Clinical characteristics and outcomes for children, adolescents and young adults with “CIC‐fused” or “BCOR‐rearranged” soft tissue sarcomas: A multi‐institutional European retrospective analysis

Abstract Background In certain rare undifferentiated small round cell sarcomas new specific molecular CIC‐DUX4/other partner, BCOR‐CCNB3/other partner, YWHAE fusions, or BCOR‐ITD (internal tandem duplication) were identified. These new “CIC fused” (CIC‐fused/ATXN1::NUTM1) and “BCOR rearranged” (BCOR fused/ITD/ YWHAE) soft tissue sarcomas (STS) are not well described. Methods Multi‐institutional European retrospective analysis of young patients (0–24 years) with CIC‐fused and BCOR rearranged STS. Results Overall, out of the 60 patients selected, the fusion status was CIC‐fused (n = 29), ATXN1::NUTM1 (n = 2), BCOR::CCNB3 (n = 18), BCOR‐ITD (n = 7), and YWHAE (n = 3), MAML::BCOR STS (n = 1). The main primaries were abdomen‐pelvic (n = 23) and limbs (n = 18). Median age was 14 years (0.9–23.8) and 0.9 (0.1–19.1) for CIC‐fused and BCOR‐rearranged groups, respectively (n = 29; p < 0.001). IRS stages were I (n = 3), II (n = 7), III (n = 35), and IV (n = 15). Overall, 42 patients had large tumors (>5 cm) but only six had lymph node involvement. Patients received mainly chemotherapy (n = 57), local surgery (n = 50), and/or radiotherapy (n = 34). After a median follow‐up of 47.1 months (range, 3.4–230), 33 (52%) patients had an event and 23 patients died. Three‐year event‐free survivals were 44.0% (95% CI 28.7–67.5) and 41.2% (95% CI 25.4–67.0) for CIC and BCOR groups (p = 0.97), respectively. Three‐year overall survivals were 46.3% (95% CI 29.6–72.4) and 67.1% (95% CI 50.4–89.3; p = 0.24), respectively. Conclusions Pediatric patients often present with large tumors and metastatic disease, especially CIC sarcomas. Overall outcome is dismal. New treatment options are needed.


| INTRODUCTION
A new subset of round/ovoid cell sarcomas histologically resembling tumors of the Ewing sarcoma family (ESFT) has been recently characterized in unclassified/undifferentiated small round cell sarcoma (USRCS or SRCSs). [1][2][3][4][5][6] Previously known as "Ewing-like sarcomas" (ELS), these tumors are currently distinguished from Ewing sarcoma in the current WHO classification. With molecular assays these tumors have been genetically characterized and are now under assessment if these entities may represent distinct biologic tumor entities. 7-10 Accordingly, a new subtype of bone sarcoma has been characterized by a BCOR::CCNB3 gene fusion. 11 This discovery offers new insight into oncogenesis and future new therapeutic considerations. Rare BCOR sarcoma was described in a group of small round cell bone tumors lacking the canonical EWSR1 translocation: only 24 BCOR::CCNB3 positive tumors could be identified in an analysis of 594 sarcoma. 11 Other "BCOR rearranged sarcomas" have been described with molecular characteristics as BCOR-ITD (internal tandem duplication). 12 These BCOR-ITD tumors are reported to be aggressive tumor entities with an ITD in the last exon of the BCOR gene. With location in the kidney, these sarcomas are described as clear cell sarcomas of the kidney (CCSK). 13 However, these sarcomas are also described as high-grade endometrial stromal sarcomas (HG-ESS) 14,15 in bone and soft tissue, referred to as USRCS, but also occurring in infants as primitive mesenchymal myxoid tumor (PMMTI). [16][17][18] CIC::DUX4 sarcoma (CDS) or CICrearranged sarcomas seem a rare subcategory of another small round cell sarcoma with histological signs of Ewing sarcoma (ES). 2 For extracranial sites, "CIC fused sarcoma" group had biological homology as the "BCOR rearranged sarcoma" group. 6 In recent analysis, ATXN1 tumors clustered with CIC tumors and YWHAE fused tumors within the BCOR tumors group. For the BCOR-ITD tumors no radiological, clinical, nor pathological specificity could be found, except for location in the central nervous system but a common transcriptomic signature with other BCOR rearranged sarcomas. 18 However, all these are described as a distinct entity from ES, and occur in various sites as soft tissue, visceral primaries and less often bone. 3,19 Therapeutic options for such tumors frequently include chemotherapy (CHT) and local surgery with radiotherapy (RT) analogous to treatment of soft tissue sarcomas or ES. 19 There are few data available regarding the clinical description of these tumors when the primary site is soft tissue. We describe the clinical characteristics, treatment, Foerderkreis Krebskranke Kinder; Fondazione Città della Speranza; S Wisnia Methods: Multi-institutional European retrospective analysis of young patients (0-24 years) with CIC-fused and BCOR rearranged STS.

| METHODS
This retrospective multi-institutional study registered all patients aged ≤24 years at the time of diagnosis with USRCS in soft tissues identified from local coordinators as CIC, BCOR, or YWHAE gene rearranged sarcomas. Location in bone, central nervous system, or kidney were exclusion criteria. Patients diagnosed between 1996 and 2021 with available clinical data and at least 3 months of follow-up were included. All specimen were reviewed by expert pathologists or by central review for those that were enrolled on clinical trials. Available molecular data were mandatory for all cases. 6 In clinical practice, diagnosis of the oncogenic fusion was performed using molecular techniques such as fluorescence in situ hybridization (FISH), reverse transcription polymerase chain reaction (RT-PCR), or targeted RNA sequencing. 6 Depending on tumor tissue availability, RT-PCR, next-generation RNA sequencing (RNA seq) or clustering, and methylation profile were performed on paraffin-embedded and/or fresh frozen tissue. 2,4,20 Recently, all cases of tumors with pathological Ewing characteristics that do not harbor the canonic EWS fusion transcript, were systematically analyzed with RNA sequencing and/or RT-PCR with specific combination of primer sets for CIC::DUX4 and BCOR and/ or FISH analysis with probes for CIC (19q13.2), for BCOR a dual color break apart probe (Xp 11.4) and for YWHEA (17p13.3). 11,21,22 The order of this techniques varies according to each institutional organization. For some old cases, analysis could have been performed retrospectively.
Due to their biological homology, all sarcomas with CIC::DUX4, CIC break, or ATXN1::NUTM1 fusion transcript were gathered in the "CIC fused sarcoma" group and sarcomas with BCOR::CCNB3, BCOR-ITD, YWHAE::NUTM2B, or MAML::BCOR fusion transcripts represent the "BCOR rearranged sarcoma" group. 6 Guardians of included patients had consented to data collection and retrospective chart review. This was performed per the requirements of the Declaration of Helsinki and in accordance with the regulations of the respective ethical committee (Ethikkommission Tübingen 413/2022BO2 from August 18, 2022).

| Definition of terms
Clinical staging was defined according to the Tumor Node Metastases (TNM) system: T1 or T2 according to the invasion of contiguous organs; N0/N1 according to the presence of lymph node metastases and M0/M1 of distant metastases. 23 Postsurgical staging was classified according to the International Rhabdomyosarcoma Study grouping (IRS-G) system. IRS-I corresponds to a complete tumor resection with microscopically complete margins, IRS-II corresponds to a macroscopic resection but invaded histologic margins and IRS-III corresponds to a macroscopic residual tumor after biopsy or immediate surgery. IRS-IV corresponds to metastatic disease. 24 Initial staging included imaging of the primary tumor and bone/lung metastases by magnetic resonance imaging (MRI) or computed tomography (CT). Whole body imaging was recommended with radionuclide bone scan or 18F-fluorodeoxyglucose positron emission tomography. Staging included bone marrow aspirate/biopsy. Tumor response was mainly assessed after 3-4 courses of CHT: complete response (CR), partial response (PR), or stable disease (SD). 25 Delayed surgical resection was defined if the resection took part after initiation of chemotherapy. "Extent of tumor resection" was defined as the best surgical result in the sum of any surgeries that were performed in one patient during the first-line therapy. Primary tumor resection was classified as microscopically complete (R0), microscopically incomplete (R1), or macroscopically incomplete (R2). 26 "Best tumor response" was the best available response at any time point after start chemotherapy. Progressive disease (PD) was defined as first event with in case of any increase in tumor volume in patients who did not achieve CR. 27

| Therapy
No dedicated protocols were available and the treatment strategy the individual decision of the treating physicians. Patients were classically treated according to Ewing's or soft tissue sarcoma (STS)/ rhabdomyosarcoma (RMS) protocols, with a combination of therapies including neoadjuvant and adjuvant CHT, surgical resection, and/or radiation therapy (RT) of the primary according to initial tumor extension and the response to initial chemotherapy. 28 Primary resection was performed if a non-mutilating procedure was possible. RT was given as individual decision of the treating center.

| Statistical analysis
Survival time was measured from the date of diagnosis (initial biopsy/surgery) to the time of last follow-up or event. Event-free survival (EFS) was defined as the time from diagnosis to the date of local relapse or any lymph node or metastatic tumor progression or death from any cause. Overall survival (OS) was calculated from the date of biopsy/surgery to death from any cause. Patients alive without local recurrence or metastasis at last contact were censored at the date of last follow-up. Kaplan-Meier method was used for survival curves and log-rank test to calculate the differences between survivals. The prognostic role of risk factors (gender, age at diagnosis [± median of the cohort], tumor size [±5 cm] and site, TNM status, IRS groups [IRS I-II vs. III vs. IV] and type of tumor groups [CIC-fused vs. BCOR groups]) were considered for their impact on EFS and OS by comparing Kaplan-Meier curves using log-rank tests and Cox univariate and multivariate models. All risk factors significant at least at p 0.25 in univariate analysis were included in the multivariable model; type of tumor group was forced in every model.
Patients with CIC sarcomas were more likely to receive primary surgery and irradiation, while BCOR tumors were more frequently treated with exclusive surgery (p = 0.04, Table 3).

| Statistical analysis and prognostic factors
In term of clinical characteristics, age >13 years was the only clinical factor significantly associated with a better EFS in univariate and multivariate analysis (Table 4). Notably, these risk factors were only present in BCOR/ YWHA tumors (p = 0.012) and not in CIC sarcomas (p = 0.84). Presence of metastases was the only factor significantly associated with a worse OS in univariate and multivariate analysis (Table 5; Figures S1 and S2). Males showed a significantly better OS in univariate analysis but not after adjustment for other risk factors (Table S3A,B).

| DISCUSSION
We reported the first clinical data of a European cohort of young patients with the novel genetically defined "CIC fused sarcomas" and "BCOR rearranged sarcomas" in soft tissues. "CIC fused sarcomas" were described in a cohort of 115 patients as often found in soft tissue sarcoma (86%) and mainly occurring in adults (median age 32 years). 9,22 This analysis includes data on children, adolescents and young adults. The definition of young adults (AYA) varies and mostly includes patients up to 24 years. However, inclusion of data of older patients in cooperation with the adult oncologists are undoubtedly needed to get an overview on a larger cohort concerning age and this data do not reflect the overall age at diagnosis.
In our study, "CIC fused" STS were mainly located in the trunk and are often presenting with metastatic disease. 29 In contrast, "BCOR rearranged sarcomas" are described as occurring more frequently in bone (64%) and in pediatric patients (76%) 30 : BCOR::CCNB was described more often in adolescents in contrast to BCOR-ITD, more frequently occurring in infants. [30][31][32] In our study, we cannot demonstrate any difference in the age of occurrence comparing BCOR::CCNB to other genetic alterations as BCOR-ITD or the YWHAE alteration in contrast to data referring to all locations. 31 Our cohort of patients harboring the BCOR rearrangement frequently had localized tumor, located in limbs and significantly younger at diagnosis than patients with "CIC fused" STS, confirming observations including locations in bone. 30 Interestingly, we confirm that age was a significant factor for EFS, especially for BCOR/YWHA tumors, but not the biological entity itself. The precise reason to explain this difference is not totally clear. However, as this finding was only present in BCOR/YWHA sarcomas, one hypothesis could be that, in some way, therapy differences may influence the outcome of patients. Patients with BCOR tumors are more frequently younger and may have receive different therapy (i.e., less radiotherapy for instance) that may explain such difference. These different genetic alterations seem to have a prognostic impact as patients with BCOR altered tumors are reported to have an overall poor prognosis. BCOR::CCNB3 fused tumors are reported to have a 5year OS of 68%, in contrast to patients with BCOR-ITD or ZC3H7B::BCOR tumors that have low survival rates, with more than 50% of patients in each group dying of relapse of their disease. 31 In our experience we report a much higher number of IRS group IV tumors in the "CIC fused sarcomas" and only two in the "BCOR rearranged sarcomas" at diagnosis. Yet, the EFS of "BCOR rearranged sarcomas" as a group is lower than that of the "CIC fused sarcoma" group. This suggests that effective therapy was likely not delivered due to the younger T A B L E 3 Summary of delivered therapy.
age in the "BCOR rearranged sarcoma" group and also probably explains the higher salvage rate at relapse given the better overall survival compared to that seen in the "CIC fused sarcoma" group. We recommend imaging guidelines developed for rhabdomyosarcoma as specific guidelines for CIC and BCOR sarcoma need to be developed. 33 The initial extent of the disease and the patients age at the time of diagnosis might be responsible for the inhomogeneous treatment. 34 The outcome of these patients is unsatisfactory with a 3year OS of 46.3% and 67.1%, respectively, but not statistically different (p = 0.24). These data correlate to the rare published data on these patient groups including bone primary: 5-year OS 43% for "CIC fused sarcoma" and 72% F I G U R E 1 Kaplan-Meier estimates presenting EFS of 60 patients with CIC-fused sarcoma (n = 31) and BCOR rearranged sarcoma (n = 29).

F I G U R E 2
Kaplan-Meier estimates presenting OS of 60 patients with CIC-fused sarcoma (n = 31) and BCOR rearranged sarcoma (n = 29).
for "BCOR rearranged sarcoma". 22,35 As in other STS, metastatic status was a clear prognostic factor for OS.
We are able to contribute detailed data on CHT and local treatment: Patients of both entities were more likely to be treated in Europe according to an STS protocol as (non-) rhabdomyosarcoma, rather than an Ewing one. However, the main drugs are similar in these protocols and the main difference lies on the total number of CHT courses. Multimodal strategy seems important but larger cohort are needed to evaluate the value of each therapy in such patients.
The main weaknesses of these data are their retrospective character, extracted from different databases or centers. As a limitation, the timing of the testing could not be evaluated. However, as patients are included diagnosed before 2015, few patients were included with the molecular diagnosis made some years after initial histologic diagnosis. The frequency of these biologically distinct cancers is extremely rare. Considering CWS patients from the European countries Germany, Switzerland, Austria, Poland, Sweden, and Finland, 11 patients with CIC-fused/ T A B L E 4 Risk factors in univariate and multivariate analysis for event-free survival for the population with CIC and BCOR sarcomas. BCOR rearranged STS of the total cohort of 247 patients with the histologic diagnosis of USRCC/UDS/ ES of STS irrespective of the molecular diagnosis were registered in the European CWS studies and the STS registry SoTiSaR in this time. As further limitation we are not able to report on numbers of patients with "CIC fused sarcoma" and "BCOR rearranged sarcoma" in bone, kidney, and CNS from all the institutions. To get a sense of the frequency and clinical data of these biologically distinct cancers larger international retrospective studies including studies and registries of bone, kidney, and CNS are necessary.

Number of patients
An international common strategy based on the somatic genetic background is undoubtedly required and a new international risk stratification considering the overall unfavorable prognostic of these entities needs to be built up.
These data also stress the need to perform additional molecular testing in case of any tumor with Ewing characteristics that do not harbor the canonic EWS fusion transcript. 36 Tumor genomic sequencing and assignment of genotype-based groups should be incorporated to help understanding of different clinical outcomes. 37 Finally, new treatment options need to be developed. Since some T A B L E 5 Risk factors in univariate and multivariate analysis for overall survival for the population with CIC and BCOR sarcomas. of them start to seem promising, international collaborations, to develop trials including new treatment strategies, are of highest importance in these rare diseases. 38,39 Applying a distinctive signature of 537 deregulated genes in CIC-rearranged sarcoma with RNA seq, Palmirini et al. in a recent paper showed that the CIC::DUX4 transcriptional profile compared that of ES was enriched in gene sets including DUSP4, ETV1/4, IGF2, and IGF2BPs, PTX 1 and 3 that are associated with activity of irinotecan, proteasome inhibitors (bortezomib and MG262), and histone deacetylase inhibitors (scriptaid and vorinostat). 40 Future common analyses on patients with these rare diseases occurring in bone and soft tissue sarcoma are needed to better compare different treatment regimen, to define prognostic factors and to develop new treatment strategies.