Non‐Meningothelial Mesenchymal Tumor of the Central Nervous System: Case Report and Literature Review

ABSTRACT

Non‐meningothelial mesenchymal tumors (NMMTs) of the central nervous system (CNS) are rare and diagnostically challenging neoplasms. Although World Health Organization (WHO) updates have incorporated molecularly defined entities, a subset of tumors remains difficult to classify because of overlapping morphology, nonspecific immunophenotype, and absence of canonical gene fusions. We report a supratentorial NMMT tumor in a 40‐year‐old man presenting with progressive left‐sided paresthesia. Neuroimaging demonstrated a parasagittal mass. Following gross total resection, histological examination revealed a predominantly myxoid proliferation with focal cortical and dural infiltration. Immunohistochemistry was noncontributory, showing focal CD99 expression. Fluorescence in situ hybridization excluded EWSR1, FUS, and CIC rearrangements. Extended RNA‐based next‐generation sequencing identified a putative low‐confidence SHISA5::BRAF fusion. This case illustrates the persistent diagnostic complexity of NMMTs despite integrated histological and molecular evaluation. The absence of canonical alterations not only complicates classification but also contributes to uncertainty regarding prognosis and optimal therapeutic strategies. Emerging evidence suggests that a subset of these tumors may be driven by alternative oncogenic mechanisms, including MAPK pathway activation beyond currently recognized fusion‐defined entities. In this context, the identification of a putative SHISA5::BRAF fusion expands the molecular spectrum of these neoplasms, although its functional relevance remains uncertain. These challenges underscore the need for cautious long‐term follow‐up and individualized multidisciplinary management.

1. Introduction

NMMTs of the CNS represent a heterogeneous and extremely rare group of neoplasms, with an estimated annual incidence of approximately 0.3–0.4 cases per 100 000 population. In large population‐based registries, mesenchymal tumors, for which non‐meningothelial subtypes are a subset, account for only around 0.3%–0.4% of all primary CNS neoplasms [1, 2].

Historically, the classification of these tumors has been complex, and it has evolved significantly. The current classification includes a specific group for NMMTs [3], where new histomolecular entities have been recognized, such as intracranial mesenchymal tumors with FET::CREB fusion, CIC‐rearranged sarcoma, and primary intracranial sarcoma with DICER1 mutation [4]. Nonetheless, the diagnosis of these tumors remains a considerable challenge due to their frequently nonspecific histology, variable immunophenotype, and morphological overlap with other soft tissue tumors, making molecular characterization essential for accurate diagnosis [2, 3, 4].

Here, we present a case of a supratentorial NMMT negative for FET::CREB fusions, exemplifying the clinicopathological, molecular, and therapeutic challenges of this group of neoplasms.

2. Clinical Summary

2.1. Clinical Features

A 40‐year‐old man with no relevant medical history presented with several months of progressive paresthesia involving the left lower limb, later extending to the left upper limb. Neurological examination was otherwise normal.

2.2. Neuroimaging

MRI demonstrated a large right parasagittal parietal mass measuring approximately 6 × 4 × 5 cm (anteroposterior × transverse × craniocaudal), with an indeterminate intra‐ versus extra‐axial origin (Figure 1). The lesion was closely apposed to the dura without marginal dural thickening, and an apparent peripheral hypervascular pedicle was noted. Its relationship with adjacent meningeal and cortical vessels was unusual, with the tumor extending both superficially and deeply to these structures and partially encasing them, accompanied by effacement of the normal cortical ribbon at the tumor–brain interface. No bony involvement was identified.

FIGURE 1.

Preoperative MRI shows a right parasagittal parietal lesion with mixed solid and cystic/necrotic architecture. (A, B) T2‐weighted images demonstrate a multilobulated cortico–subcortical mass with heterogeneous signal; (B) highlights multiple flow voids and vascular encasement at the meningocortical interface, indicating marked intratumoral vascularity. Margins are relatively well defined against the deep white matter, favoring a cortex–meningeal interface rather than diffuse white‐matter infiltration. (C) The ADC map shows no convincing low‐ADC component. (D, E) Post‐contrast T1‐weighted images reveal heterogeneous patchy enhancement with interspersed non‐enhancing areas; cranial dural attachment is best seen in (E). (F) SWI shows intralesional susceptibility foci consistent with blood products and prominent vessels. Thick arrows mark solid tumor components in (A), (C), and (D); asterisks in (A) indicate cystic/necrotic regions (corresponding to high diffusivity in (C) and non‐enhancing areas in (D)); thin arrows in (B) and (F) denote vascular structures; and a dashed circle in (E) marks dural attachment.

Post‐contrast imaging demonstrated heterogeneous patchy enhancement, with enhancing solid areas interspersed with non‐enhancing cystic or necrotic components and regions of intermediate T2 signal (Figure 1A,B). Diffusion‐weighted imaging showed no restriction (Figure 1C). Multiple intralesional vessels were conspicuous (Figure 1F), and no diffuse infiltrative‐appearing component was present within the adjacent white matter (Figure 1D,E).

2.3. Surgical Findings

After completion of the diagnostic workup, a right parasagittal craniotomy with intraoperative motor mapping was performed. Intraoperatively, the lesion was found to be intra‐axial, without dural attachment, and located beneath a bridging vein draining into the superior sagittal sinus. The tumor appeared well circumscribed, with sharp borders, a slightly yellowish coloration, and a marked tendency to bleed, consistent with a highly vascular lesion (Figure 2).

FIGURE 2.

Intraoperative image of the right parasagittal parietal lesion demonstrating a well‐circumscribed intra‐axial mass beneath a bridging vein and without dural attachment. 1, motor cortex. T, tumor.

Frozen‐section (Figure 3A) analysis suggested a low‐grade neoplasm. In view of this preliminary diagnosis, motor mapping was continued up to 1 mA, aiming to achieve an optimal onco‐functional balance. Gross total resection was achieved, which was subsequently confirmed by postoperative magnetic resonance imaging.

FIGURE 3.

Histopathological features of the tumor on hematoxylin and eosin (HE) staining (A–F). (A) Frozen section shows a low‐grade neoplasm. (B) An extra‐axial mesenchymal proliferation with preserved underlying cortical parenchyma is shown. (C) Moderately cellular spindle cell proliferation is embedded in a myxoid stroma, with scattered amianthoid‐like bodies. (D) Tumor infiltration of the adjacent brain parenchyma through perivascular (Virchow–Robin) spaces. (E) Areas of the tumor with a branching (“staghorn‐like”) vascular pattern, reminiscent of a solitary fibrous tumor. (F) Amianthoid bodies. Scale bars: (A) 200 μm, (B) 100 μm, (C) 200 μm, (D) 400 μm, (E, F) 200 μm.

Postoperatively, the patient showed a favorable clinical course, with no neurological deficits except for minimal intermittent paresthesia and mild apraxia in the left lower limb.

3. Pathological Findings

The pathological analysis determined that the intracranial lesion corresponded to a predominantly leptomeningeal myxoid mesenchymal neoplasm with superficial cortical and dural infiltration (Figure 3B).

The lesion consisted of a predominantly leptomeningeal spindle cell proliferation embedded in an abundant fibromyxoid stroma. The tumor lacked a distinctive architectural pattern overall, although focal areas exhibited a prominent branching (“staghorn‐like”) vascular network reminiscent of solitary fibrous tumor and scattered amianthoid collagen bundles (Figure 3C,E,F). The neoplasm infiltrated the dura and extended into the adjacent cortex through Virchow–Robin spaces, with focal cortical involvement (Figure 3D).

The tumor was composed of relatively uniform spindle cells with scant eosinophilic fibrillary cytoplasm and elongated nuclei displaying finely dispersed chromatin and inconspicuous nucleoli. In addition, minimal cytological atypia was found, mitotic figures were not readily identified, and no necrosis was observed. The growth pattern was heterogeneous, alternating densely cellular fasciculated areas within a fibromyxoid background with poorly defined hypocellular nodules embedded in a fibrous stroma containing amianthoid collagen bundles.

Immunohistochemical analysis demonstrated diffuse vimentin expression with patchy CD99 positivity, cyclin D1 expression, cytoplasmic and paranuclear dot‐like WT1 staining, very focal CD34 positivity, variable smooth muscle actin expression, weak SATB2 staining, and only isolated calretinin‐positive cells. Nuclear expression of H3K27me3, INI1, and BRG1 was retained. The tumor cells were negative for GFAP, OLIG2, SOX10, S100, synaptophysin, EMA, AE1/AE3, CAM5.2, desmin, MyoD1, myogenin, STAT6, MUC4, BCOR, NKX2.2, SSTR2A, ERG, TLE1, SSX1, SS18‐SSX, pan‐TRK, HMB45, and Melan‐A. The Ki‐67 labeling index was approximately 20%. Overall, the immunophenotype excluded glial, neuronal, meningothelial, epithelial, melanocytic, and myogenic differentiation, while also arguing against several fusion‐associated mesenchymal neoplasms but remaining otherwise nonspecific, prompting further molecular characterization.

Break‐apart fluorescence in situ hybridization (FISH) was performed on formalin‐fixed, paraffin‐embedded (FFPE) tumor tissue using ZytoLight SPEC dual‐color break‐apart probes (ZytoVision, Germany) targeting the EWSR1 (22q12.2), FUS (16p11.2), and CIC (19q13.2) loci. No rearrangements involving EWSR1, FUS, or CIC were identified. Comprehensive molecular profiling was undertaken using targeted next‐generation sequencing. DNA analysis was performed with the Oncomine Precision Assay (Thermo Fisher Scientific) on the Genexus GX5 platform, which interrogates the following genes: AKT1, AKT2, AKT3, ALK, AR, ARAF, BRAF, CDK4, CDKN2A, CHEK2, CTNNB1, EGFR, ERBB2, ERBB3, ERBB4, ESR1, FGFR1, FGFR2, FGFR3, FGFR4, FLT3, GNA11, GNAQ, GNAS, HRAS, IDH1, IDH2, KIT, KRAS, MAP2K1, MAP2K2, MET, MTOR, NRAS, NTRK1, NTRK2, NTRK3, PDGFRA, PIK3CA, PTEN, RAF1, RET, ROS1, SMO, and TP53, and revealed no clinically significant sequence variants.

RNA‐based fusion analysis was carried out using the FusionPlex Sarcoma Panel V.2 (ArcherDX), which interrogates 63 sarcoma‐associated genes for fusion events. This analysis identified a putative low‐confidence SHISA5::BRAF fusion involving SHISA5 Exon 5 and BRAF Exon 10.

Thus, the final diagnosis was a myxoid NMMT negative for FET::CREB fusions, with a suspected SHISA5::BRAF rearrangement.

4. Discussion

Despite the recognition of new NMMT entities in the 2021 WHO classification, a significant number of cases remain that cannot be precisely classified even after extensive studies. This conceptual heterogeneity directly translates into clinical practice, where definitive diagnosis continues to rely on careful integration of morphology, immunohistochemistry, and molecular biology [3, 4].

As summarized in Table 1, most reported intracranial NMMTs harbor recurrent FET::CREB‐family fusions, most commonly involving EWSR1::CREM, EWSR1::CREB1, or EWSR1::ATF1, establishing this alteration as the dominant molecular signature of this entity. However, accumulating evidence indicates that a subset of intracranial mesenchymal tumors lacks these canonical alterations and remains difficult to classify despite comprehensive histomolecular evaluation. Recent reports have described tumors without detectable gene fusions, as well as cases harboring noncanonical or isolated molecular alterations that do not fit within current WHO‐defined entities.

TABLE 1.

Clinicopathological and molecular features of previously reported intracranial myxoid mesenchymal tumors.

Study	No. of cases	Age/sex	Location	Histopathological features	Molecular alteration	Treatment	Outcome/follow‐up
White et al. [5]	1	9 years/M	Supratentorial extra‐axial	Myxoid mesenchymal tumor	EWSR1::CREM	GTR	Recurrence at 6 months
Komatsu et al. [6]	1	53 years/F	Supratentorial intraventricular	Myxoid, reticular growth	EWSR1::CREB1	GTR	N/A
Ballester et al. [7]	1	67 years/M	Supratentorial extra‐axial	Myxoid mesenchymal tumor	EWSR1::ATF1	SR	Stable at 3.5 months
Liu et al. [8]	1	31 years/M	Supratentorial extra‐axial	Myxoid mesenchymal tumor	EWSR1::CREM	GTR	Stable at 14 months
Ward et al. [9]	1	48 years/F	Supratentorial intraventricular	Myxoid mesenchymal tumor	EWSR1::ATF1	GTR + radiotherapy	Stable at 12 months
Kambe et al. [10]	1	65 years/F	Infratentorial	Myxoid mesenchymal tumor	EWSR1::CREM	GTR	Stable at 4 months
Sloan et al. [11]	20	9–70 years/75% F, 25% M	Predominantly extra‐axial	Myxoid/spindle cell spectrum	FET::CREB fusions (EWSR1/FUS)	GTR or SR ± adjuvant therapy	Recurrence rate 70%, between 1 and 57 months
Tauziède‐Espariat et al. [12]	1	41 years/M	Supratentorial extra‐axial	Myxoid mesenchymal tumor	SMARCA2–CREM fusion	N/A	N/A
Sasaki et al. [13]	1	7 years/F	Supratentorial extra‐axial	Myxoid mesenchymal tumor	EWSR1::CREM + MAP3K13 mutation	GTR	Stable at 9 months
Shaikh et al. [14]	1	27 years/F	Supratentorial intra‐axial	Non‐myxoid variant	EWSR1::ATF1	GTR	N/A
d'Amati et al. [15]	1	43 years/M	Supratentorial extra‐axial	Mesenchymal tumor, atypical features	COX14::PTEN rearrangement	GTR	Recurrence at 2 months
D'Antonio et al. [16]	1	27 years/F	Infratentorial	Myxoid mesenchymal tumor	FET::CREB	GTR + radiotherapy + immunotherapy	Complete remission
Ozkizilkaya et al. [17]	1	37 years/F	Supratentorial intraventricular	Myxoid mesenchymal tumor	FET::CREB	GTR	Recurrence at 36 months
Shen et al. [18]	1	73 years/F	Supratentorial intra‐axial	Myxoid mesenchymal tumor	No alteration identified	GTR	Recurrence at 7 days
Mezzacappa et al. [19]	5	13–32 years/60% F, 40% M	Supratentorial	Myxoid mesenchymal tumors	CREB‐family fusions	GTR or biopsy ± adjuvant therapy	Recurrence rate 33%, between 1 and 24 months
Imam et al. [20]	1	44 years/M	Supratentorial	Myxoid mesenchymal tumor	FET::CREB	GTR	Stable at 12 months
Sono et al. [21]	1	45 years/M	Supratentorial extra‐axial	Myxoid tumor with rhabdoid features	FET::CREM	GTR	Stable at 36 months
Hu et al. [22]	6	20‐year median age/33% F, 66% M	Supra and infratentorial	Myxoid mesenchymal tumors	EWSR1::ATF1, EWSR1::CREB1, EWSR1::CREM	GTR and subtotal resection	N/A
Carlisle et al. [23]	1	51 years/F	Supratentorial intra‐axial	Myxoid mesenchymal tumor	No alterations identified	GTR	Stable at 84 months

Abbreviations: GTR, gross total resection; N/A, not available; SR, subtotal resection.

The persistence of unclassified cases underscores the need for continued refinement of the current classification system through integrated molecular profiling and long‐term clinical correlation to better define biologically coherent subgroups.

Neuroimaging plays a critical role in the initial characterization and surgical planning of our case. The dural‐based mass with cortical involvement lacked classic extra‐axial signs. Although meningioma was considered, marked vascularity and heterogeneous architecture broadened the differential to include solitary fibrous tumor, hypervascular metastasis, and high‐grade glioma; absence of diffuse white‐matter infiltration favored a primarily meningeal process.

Given the possibility of an intra‐axial lesion adjacent to the motor cortex and corticospinal tract, resection was performed under continuous neurophysiological monitoring with direct cortical and subcortical motor mapping. Considering the patient's young age, preserved neurological status, and frozen‐section findings suggestive of a low‐grade tumor, resection was maximized while maintaining functional safety. This deliberate onco‐functional balance aimed to achieve optimal tumor removal while preserving motor pathways.

From a histopathological standpoint, the diagnostic challenge persisted. These tumors may exhibit collagenous or myxoid stroma, with spindle, epithelioid, or reticular architecture, often mimicking meningiomas or soft tissue sarcomas [4, 11]. In our case, alternating fasciculated areas within a fibromyxoid background and regions of fibrous stroma with amianthoid‐like fibers further illustrated this morphological overlap. As noted by Sloan et al., even tumors within the same molecular group may show divergent morphology and variable immunophenotypes, reinforcing the need for molecular correlation in accurate classification [11].

Although the lesion showed focal cortical infiltration, several features argued strongly against a circumscribed glial neoplasm. First, its predominantly extra‐axial growth pattern, with dural attachment, expansion of the subarachnoid space, and extension into the underlying cortex through Virchow–Robin spaces, favored a meningeal‐based mesenchymal process rather than a primary intra‐axial glial tumor. Second, the tumor exhibited a fibromyxoid rather than fibrillary stroma, lacking the characteristic architectural features of circumscribed gliomas. Finally, the immunophenotype did not support glial or glioneuronal differentiation, as tumor cells were negative for GFAP, OLIG2, SOX10, and synaptophysin. Taken together, the anatomical distribution, histomorphological characteristics, and immunoprofile effectively excluded a circumscribed glioma from the differential diagnosis.

On the other hand, the immunohistochemical profile supported the morphological impression of an undifferentiated mesenchymal neoplasm. Diffuse vimentin expression was consistent with mesenchymal differentiation, whereas the absence of glial (GFAP, OLIG2), neuronal (synaptophysin), meningothelial (EMA, SSTR2A), epithelial (AE1/AE3, CAM5.2), melanocytic (HMB45, Melan‐A, SOX10), and myogenic (desmin, MyoD1, myogenin) markers argued against meningioma, epithelial and melanocytic neoplasms, and tumors showing myogenic differentiation. Although amianthoid collagen bundles, a feature that may also be encountered in solitary fibrous tumor, were identified, this diagnosis was excluded by the absence of nuclear STAT6 expression. Furthermore, the lack of expression of MUC4, BCOR, NKX2.2, SS18‐SSX, and pan‐TRK made several morphologically similar fusion‐associated mesenchymal neoplasms, including BCOR‐altered sarcomas, synovial sarcoma, and NTRK‐rearranged spindle cell neoplasms, unlikely. Patchy CD99 positivity, cyclin D1 expression, cytoplasmic and paranuclear dot‐like WT1 staining, focal CD34 positivity, weak SATB2 staining, and variable smooth muscle actin expression has been reported across a broad spectrum of soft tissue tumors.

Collectively, the morphologic and immunohistochemical findings supported the interpretation of an undifferentiated mesenchymal neoplasm by excluding the principal histologic mimics while failing to identify an immunophenotypic profile characteristic of any recognized entity. Consequently, further molecular characterization was required to refine the diagnosis.

Molecular studies have delineated more reproducible subgroups within CNS mesenchymal neoplasms. Sloan et al. identified 20 intracranial tumors with FET–CREB fusions, establishing a distinct entity characterized by predominantly extra‐axial location and morphological variability [11], a pattern confirmed in subsequent series [16]. However, a subset of lesions lacks these defining fusions and remains unclassified [24]. Our case belongs to this latter group, characterized by the absence of canonical fusions and a noncontributory immunophenotype.

Within this molecular landscape, the isolated identification of a BRAF alteration (SHISA5::BRAF fusion) represents an exceptional finding. Although BRAF alterations have been increasingly recognized in a variety of CNS and mesenchymal tumors, most reported cases involve canonical mutations (e.g., V600E) or recurrent fusion partners. Rare BRAF fusions have been described in intracranial neoplasms, including ARHGAP45::BRAF in histiocytic sarcoma and PTPRN2::BRAF in meningioma [25, 26]. However, fusion partners are highly heterogeneous, and many represent isolated case reports.

To the best of our knowledge, no prior cases of SHISA5::BRAF fusion have been reported in the literature, either in CNS tumors or in mesenchymal neoplasms. This finding further expands the molecular spectrum of BRAF‐altered tumors and raises the possibility that additional, as yet uncharacterized fusion partners may underlie a subset of currently unclassified neoplasms [27].

BRAF alterations result in constitutive activation of the MAPK pathway, analogous to other oncogenic BRAF fusions such as KIAA1549::BRAF or TOM1L2::BRAF described across CNS tumor types [27]. Emerging evidence suggests that a subset of NMMTs may be driven by alternative oncogenic mechanisms, including MAPK pathway activation outside currently recognized fusion‐defined entities. However, the functional relevance of SHISA5::BRAF as a fusion partner remains unknown, and this rearrangement may represent a passenger event rather than a true oncogenic driver. This molecular heterogeneity, already highlighted in recent reviews [3, 4], reinforces the concept that biologically distinct tumors remain unclassified despite extensive histomolecular characterization.

Clinically, this biological diversity is mirrored by variable behavior. Although many cases follow an indolent course after complete resection, recurrence and progression occur in approximately one‐third of patients, with a small percentage of mortality [4, 11]. Accordingly, prolonged follow‐up is warranted even after gross total resection.

From a therapeutic standpoint, maximal safe surgical resection remains the cornerstone of management. The role of adjuvant radiotherapy or chemotherapy is not well‐defined and is generally reserved for subtotal resection, recurrence, or aggressive histological features, with evidence largely derived from small series [15, 24, 28]. While BRAF inhibitors, alone or combined with MEK inhibitors, have demonstrated efficacy in tumors harboring canonical activating mutations such as BRAF V600E or V600K, their activity in BRAF fusion‐driven neoplasms is less predictable [4, 28, 29]. Given the uncertain functional relevance of SHISA5::BRAF, the therapeutic implications of this finding remain speculative, reinforcing the need for individualized, multidisciplinary management [30].

More broadly, NMMTs of the CNS remain diagnostically challenging due to their morphological overlap and inconsistent immunophenotypes. This case highlights the limitations of conventional approaches and underscores the importance of comprehensive molecular testing, particularly in tumors that remain unclassified after standard evaluation. The identification of a rare BRAF fusion may expand the molecular spectrum of these neoplasms and supports the growing recognition that a subset of intracranial mesenchymal tumors falls outside current classification frameworks. In this context, careful long‐term follow‐up and continued refinement of molecular classification systems are warranted.

Funding

The authors have nothing to report.

Ethics Statement

The authors have nothing to report.

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data supporting the findings of this study are available from the corresponding author upon reasonable request, subject to applicable privacy and confidentiality restrictions.