UTROSCTs are rare neoplasms of unclear histogenesis that are typically located in the uterine corpus and rarely in the endocervix. The morphological features of UTROSCT overlap with those of ovarian sex-cord tumors. The World Health Organization classifies UTROSCTs in the category of “endometrial stromal and related tumors” . However, genetically, these tumors are separate entities that harbor different molecular abnormalities and have different prognosis. The majority of low-grade endometrial stromal sarcomas (LG-ESSs) harbor a fusion gene, most commonly involving JAZF1, SUZ12, and/or PHF1 genes, i.e., JAZF1-SUZ12 fusion (80%), JAZF1-PHF1 fusion (6%), or EPC1-PHF1 fusion (4%) [10-12]. The rare fusions of MEAF6-PHF1 (3%) and MBTD1-CXORF67 (2%) have also been reported in LG-ESSs [12-14]. In contrast, high-grade endometrial stromal sarcomas (HG-ESSs) have been found to harbor YWHAE-NUTM2A/B or ZC3H7B-BCOR fusion genes and to exhibit more aggressive behavior [15-18].
Recently, characteristic fusions involving ESR1 and GREB1, which are key factors in sex hormone pathways, were identified in 36 cases of UTROSCT; these included ESR1-NCOA3 (N = 15), ESR1-NCOA2 (N = 8), GREB1-NCOA1 (N = 5), GREB1-NCOA2 (N = 4), GREB1-CTNNB1 (N = 1), GREB1-NR4A3 (N = 1), GREB1-SS18 (N = 1), and NCOA2 rearrangement with an unknown partner gene (N = 1) [1-4, 19, 20]. Although most cases exhibit benign behavior, UTROSCT is considered to have uncertain malignant potential owing to its low rate of recurrence (5.9%) . However, a recent large series reported that up to 23.5% of UTROSCT cases develop extra-uterine recurrence . The intrinsic molecular mechanisms of UTROSCT that underly its different clinical behaviors are still unclear. In the current case of recurrent UTROSCT in a postmenopausal woman with GREB1-NCOA2 fusion gene, it is possible that the GREB1 rearrangement contributed to the aggressive behavior.
The clinical symptoms of UTROSCT are not specific. Typically, UTROSCTs are well circumscribed but unencapsulated. True myometrium invasion is uncommon but can be present in some cases and may be correlated with aggressive behavior. The tumor cells are organized in sheets, nests, trabeculae, or hollow or solid tubules with a repetitive pattern of cordlike or tubular growth; more rarely, the cells will have retiform or glomeruloid appearance or papillae and a solid pattern predominance. Neoplastic cells are small, round, ovoid to spindle with monotonous nuclei, inconspicuous or small to medium with distinct nucleoli, and occasionally have vascular invasion, heterologous elements, or necrosis. Recently, three cases of UTROSCTs with extensive rhabdoid differentiation and malignant behavior have been reported UTROSCTs exhibit a co-expression of epithelial, smooth muscle, sex-cord markers, and steroid receptors.
GREB1-rearranged uterine tumors often show prominently fascicular spindle cells or trabecular/cord-like arranged epithelioid cells and an inconspicuous immunophenotype of sex-cord differentiation . Similarly, our case showed predominantly diffuse sheets of short spindle cells and nests of epithelioid cells in the primary tumor. Immunohistochemically, the tumor cells were positive for epithelial markers (AE1/AE3, EMA), a smooth muscle marker (desmin), and less specific sex-cord markers (CD56, WT-1, CD99), but negative for relatively specific sex-cord markers (α-inhibin, calretinin, FOXL2, and SF1). Although without well-formed tubules or retiform structure, the extensive sex-cord-like patterns in our case, including nests and trabeculae/cords, strongly suggested the diagnosis of UTROSCT rather than other mesenchymal tumors, such as LG-ESS with sex-cord differentiation. The details of five cases of GREB1-rearranged UTROSCT reported previously  contrasted with the cases described by Cheng-Han et al. and with the current case, because all five cases in that series had prominent sex-cord differentiation and were positive for α-inhibin and/or calretinin immunostaining. Whether GREB1-rearranged UTROSCT had specific morphological and immunohistochemical features should be investigated further when additional cases have been reported.
Regarding prognosis, GREB1-rearranged tumors tended to occur in signiﬁcantly older women than did UTROSCT with ESR1 fusions (51–70 years old, median: 65 years versus 34–55 years, median 47 years) ; moreover, GREB1-rearranged tumors tended to be larger and more mitotically active (0–14 versus 0–1) and appeared to behave more aggressively. The current case was a 57-year-old postmenopausal woman with a 10.0-cm uterine mass and three mitotic figures per 10 high-powered fields in the primary and recurrent tumors. The tumor cells were determined by FISH to have NCOA2 rearrangement and by RNA sequencing to harbor a GREB1-NCOA2 fusion. These pathological and clinical features of our case were consistent with the cases described by Lee et al. and provided evidence for the recurrent aggressive behaviors of GREB1-rearranged uterine tumors.
Seven cases of GREB1-rearranged uterine tumors were previously summarized (4); of these, six cases had follow-up information, and two cases developed pelvic dissemination and/or lung metastasis with more than 1 year of follow-up. A recent report  described 26 cases of UTROSCT with NCOA1–3 rearrangement. Of these, five cases were GREB1-rearranged UTROSCT. The only recurrent case in that report harbored GREB1-NCOA2 fusion. The current case is the fourth known recurrent case with GREB1 rearrangement in the literature. The tumor recurred 30 months after a total hysterectomy with bilateral salpingo-oophorectomy.
To date, 12 cases of GREB1-rearranged uterine sarcoma have been reported in the English literature [1-4, 19]; the current case is the 13th known case. The clinicopathological features of all known GREB1-rearranged tumors, including our case, are summarized in Table 1. The mean age of the patients was 64.8 years (range, 51‒74 years). Follow-up information was available for nine cases (69.3%, 9/13), with a mean follow-up interval of 23.7 months (range, 0.75–66 months). Of these nine cases, four cases were recurrent or metastasized cases (44.4%). Of these, three tumors recurred in the pelvis (at 17, 30, and 66 months) and one metastasized to the lung after the initial diagnosis (at 24 and 132 months). The other five patients were alive without evidence of disease at the last follow-up (range, 0.75–54.3 months). The average tumor size was 9.1 cm (range, 4.2‒14.9 cm). Microscopically, the tumor growth pattern of most tumors (10/12, 83.3%) was infiltrative growth (including three focal invasions; cases 2, 3, and 4), and two were well demarcated. Mitotic figures ranged from 0‒14 per 10 high-powered field (2 mm2; mean: 3.6). Lymphovascular invasion and necrosis were each present in two cases (2/12, 16.7% and 2/7, 28.6%, respectively).
Of these 13 cases of GREB1-rearranged uterine tumors, eight cases (including our case) had a definite pathological diagnosis of UTROSCT, with typical morphological and/or immunohistochemical features. Five patients were initially diagnosed as having undifferentiated uterine sarcoma with nuclear uniformity, GREB1-rearranged sarcoma, atypical mesenchymal tumor, and sarcoma, not otherwise classifiable (cases 1, 2, 3, 4, and 6, respectively) (Table 1). Cheng-Han reported four cases (cases 1, 2, 3, and 4) as “GREB1-rearranged uterine sarcomas with variable sex-cord differentiation” because of limited sex cord differentiation and more aggressive clinical behavior. Notably, of these five cases, three cases (cases 2, 3, and 4) showed a focal trabeculae/cord pattern, which is an important morphologic clue for UTROSCT diagnosis. Except the four cases reported by Cheng-Han et al., seven patients with GREB1-NCOA2/NCOA1 fusion genes were diagnosed as having UTROSCT by authors (cases 5, 8, 9, 10,11, and 12), which suggested that at least a considerable proportion of these tumors with GREB1-rearrangement showed typical morphologic features of “UTROSCT”. Recently, three cases of UTROSCT with extensive rhabdoid differentiation, malignant behavior, and ESR1-NCOA2 fusions have been reported . Whether these tumors with specific gene rearrangement and aggressive clinical behavior should be regarded as poorly differentiated UTROSCT or as distinct uterine sarcomas need to be further investigated when additional cases have been reported. However, two cases (cases 1 and 6) showed morphologic features of high-grade sarcoma, predominantly composed of diffuse sheets of epithelioid, spindle, or polygonal cells and without conspicuous sex-cord–like patterns or the sex-cord-like differentiation immunophenotype. This could indicate that molecular detection, including the GREB1 rearrangement, should be used with cases of HG-ESS, undifferentiated sarcoma, and other high-grade sarcomas, which cannot be classified as any specific uterine sarcoma subtype, to clarify whether they are poorly differentiated UTROSCT or a distinct sarcoma subtype.
A systematic review by Blake et al. and a large case series by Moore et al. reported GREB1-rearranged uterine tumor recurrence rates of 5.9% and 23.5%, respectively; however, neither of these reports addressed the genetic ﬁndings [5, 6]. According to Moore et al, cases of UTROSCT that behaved in a malignant manner were associated with older age, larger tumor size, and higher mitotic activity . These aggressive and/or older patients included in the study by Moore et al. may have had a high frequency of GREB1 rearrangement or ESR1-NCOA2 fusions; however, the study did not address the molecular events. Additionally, the relatively higher recurrence rate in the series by Moore et al. may in part reflect a referral bias related to consultation cases in their series, which more often presented with unusual clinical behavior. Of these recurrent GREB1-rearranged cases, 3/4 harbored the GREB1-NCOA2 fusions and 1/4 the GREB1-CTNNB1 fusions. Interestingly, of the 4 UTROSCT cases harboring ESR1-NCOA2 with follow up information, three cases showed recurrence, as reported by Jennifer A. Bennett et al [1, 20]. In contrast, no cases showed recurrence among the 10 cases of ESR1-NCOA3 positive UTROSCT with follow up information . These limited recurrent cases specify that uterine tumor/sarcoma with GREB1-rearrangement or ESR1-NCOA2 fusions may have a high risk of recurrence/metastasis. Although the case number is limited, the higher recurrence rate of GREB1-rearranged or ESR1-NCOA2 fusion uterine tumors could indicate that these types have a greater tendency of aggressive behavior. Additional cases are needed to clarify the correlation between morphologic features, prognosis, and intrinsic molecular events.
Due to the rarity of this type of tumor, there is no well-established treatment protocol for UTROSCT, as there are limited data available for guiding clinical management. The initial clinical treatment strategies for UTROSCT include tumor resection, total abdominal hysterectomy alone, or total abdominal hysterectomy with bilateral adnexectomy, in consideration of patient age and parity. Recurrent cases are managed by repeated surgical procedures, and no evidence supports the usefulness of chemotherapy for recurrent cases . Additional investigation of the correlation between specific molecular events and prognostic significance is essential for the development of therapeutic strategies, especially for tumors with aggressive behavior.
GREB1 encodes growth regulation by estrogen in breast cancer 1, a protein transcriptionally driven by estrogen-bound ER. Furthermore, GREB1 is a key factor of the canonical estrogen/ER signaling pathway . Functionally, GREB1 is one of the most important tamoxifen/RU486-suppressed ER pathway downstream effectors, which indicates that GREB-rearranged tumors may respond to tamoxifen and RU486 [4, 22]. ESR1 encodes estrogen receptor 1, a ligand-dependent transcription factor. Binding with estrogen, ESR1 is not only essential for sexual development, reproductive function, and bone formation but is also involved in pathologic processes, including breast cancer, endometrial cancer, and osteoporosis. Several mutations in the ligand-binding domain of ESR1 have been found to be correlated with resistance to hormone therapy in ER-positive breast cancer [23, 24]. ESR1-rearranged UTROSCT may not be susceptible to an estrogen blockade because the ER ligand-binding domain is lost in fusions involving ESR1 (4). The partner genes of GREB1- or ESR1-rearranged UTROSCT, including NCOA1–3, NR4A3, SS18, and CTNNB1, all encode transcription factors, and the basic function of each, as reported previously , is listed in Table 2. Five cases of UTROSCT that are suspected to be related to tamoxifen treatment have been reported ; however, the molecular features of these tamoxifen treatment-related cases were not determined. The complicated mechanism underlying chromosome translocation involving different functional genes, prognosis, and response to tamoxifen is still unclear. Although hormonal treatment has not yet been explored for UTROSCT, such in vitro and in vivo research may be necessary.
In conclusion, we reported a case of recurrent UTROSCT in a 57-year-old woman with the GREB1-NCOA2 fusion gene. This case provides further evidence that uterine tumors with GREB1-rearrangement may have a high recurrence/metastasis risk. Further studies are necessary to elucidate the clinical features of UTROSCT, especially the prognosis, potential treatments, and the range of possible molecular events.