Intracranial Mesenchymal Tumor with FET-CREB Gene Fusion: A Case Report of Late Recurrence

Abstract

Intracranial mesenchymal tumors with FET‑CREB fusion represent a rare tumor entity newly defined in the 5th edition of the WHO classification of central nervous system tumors. Due to overlapping morphological features, these neoplasms were previously diagnosed under different names, including intracranial angiomatoid fibrous histiocytoma and intracranial myxoid mesenchymal tumor. Clinically, they need to be differentiated from meningioma, high‑grade glioma, solitary fibrous tumor, and other intracranial neoplasms. Most supratentorial lesions achieve good outcomes after gross total resection, and tumor recurrence is considered uncommon. We present a 58‑year‑old female patient who experienced local tumor recurrence 27 months after complete resection of a primary supratentorial cystic mass located in the right temporoparieto‑occipital region. The primary lesion was predominantly cystic with minimal solid components. Microscopically, oval and spindle‑shaped tumor cells infiltrated the cystic wall without obvious brain parenchymal invasion, accompanied by linear hemosiderin deposition. Immunohistochemically, tumor cells were diffusely positive for vimentin, MUC4, CD99, synaptophysin, EMA, and ALK (D5F3). Focal positivity was observed for CD68 and broad‑spectrum cytokeratin. All other markers including SMA, MyoD1, HMB45, myogenin, TFE3, MelanA, SOX10, SSTR2, PR, GFAP, and Oligo2 were negative. INI1 expression was intact in all tumor nuclei. The Ki‑67 proliferation index increased from 5%–10% in the primary lesion to 10%–15% in the recurrent specimen. Both primary and recurrent tumors harbored identical EWSR1::ATF1 fusion confirmed by next‑generation sequencing and FISH analysis. The patient received no adjuvant therapy after the first operation, and delayed recurrence was detected during routine MRI follow‑up. A combination of radiological findings, histopathology, immunohistochemistry, and molecular testing is essential for accurate diagnosis of this rare tumor. This case describes late tumor recurrence in an adult patient with typical low‑risk features, including supratentorial location and complete surgical resection. Our findings expand the clinical and pathological spectrum of EWSR1::ATF1‑fused intracranial mesenchymal tumors. Long‑term imaging surveillance is necessary even for patients with favorable clinical factors. Standardized follow‑up protocols and optimal adjuvant treatment strategies remain undefined and require further validation in larger cohorts.

Introduction

Intracranial mesenchymal tumors harboring FET‑CREB fusion are distinct non‑meningothelial neoplasms according to the 2021 WHO CNS5 classification.¹ These tumors arise from gene fusions involving FET family genes (EWSR1, FUS) and CREB family transcription factors (ATF1, CREB1, CREM).² Before the updated classification, inconsistent pathological diagnosis was common because of similar morphological appearances among different mesenchymal tumors.³ Subsequent epigenetic studies confirmed two molecular subgroups with significantly different prognostic behaviors.^3,4

Different fusion variants exhibit distinct clinical outcomes. EWSR1::ATF1 fusion is the most common subtype among intracranial FET-CREB tumors and is closely associated with unfavorable progression‑free survival, with a 2‑year PFS rate of approximately 62%. In comparison, tumors with EWSR1::CREB1 fusion tend to have a better prognosis.⁵ Established favorable prognostic indicators include adult onset, supratentorial location, and gross total resection. Infratentorial location, pediatric age, and subtotal resection are recognized high‑risk factors for recurrence.⁶ Radiologically, these tumors typically present as cystic‑solid masses with prominent peritumoral edema, often mimicking other common intracranial tumors such as glioma and meningioma.⁷ A multimodal diagnostic approach combining morphology, immunohistochemistry, and molecular detection is required for definitive diagnosis.⁸

Most previously reported recurrent cases occurred in high‑risk populations, such as pediatric patients or incompletely resected lesions. Recurrence in adult patients with completely resected supratentorial tumors is extremely rare. In this study, we report a delayed recurrence case with paired primary and recurrent tissues, aiming to provide additional clinical evidence for understanding the biological behavior of EWSR1::ATF1‑fused tumors.

Case Presentation

A 58‑year‑old female patient was admitted to our hospital in January 2023 due to persistent headache for three weeks. She had no nausea, vomiting, or neurological deficits. Cranial CT and MRI revealed a large cystic mass in the right temporoparieto‑occipital lobe with extensive peritumoral edema and imaging signs of uncal herniation.⁸ Radiological differential diagnoses included granulomatous lesion, meningioma, high‑grade glioma, lymphoma, and intracranial metastasis.

The patient underwent standard craniotomy. Intraoperative observation showed a well‑demarcated cystic lesion with small solid mural nodules. After cyst fluid evacuation, all solid tumor components were completely removed. Gross total resection was achieved with a clear tumor‑brain interface. The resected specimen measured 2.5 cm × 1.8 cm × 1.6 cm, with a thin cyst wall and pale yellow cystic fluid.

The patient’s headache resolved immediately after surgery. Postoperative cranial CT and MRI showed transient cerebral edema with mild uncal herniation, which gradually resolved within 20 days (Figure 1A–D). No permanent neurological dysfunction occurred. The patient did not receive radiotherapy, chemotherapy, or targeted therapy after the initial resection.

Figure 1.

Preoperative cranial CT and MRI features of the primary tumor. (A) Axial non-contrast CT shows a well-circumscribed hypodense cystic lesion with peritumoral edema. (B) Sagittal T1-weighted MRI demonstrates a large cystic hypointense mass. (C) Axial T2-weighted MRI shows extensive peritumoral edema. (D) Post-contrast T1WI reveals mural nodule enhancement within the cystic lesion.

Routine follow‑up MRI in April 2025, 27 months after surgery, detected a small recurrent nodule at the original tumor bed. The patient underwent a second craniotomy, and the recurrent lesion was totally resected. Outpatient clinical follow-up from April 2025 to May 2026 demonstrated no evidence of tumor recurrence.

Pathological and Molecular Findings

Formalin‑fixed paraffin‑embedded tissues from both primary and recurrent tumors were processed for histological staining, immunohistochemistry, and molecular analysis.

Histopathological Features of the Primary Tumor

Microscopically, the primary tumor showed nested and fascicular architecture with abundant hyalinized collagen fibers. Myxoid stroma was not prominent. Numerous thin‑walled capillaries formed pseudovascular structures within the lesion. Tumor cells comprised a single tumor cell population with two morphological variants (round-to-oval cells with eosinophilic cytoplasm and focal rhabdoid features, and epithelioid cells with clear cytoplasm), showing mild nuclear atypia and rare mitoses. No brain invasion was identified. The tumor was surrounded by a fibrous pseudocapsule with peripheral lymphocytic infiltration (Figure 2A–F).

Figure 2.

Histopathological features of the primary tumor (H&E staining). (A) An infiltrative growth pattern is evident in the absence of overt brain-parenchymal invasion. (×4). (B) A nodular architecture is surrounded by densely hyalinized, “asbestos-like” fibers. (×10). (C) In selected foci, a pseudoangiomatous pattern is observed, in which neoplastic cells encircle hyalinized vascular channels. (×10). (D) Scattered, variably ectatic vessels are interspersed throughout the tumor. (×20). (E) and (F) A single tumor cell population with two morphological variants (rhabdoid cells and epithelioid clear cells) is evident. The tumor cells exhibit round nuclei; cytoplasm ranges from clear to eosinophilic, and focal rhabdoid morphology is noted. Nuclear atypia is minimal, and definitive mitotic figures are not identified (×20).

Immunohistochemical Results

Immunohistochemical staining results were consistent between primary and recurrent lesions: diffuse positivity for vimentin, MUC4, CD99, synaptophysin, EMA, and ALK (D5F3); focal positivity for CD68 and AE1/AE3; and consistent negativity for SMA, MyoD1, HMB45, myogenin, TFE3, MelanA, SOX10, SSTR2, PR, GFAP, and Oligo2. INI1 nuclear expression was retained. The Ki‑67 index was 5%–10% in the primary tumor and increased to 10%–15% in the recurrent tissue (Figure 3, Table 1 and Table 2).

Figure 3.

Immunohistochemical staining profiles of the primary tumor. All key markers are shown with consistent staining results. INI1 shows intact nuclear expression.

Table 1.

Immunohistochemical Markers, Clones, Vendors, and Staining Results

Antibody	Clone	Source	Staining Pattern
Vimentin	V9	Quanhui	Diffuse positive
MUC4	8G7	MXB	Diffuse positive
CD99	EP8	Gene	Diffuse positive
Synaptophysin	LBP1-syn	AMP	Diffuse positive
EMA	LBP1-EMA	AMP	Diffuse positive
ALK	D5F3	Roche	Diffuse positive
CD68	LBP1-CD68	AMP	Focal positive
SMA	Asm-1	Quanhui	Negative
MyoD1	Mx049	MXB	Negative
HMB45	LBP1-HMB45	AMP	Negative
Myogenin	F5D	MXB	Negative
TFE3	MRQ-37	MXB	Negative
MelanA	LBP1-MelanA	AMP	Negative
SOX10	EP268	MXB	Negative
SSTR2	EP149	MXB	Negative
Cytokeratin	LBP1-AE1/AE3	AMP	Focal positive
PR	IE2	Roche	Negative
GFAP	GA5	Quanhui	Negative
Oligo2	LBP2-oligo2	AMP	Negative
INI-1	OTIR4G9	ZSBIO	Nuclear expression retained
Ki67	LBP2-ki67	AMP	Primary:5%–10%; Recurrent:10%–15%

Table 2.

Comparative Features of Primary and Recurrent Tumors

Feature	Primary Tumor (2023)	Recurrent Tumor (2025)
Cellular density	Moderate (500/mm2)	Increased (800/mm2)
Nuclear atypia	Mild	Mild
Mitotic activity	Rare	Rare
Ki-67 index	5–10%	10–15%
Fusion status	EWSR1::ATF1 positive	Identical EWSR1::ATF1 fusion
Immunophenotype	Consistent	Unchanged

Molecular Testing Results

Targeted next‑generation sequencing identified identical EWSR1::ATF1 fusion in both primary and recurrent tumor tissues. Break‑apart FISH analysis further confirmed consistent EWSR1 gene rearrangement in paired specimens (Figure 4).

Figure 4.

FISH detection of EWSR1 gene rearrangement. Dual-color break-apart FISH probes corresponding to the 5′ green and 3′ red segments of the EWSR1 locus were applied for molecular testing. Two hundred discrete tumor cell nuclei were randomly counted, with a rearrangement-positive rate of 30%. The institutional diagnostic criterion defined specimens with ≥15% split fluorescent signals as EWSR1 rearrangement-positive. Concordant EWSR1 break-apart signals were validated in primary and recurrent tumor tissues, confirming EWSR1 rearrangement consistent with the EWSR1::ATF1 fusion identified by NGS. (White arrows indicate tumor cell nuclei with split-apart signals of the EWSR1 gene, confirming EWSR1 gene rearrangement in these cells).

Radiological and Pathological Features of Recurrent Lesion

Twenty‑seven months after initial surgery, follow‑up MRI detected an abnormal nodule at the resection cavity. The recurrent lesion showed mixed signal intensity on T1WI and T2WI, with peripheral rim enhancement and internal nodular enhancement accompanied by obvious peritumoral edema⁷ (Figure 5A–C).

Figure 5.

Imaging and histopathology of the recurrent tumor. (A) T1-weighted image shows an irregular, heterogeneously iso- to hypointense lesion. (B) T2-weighted image demonstrates a mass-like iso- to hyperintense lesion with a peripheral hypointense rim. An adjacent oval cystic lesion with a small fluid-fluid level is also noted. Extensive perilesional edema is present. (C) Gadolinium-enhanced T1-weighted image reveals rim enhancement of the lesion with a nodular, markedly enhancing focus within. The adjacent cystic lesion shows peripheral enhancement. (D) Increased cellularity and diminished “asbestos-like” fibers. (×4). (E) A dense, circumferential lymphocytic cuff persists; no invasion into adjacent brain parenchyma is observed. (×4). (F) Foci retaining the pseudoangiomatous pattern exhibit increased vascular density. (×20). (G) The first cell population: round-to-oval cells with eosinophilic cytoplasm and rhabdoid features. (×20). (H) The second population comprises epithelioid cells with round nuclei and clear cytoplasm. Cytologic atypia remains mild, and mitotic activity is absent. (×20).

The recurrent lesion measured approximately 0.5 cm in maximum diameter. Histologically, it shared identical morphological features with the primary tumor, including consistent cell morphology and architectural pattern. Compared with the primary lesion, the recurrent tumor presented higher cellular density and elevated Ki‑67 proliferation index, without increased nuclear atypia or mitotic activity. Immunohistochemistry and FISH results confirmed identical immunophenotype and EWSR1 gene fusion, verifying true local tumor recurrence⁹ (Figure 5D–H). Table 2 summarizes the comparative pathologic findings and Ki-67 labeling indices between the primary and recurrent tumors.

Discussion

FET‑CREB fusion‑positive intracranial mesenchymal tumors are rare neoplasms with low overall recurrence rates. For adult patients with completely resected supratentorial lesions, recurrence is considered extremely uncommon.³ However, our patient developed delayed recurrence despite all favorable clinical factors, suggesting that EWSR1::ATF1 fusion may confer additional recurrence risk independent of traditional clinical factors, although this requires validation in larger cohorts.⁵

Previous recurrent cases were mostly reported in children, infratentorial lesions, or subtotally resected tumors. This case provides new evidence that EWSR1::ATF1‑driven tumors may recur even in low‑risk adult patients, expanding the biological spectrum of this rare tumor subtype.⁶

Epigenetic methylation profiling has been proven valuable for precise subtyping and prognostic evaluation. Two distinct methylation subgroups have been identified, with significantly different progression‑free survival.⁴ Unfortunately, methylation analysis was not available in our case due to institutional limitations, restricting further exploration of the epigenetic mechanism underlying tumor recurrence.

In a recent 2025 study, Panico et al reported a rapidly progressive intracranial mesenchymal tumor without FET‑CREB fusion, which achieved long‑term disease control after combined surgical resection and particle radiotherapy.¹⁰ Combined with their findings, we recognize the significant molecular heterogeneity among morphologically similar intracranial mesenchymal tumors. Adjuvant radiotherapy may improve local control for high‑risk fusion‑positive cases, although standardized treatment regimens remain unclear.¹⁰

A recent cohort study reported EWSR1::ATF1 fusion as a potential independent adverse prognostic factor with a high hazard ratio.⁵ Nevertheless, as a single case report, our study cannot establish definitive clinical guidelines. We only suggest intensified long‑term surveillance for EWSR1::ATF1‑positive patients, even if they have low‑risk clinical features. Further multi‑center studies with epigenetic data are needed to optimize postoperative follow‑up and therapeutic strategies.³

Our study has several limitations. First, methylation profiling was not performed, limiting accurate molecular subtyping.⁴ Second, only one case was analyzed, lacking statistical power. Third, the follow‑up duration after the second resection remains relatively short. Fourth, no adjuvant therapy was applied after initial surgery; therefore, the effect of postoperative treatment on recurrence prevention cannot be evaluated.¹⁰

The FET family genes encode RNA-binding proteins. Gene fusion events involving these genes can generate aberrant fusion proteins, which may participate in the initiation and progression of tumors. FET-CREB fusion-positive intracranial mesenchymal tumors exhibits a broad morphological spectrum, including cytological features such as round, oval, and spindle-shaped cells with eccentric nuclei resembling rhabdomyoblasts. Across extracranial FET-CREB tumors, especially those with EWSR1-CREB, over a dozen cases have been reported at various body sites.^4,11–17 Their correlation with tumor phenotypes has not been fully elucidated across tumor entities, including angiomatoid fibrous histiocytoma (AFH), clear cell sarcoma (CCS), intracranial myxoid mesenchymal tumor (IMMT), primary pulmonary myxoid sarcoma (PPMS), gastrointestinal neuroectodermal tumor (GNET), and salivary gland clear cell carcinoma. However, there are differences in clinical presentation (location, age of diagnosis, prognosis), histopathology, and molecular aspects (frequency of gene fusion partners).^4,13

AFH is characterized primarily by an EWSR1-CREB1 gene fusion (83% of cases in five series), occasionally with an ATF1 gene fusion (17% of cases in five series).^6,13,18 AFH is a slowly progressive soft tissue tumor common in children and young adults, most commonly with EWSR1-CREB1 and occasionally with EWSR1-ATF1 fusion. Histopathologically, it features an incomplete thick fibrous pseudocapsule with hemosiderin deposits, and the tumor is composed of oval to short spindle-shaped histiocytoid cells with irregular cystic expansions and pseudovascular spaces; thin-walled small vessels are also present.^6,9

CCS is a malignant tumor found in soft tissues, characterized histopathologically by large nests of clear cells and expression of melanocytic markers (HMB45 and MelanA); over 90% of CCS tumors harbor an EWSR1-ATF1 fusion. Intracranial mesenchymal tumors with FET-CREB fusion are negative for melanocytic markers and can be distinguished from CCS.^12,13

A minority of tumors with round-cell morphology and collagenous stroma may be misdiagnosed as desmoplastic small round cell tumor (DSRCT),¹⁹ a malignant and aggressive childhood tumor characterized by a t(11;22)(p13;q12) chromosomal translocation that generates the EWSR1-WT1 fusion oncogene.²⁰ Morphology is sometimes difficult to distinguish from intracranial mesenchymal tumors fused with FET-CREB, especially when the tumor also expresses immunohistochemical markers such as desmin, EMA, and synaptophysin. DSRCT is rare in the intracranial region and mostly occurs in the retroperitoneum, pelvic cavity, nasal cavity, paranasal sinuses, and other areas.^18–20

Meningiomas and solitary fibrous tumors share similarities in location and radiological features with intracranial mesenchymal tumors with FET-CREB fusion but can be distinguished histopathologically. Meningiomas may have myxoid stromal components,⁷ but lack the characteristic fibrous pseudocapsule and lymphoplasmacytic cuff surrounding the tumor, as well as the diffuse desmin expression seen in FET-CREB fusion tumors. Additionally, meningiomas express the diagnostic biomarker SSTR2, which is absent in FET-CREB fusion tumors.^7,21 Compared to solitary fibrous tumors, FET-CREB fusion intracranial mesenchymal tumors lack nuclear STAT6 immunopositivity and the characteristic NAB2-STAT6 fusion.^13,22,23

This study is a retrospective single-case analysis. The recurrent case presented herein is insufficient to confirm that EWSR1-ATF1 fusion serves as an independent high-risk prognostic biomarker, and the prognostic predictive value of this molecular alteration remains to be validated in large-scale, multicenter cohort studies. As a rare supratentorial recurrent tumor, this case exhibits a unique clinical phenotype and provides individualized clinical data to enrich the disease spectrum of this rare tumor entity.

Conclusion

Intracranial mesenchymal tumors harboring FET-CREB gene fusions are exceedingly rare in clinical practice. They present a complex clinicopathological spectrum and substantial heterogeneity, which renders them highly susceptible to misdiagnosis with a variety of central nervous system and soft tissue neoplasms. Accurate diagnosis of this tumor requires integrated evaluation of radiological, histopathological, and molecular biological findings, since morphological observation or immunohistochemical detection alone cannot establish a definitive diagnosis. We report a rare delayed recurrence case in an adult patient with a completely resected supratentorial EWSR1::ATF1‑fused tumor. Our observation indicates that EWSR1::ATF1 fusion confers inherent recurrence risk independent of conventional favorable clinical factors.⁵

This recurrent case further enriches the clinical follow-up data of intracranial mesenchymal tumors with FET-CREB fusions and highlights the clinical value of integrating molecular subtyping with phenotypic features for individualized diagnosis and treatment. Further accumulation of large-sample clinical data is warranted to clarify prognostic discrepancies among distinct gene fusion subtypes and pathological phenotypes, and to optimize the diagnostic, therapeutic, and follow-up criteria for this rare tumor entity.

Funding Statement

No funding was received for this study.

Ethical Approval and Patient Consent

This study was a retrospective single‑case report without invasive experimental procedures. Formal ethical approval was exempted according to local institutional regulations. The study complied with the principles of the Declaration of Helsinki. Written informed consent was obtained from the patient for publication of clinical data, pathological images, and radiological images. All personal information was fully de‑identified to ensure patient privacy.

Disclosure

The authors declare no competing interests.