Advancing Understanding and Therapeutic Strategies for NUT Sarcomas: Comprehensive Review of the Literature and Two Cases

Soft tissue sarcomas (STSs) are a family of rare mesenchymal malignancies accounting for fewer than 1% of solid tumors in adult patients and 15% of pediatric patients in the United States.^[1,2] There are more than 70 distinct types of STSs,^[3] with a subset driven by gene fusions.^[4] One such fusion involves the nuclear protein in testis (NUT) protein, originating from the NUT Family Member 1 (NUTM1) gene situated in chromosome 15. Under physiological conditions, the NUT protein is expressed in post-meiotic spermatids, contributing to chromatin condensation by modulating acetylation patterns. However, when NUTM1 fused with genes responsible for transcription regulation, it transforms into an oncogenic driver, promoting the development of a spectrum of emerging neoplasms. NUT sarcoma can be considered a novel sarcoma subtype driven by NUTM1 fusions distinct from other NUT-rearranged tumors, such as NUT carcinomas, poromas and porocarcinomas of the skin, and NUT-rearranged hematological malignancies.^[5]

NUT carcinomas are rare and aggressive cancers notable for their poor prognosis with a median overall survival of 6.7 months.^[6] The first International Symposium of NUT Carcinoma suggested chemotherapy, especially ifosfamide-based regimens, as a suitable upfront treatment in the advanced setting.^[7] In addition, results from clinical trials for NUT carcinoma suggest that bromodomain and extraterminal (BET) inhibitors may have therapeutic value, whether administered as monotherapy or in combination with other agents.^[8–10] On the contrary, our understanding of NUT sarcoma is still evolving, marked by a substantial gap in our grasp of its oncogenesis, treatment strategies, and prognostic considerations.

The present article introduces two new cases of NUT sarcoma; comprehensively reviews all published cases; and aims to foster a discussion regarding the prognosis, available treatment options, and pathogenesis of NUT sarcomas. This comprehensive examination is designed to invigorate drug development efforts for this rare cancer type. Clinical considerations and targeted therapeutic strategies should consider the biological effect of each NUT partner gene described to date.

The patients whose case reports are presented in this article consented to publication of their deidentified clinical data.

A 55-year-old woman with a medical history notable for resected skin melanoma on the left hip, Hashimoto thyroiditis, and minimally symptomatic CREST (calcinosis, Raynaud phenomenon, esophageal dysmotility, sclerodactyly, and telangiectasia) syndrome presented to the emergency department with complaints of back pain.

The patient's evaluation included an abdominal and pelvic computed tomography (CT) scan, which revealed a hypoenhancing mass involving the central left kidney, measuring 4.2 × 2.9 cm. The mass showed infiltration into the midpole calyx and renal parenchyma, prompting concern for urothelial carcinoma. Metastasis was ruled out through a positron emission tomography (PET)-CT scan. Subsequently, a CT-guided core biopsy was conducted.

The biopsy pathology report indicated a high-grade undifferentiated malignant neoplasm displaying sarcomatoid and rhabdoid features. A radical right nephrectomy was performed. The subsequent pathology report diagnosed a malignant epithelioid neoplasm infiltrating the renal parenchyma and adjacent soft tissue. The tumor measured 4.2 × 3.5 × 2.3 cm. Tumor necrosis was evident in 10% of the sample, and a mitotic rate of 12 per 10 high-powered fields (HPFs; ∼2.2 mm²) was noted. Lymphovascular and perineural invasion were present and resection margins were negative. One local lymph node and 3 retroperitoneal lymph nodes were negative for malignancy. Morphologically, the tumor displayed poorly differentiated epithelioid cells with eccentric nuclei and prominent nucleoli. These cells were organized into nests, cords, and single cells. The differential diagnosis included high-grade sarcoma versus renal cell carcinoma with sarcomatoid differentiation (Fig. 1A). Immunohistochemical (IHC) analysis revealed positive staining for PAX8 (rare), cathepsin K (focal weak), pancytokeratin cocktail, vimentin, desmin, estrogen receptor, Melan A, and INI1/SMARCB1. The Ki-67 stain exhibited a proliferative index of approximately 60%. Conversely, tumor cell stains for PAX2, GATA3, CAIX, PR, EMA, CK7, CK20, CK5/6, P63, TTF1, CD10, RCC, ALK1, S100 protein, SOX10, HMB45, mammaglobin, GATA3, WT1, ERG, SMA, and caldesmon were negative. Subsequent sarcoma gene fusion panel testing revealed the presence of an MXI1::NUTM1 fusion. The patient was thus diagnosed with a MXI1::NUTM1 rearranged sarcoma arising from the left kidney. A PET-CT scan performed 6 months after surgery showed no fluorodeoxyglucose (FDG)-avid lesions suggestive of recurrent or metastatic disease confirming a remission period of 6 months. The patient is currently on surveillance.

A 67-year-old man presented with a palpable painful left groin mass. A CT scan revealed a 6.2-cm left pelvic mass located between the abdominal wall and the reservoir component of a penile implant. The imaging features were nonspecific, and PET-CT ruled out metastasis.

A biopsy of the mass noted a poorly differentiated malignant tumor with epithelioid features. IHC analysis demonstrated positive staining for vimentin and negative staining for pan-cytokeratin cocktail, EMA, synaptophysin, NKX3.1, CD45, MUM1, CD163, myogenin, MyoD1, and SOX10. Stains for kappa and lambda were noncontributory. Special stains for GMS and AFB yielded negative results.

The mass was subsequently resected via radical orchiectomy including the pelvic mass, penile prosthesis reservoir, left testicle, and spermatic cord. The pathology report noted poorly differentiated sarcoma with epithelioid and plasmacytoid features involving fibrous tissue (Fig. 1B). The tumor measured 7.1 cm, exhibited 5% tumor necrosis, and a mitotic rate of 6 per 10 HPFs or approximately 2.2 mm². Lymphovascular invasion was present and resection margins were free of tumor. The testis displayed areas of atrophy and fibrosis without tumor involvement. Histologically, the mass displayed diffuse growth of cells with small round, epithelioid, and plasmacytoid and rhabdoid morphology within a myxoid background and scattered areas of cartilaginous differentiation. IHC staining demonstrated positive results for vimentin, CD56, and BCL2, with focal positivity for S100 protein, CD99, and GFAP. Negative staining was observed for cytokeratin, EMA, desmin, SMA, myogenin, CD34, CD117, beta-catenin, synaptophysin, NSE, WT1, NKX3.1, and others. Elevated Ki-67 proliferation index (up to 50%) was noted. Molecular studies identified a CIC::NUTM1 fusion. Thus, a diagnosis of CIC::NUTM1 rearranged high-grade sarcoma of the pelvis was established.

The patient initially received 1 cycle of vincristine, doxorubicin, and cyclophosphamide; however, the subsequent multidisciplinary consensus recommendation was to withhold further chemotherapy and perform postoperative radiation therapy. Radiation was successfully completed 5 months after resection. CT scan 8 months after resection revealed a new intramuscular mass in the left abdominal wall musculature, measuring 11 cm, that extended to the contiguous right postsurgical changes. An ultrasound-guided biopsy was performed with pathology revealing a metastatic poorly differentiated sarcoma with epithelioid and plasmacytoid and rhabdoid features consistent with the patient’s known primary.

The case was brought before the tumor board at The University of Texas MD Anderson Cancer Center, and it was decided to pursue systemic therapy due to rapid local progression that rendered immediate surgery impractical. Despite the initial recommendation favoring trabectedin, the patient underwent two cycles of single-agent doxorubicin at 75 mg/m² with dexrazoxane at the patient’s local hospital. Subsequently, the treatment approach was modified to combination therapy with doxorubicin at 60 mg/m² bolus plus dexrazoxane on day 1, along with trabectedin at 1.1 mg/m² on day 1 of a 21-day cycle. The patient received 2 doses of this regimen. Restaging scans conducted 3 months after the beginning of treatment revealed stable disease with areas of necrosis within the tumor, signaling a favorable response (Fig. 2).

The literature search was conducted on Nov 15, 2024, using the PubMed database with the search term NUT sarcoma. Articles deemed relevant to the topic were selected, and additional references cited within those articles were also included to broaden the scope of the review. Our review of the literature revealed 61 cases of NUT-rearranged sarcoma published in addition to the two cases we presented herein (Supplemental Table S1).

Among the 63 patients, the median age was 40 years, and there was a male-to-female ratio of 1:1.03. The most frequent primary locations were the colon (9 patients), lung (6 patients), and kidney (4 patients). The most prevalent NUT partner genes were the MAD family in 52% of patients (33 patients total, including MGA [n = 12], MXD4 [n = 12], MXD1 [n = 2], and MXI1 [n = 7]), CIC in 30% of patients (n = 19), and bromodomain proteins in 8% of patients (n = 5 patients, including BRD4 [n = 4] and BRD3 [n = 1]). Other fusion partners were BCORL in one patient, ZNF532 in another, and NSD3 in a third patient. The gene partner was not reported for three of the patients. The median age for patients with MAD translocations was 47 years, whereas for CIC, it was younger at 18 years. The male-to-female ratio was 1:0.36 for patients with CIC translocations; however, the ratio was more favored toward female patients at 1:1.75 for MAD translocations. Among patients with MAD, 18% (6 of 33) were reported to have died of the disease, whereas 57% (11 of 19) of patients with CIC translocations were reported to have died of the disease.

The histologic features across the pathology samples were diverse, epithelioid, rhabdoid, spindle, round, and fibrosarcomatous cells, with many tumors exhibiting mixed histology. Of the pathology samples reported, 30 showed round cell morphology: 14 in CIC-rearranged tumors, 10 in MAD-rearranged tumors, four in BRD-rearranged tumors, and two with other rearrangements. A total of 14 pediatric cases (age < 18 years) were identified and included eight with CIC rearrangement, four with MAD rearrangement, and one with BRD rearrangement. Primary NUT sarcoma tumors originated from various locations, including the gastrointestinal system (14 patients), lung and pleura (8 patients), kidney (4 patients), brain/dura (4 patients), and spine/skull (10 patients), among others. Notably, all tumors arising from the spine or skull were CIC-rearranged.

Thirty-eight of 63 patients had localized disease at diagnosis and underwent surgical intervention; however, 19 of these cases experienced relapse. Radiotherapy was used in 10 patients: eight in combination with surgery and/or systemic therapy with a curative intent, and two patients received radiation with palliative intent. Twenty-one (33%) patients underwent systemic therapy with chemotherapy (e.g., vincristine plus doxorubicin plus cyclophosphamide and ifosfamide plus etoposide [VAC-IE regimen] or epirubicin plus ifosfamide regimen). Other systemic therapies included targeted therapy (1 patient treated with anlotinib) or immunotherapy (2 patients were treated with pembrolizumab and a second patient was treated on clinical trial with autologous activated natural killer (NK) cells in combination with pembrolizumab, interleukin (IL)-15 and cyclophosphamide). Among the 21 patients who received systemic therapy, 12 received neoadjuvant/adjuvant treatment and nine patients were treated in the advanced setting. For those treated in the metastatic setting, three (33%) patients showed disease control as defined by stable disease or response to treatment. Response Evaluation Criteria in Solid Tumors (RECIST) 1.1 assessments were available only for one patient treated with the combination of trabectedin and doxorubicin, who exhibited stable disease.^[11] Another patient treated with ifosfamide, carboplatin, and etoposide (ICE) was reported to have “initially responded” to therapy but subsequently progressed after seven cycles of therapy.^[12] A third patient was reported “to be responsive to chemotherapy, as the primary tumor size shrank, and the metastatic pulmonary lesions were undetectable 6 months after the onset of chemotherapy.”^[13] Two patients receiving chemotherapy experienced progression, whereas the response type was not reported for 4 additional patients.

Out of the total patient cohort, 21 patients were reported to have died, with a median survival of 14 months. Conversely, 29 patients were alive at the time of the report publication with 17 showing no evidence of disease and 12 exhibiting evidence of disease at the time of report publication. Survival or progression data were not available for 13 of the patients.

NUT sarcoma is an extremely rare tumor, with only 63 reported cases documented in the literature. This malignancy exhibits a slightly higher incidence among women and tends to manifest at a relatively young age. Notably, 60% of the NUT-rearranged sarcomas are diagnosed in early stages, although half of the reported cases experienced relapse despite curative-intent surgical intervention. Overall survival of the evaluable patients was 13.4 months. Furthermore, there is limited evidence on the efficacy of systemic therapies in the treatment of this tumor type, although trabectedin, ifosfamide, and doxorubicin-based regimens may play a role in this setting. Consequently, there is an urgent and compelling need for the development of effective systemic therapies for NUT sarcoma.

NUT carcinoma is often diagnosed at an advanced stage, occurs more frequently in men, and has a median overall survival of 6.7 months.^[6] However, formal comparisons between the two entities are not feasible because of the limited available data. Furthermore, specific guidelines for the systemic treatment of NUT carcinoma are being developed based on accumulating clinical experience.^[7] NUT sarcoma is an underdiagnosed malignancy, prompting us to advocate for the inclusion of NUT IHC as a screening method in patients presenting with sarcoma lacking a specific subtype diagnosis. A positive NUT IHC result should prompt further confirmation with fusion partner gene detection using DNA and preferably RNA-based next-generation sequencing (NGS). Alternatively, centers with ample resources may opt to perform NGS from the outset, eliminating the need for IHC screening. The primary gene partners associated with NUT sarcoma are the MAD family (MGA, MXD4, and MXD1), along with the CIC gene. It is notable that bromodomain proteins (BRD4 and BRD3), which are the favored fusion partners in NUT carcinoma, are infrequent in NUT sarcomas.

Although the numbers of cases are too small to draw statistically significant differences observed in age, gender distribution, and mortality rates between patients with MAD translocations and those with CIC translocations, the trends raise crucial clinical considerations. The older median age and higher female-to-male ratio in patients with MAD translocations emphasize distinct demographic characteristics. In addition, the contrast in mortality rates, 18% for MAD and 57% for CIC translocations, respectively, shows a trend toward increased disease aggressiveness in CIC fusion partner tumors.

Histological examination remains the cornerstone of diagnosis and the only method to differentiate NUT sarcoma from other NUT-rearranged neoplasms, including NUT carcinoma. Histological features include spindle or epithelioid cells displaying rhabdoid or sarcomatoid characteristics, with many cases featuring round cells, particularly in CIC::NUTM1 rearranged sarcomas. Most NUT sarcomas are histologically categorized as high grade, although isolated case reports have been classified as intermediate- or low-grade sarcomas. Importantly, even those cases exhibiting lower or intermediate-grade characteristics tend to behave aggressively, often progressing to metastatic disease.^[14–17] Consequently, it is important to approach all NUT sarcomas as high-grade malignancies, irrespective of their histological grading.

The clinical experience with systemic therapies in NUT sarcoma is limited. Among the 13 documented patients with NUT sarcoma treated with systemic therapies, we present a case involving the combination of trabectedin and doxorubicin, which achieved stable disease. Another case treated with ICE showed an initial response but subsequently progressed after seven cycles. A third patient with NUT sarcoma was reported to be responsive to chemotherapy; however, the chemotherapy type was not reported. In contrast, more data exist for NUT carcinoma, guiding the first International Symposium on NUT Carcinoma to recommend frontline chemotherapy, particularly ifosfamide regimens.^[7] Given the limited experience with NUT sarcoma, establishing guidelines for frontline systemic treatment recommendations remains challenging. However, trabectedin-based or ifosfamide-based regimens, especially in combination with doxorubicin (the backbone of soft tissue sarcoma), may represent a reasonable approach to managing these tumors in the absence of clinical trials. Given the potential variance in pathogenesis and tumor response to treatments between NUT carcinomas and NUT sarcomas, therapeutic approaches should be molecularly tailored toward the distinct oncogenic mechanisms driving these entities (Fig. 3).^[18] Clinical trials focused on NUT sarcomas and NUT carcinomas can enhance collaborative efforts to develop new therapies rooted in molecular targeting; however, results should be interpreted with caution, given the distinct biological differences between sarcomas and carcinomas.

The pathogenic models for BRD4 and BRD3::NUTM1 tumors are better elucidated compared with tumors involving other NUT fusion partners. The physiological bromodomain (BRD) protein functions as the “reader” of acetylated histones by interacting with its two bromodomains (located at the N terminus),^[19] facilitating the recruitment of transcription factor proteins through its extraterminal domain (situated at the C terminus).^[20] One of the factors recruited is positive transcription elongation factor (PTEF) beta, which phosphorylates RNA polymerase, ultimately promoting gene transcription.^[21] Conversely, NUT, primarily expressed in germ cells, plays a dual role in modifying acetylation patterns that ultimately facilitate chromatin compaction. It engages in acetylating specific genes by recruiting the EP300 histone acetyltransferase,^[22] while concurrently promoting deacetylation in most other genes through its histone deacetylase (HDAC) function.^[23] Therefore, the BRD::NUTM1 fusion drives a dysregulated transcription pattern within cells. On the one hand the fusion fosters focal histone hyperacetylation and subsequent transcription supported by the BRD protein in the designated “megadomains” at gene promoters, such as MYC, which finally leads to increased proliferation. Simultaneously, the fusion facilitates a broad-scale gene deacetylation, hampering the proper differentiation of cells.^[24,25] In this context, BET inhibitors exert their effect by disrupting the interaction between BRDs and acetylated lysine through competition (small molecules mimicking acetylated lysines) or directly inhibiting the BRDs (e.g., BRD inhibitors, dual BRD inhibitors, and BRD4 selective inhibitors), ultimately preventing uncontrolled proliferation.^[26] Among the molecules currently in development, the BET inhibitor molibresib showed a 21% response rate in NUT carcinomas in a phase 1 clinical trial.^[10]

The pathogenesis of MAD family proteins fused to the NUT protein differs significantly from those of bromodomain proteins. Under normal physiological conditions, MAD proteins engage in competition with MYC proteins to heterodimerize with MAX. The MYC-MAX complex binds to promoter regions of genes known as E-boxes, promoting the transcription and proliferation of the target MYC genes, which include SOX2, TP63, and MYB.^[27] In contrast, the MAD-MAX heterodimer also binds to E-boxes, the same target of MYC genes, but represses their transcription.^[28] However, when MAD proteins are fused to the NUTM1 gene, the NUT component promotes focal acetylation on the target MYC genes, subsequently facilitating transcription of these genes.^[29] Therefore, patients with MAD::NUTM1 fusions do not exhibit MYC overexpression, but they do express the target MYC genes.^[30] Here, BRD proteins may play a pivotal role in “reading” these abnormal acetylation patterns and promoting transcription, even when they are not part of the fusion. Therefore, although some publications have suggested that BET inhibitors may not be effective in non-BRD fusion NUT tumors,^[30] we hypothesize that these tumors involve a comprehensive disruption of the cell's transcriptomic machinery, where BRD proteins serve as readers of this abnormal acetylation pattern ultimately promoting transcription. To date, MAD inhibitors have not been developed.

The pathogenesis of CIC::NUTM1 fusions remains largely unknown. However, it is believed to bear a closer resemblance to CIC::DUX4 tumors than to the sarcomas with other NUT fusions, primarily due to the similarity in gene expression signatures observed between CIC::NUTM1 tumors and CIC::DUX4 tumors.^[31] In this study, we showed that CIC-rearranged NUT sarcomas more frequently exhibited round cell features, primarily affected the pediatric population, and were more likely to originate from the skull or spine compared with other NUT sarcomas. These characteristics are commonly associated with CIC::DUX4 sarcoma.^[32] Physiologically, CIC serves as a transcriptional repressor of E twenty-six (ETS) variant transcription factor (ETV) 1/4/5 genes and these ETV 1/4/5 genes typically function as transcription factors, binding to ETF-binding motifs present in the promoters of various target genes.^[33] The CIC portion of the fusion may attach to ETV genes, whereas the NUT portion may hyperacetylate these genes via EP300.^[34] This process may activate the transcription of the ETV proteins that ultimately will transcribe multiple target genes.

For our patient in case 2 with the CIC::NUTM1 fusion, we theorized that trabectedin might offer a potential therapeutic option, as trabectedin may play a role in inhibiting transcription through the blockade of RNA polymerase II, inhibiting DNA separation, and displacing the CIC::NUTM1 fusion from their target gene promoters. These mechanisms of action have been substantiated in preclinical studies for FUS::DDIT3 fusions seen in myxoid liposarcoma.^[35] In addition, several clinical trials and retrospective studies have shown that trabectedin is effective in translocation-associated sarcomas.^[36–38] However, there was no superiority of trabectedin when compared with doxorubicin-based chemotherapy in first-line treatment in a phase 3 clinical trial.^[39]

An alternative therapeutic approach for NUT sarcomas could involve targeting the NUT portion of the fusion using various strategies. This approach may demonstrate effectiveness in treating NUT sarcomas regardless of the specific fusion partner. One potential strategy could be aimed at mitigating the hyperacetylation induced by the NUT fusion by targeting the function of the EP300 histone acetyltransferase with molecules such as A-485.^[40] Furthermore, a complementary approach might entail targeting HDAC activity present in the NUT portion of the fusion gene, using HDAC inhibitors like vorinostat, with the goal of supporting the proper differentiation of cells. Encouragingly, two cases of NUT carcinoma reported partial response with vorinostat,^[23,41] although its effectiveness in treating NUT sarcoma requires evaluation.

Given the significant insights gained from the establishment of the NUT carcinoma registry and the collaborative efforts to study these rare tumors, it would be highly beneficial to create a dedicated NUT sarcoma registry.^[42] Such a registry could facilitate the collection of clinical data, promote research collaboration, and enhance our understanding of NUT sarcoma, ultimately leading to improved diagnosis, treatment options, and outcomes for patients.

In conclusion, this comprehensive literature review presents two cases of NUT sarcomas and examines an additional 61 cases, highlighting the rarity of this disease. This article underscores the importance of prompt diagnosis through IHC and/or NGS testing, advocates for the establishment of a NUT sarcoma registry, and emphasizes the urgent need for clinical trials to drive drug development. Promising biology-driven targeted therapies, such as BET inhibitors, histone acetyltransferase inhibitors, and HDAC inhibitors, should be further explored. In the absence of clinical trials, the findings from this review suggest that trabectedin-based or ifosfamide-based regimens, particularly when combined with doxorubicin, may offer a reasonable approach for frontline therapy in NUT sarcomas.

Supplemental materials are available online with the article.