressed in lung tissue (81). However, patients have safely been treated with HER2 CAR T cells at much lower doses, including one DSRCT patient treated with 1 × 10⁷/m² cells (42). Such life-threatening complications can potentially be avoided if the BsAbs are first administered into a “cold” compartment to bind to tumors before TILs arrive.

Finally, it is well known that tumors evolve resistance to treatments by downregulating expression of target molecules, and by upregulating expression of immune checkpoint molecules. In the context of immunotherapy, this is called immune escape (82). It has been demonstrated that tumors upregulate PD-1 and CD47 in response to treatment with BsAbs (77, 83). It has also been shown that tumors are capable of downregulating the targets of CARs, evidenced by the fact that 11% of B-cell malignancy patients who initially responded to treatment with CD19 CAR T cells eventually relapsed with CD19- tumors (84). One potential strategy to combat this phenomenon is to develop treatments that target multiple tumor antigens, including bispecific CAR T cells and T cell engaging antibodies with more than one tumor-targeting domain (82).