Introduction

Solitary fibrous tumors (SFTs) are rare mesenchymal tumors that can occur at any location [1,2,3,4]. In the 2021 World Health Organization (WHO) classification of tumors of the central nervous system (CNS), SFTs are graded from 1 to 3 according to mitotic activity and tumoural necrosis [5]. SFTs with < 2.5 mitoses/mm2 (< 5 mitoses/10 high-power field [HPF]) and no necrosis were considered grade 1; SFTs with ≥ 2.5 mitoses/mm2 (≥ 5 mitoses/10 HPFs) and no necrosis were considered grade 2; tumors with ≥ 2.5 mitoses/mm2 (≥ 5 mitoses/10 HPFs) and necrosis were considered grade 3. Most intracranial SFTs are dura-based; thus, the skull base, parasagittal, and falcine regions are common sites, whereas the cerebellopontine angle (CPA), pineal gland, and sellar region are uncommon sites [5].

SFT and hemangiopericytoma (HPC) were initially thought to be different tumors; however, the identification of NAB2-STAT6 in both SFT and HPC suggests that they comprise the same tumor entity [3, 6, 7]. The presence of the fusion gene results in nuclear expression of STAT6 by immunohistochemistry [6]. NAB2-STAT6 fusion variants have been classified into two major groups based on their hypothesized functional effects: NAB2 exon 4-STAT6 exon 2/3 and NAB2 exon 5/6/7-STAT6 exon 16/17/18 [1, 5, 8]. Several studies have suggested that the fusion gene variants are associated with clinicopathological features [1, 2, 5, 7, 9]. NAB2 exon 4-STAT6 exon 2/3 fusion variants have been reported to be associated with pleuro-pulmonary development, lower mitotic counts, and a lower risk of adverse events [1, 2, 5, 7, 9]. On the other hand, NAB2 exon 5/6/7-STAT6 exon 16/17/18 fusion variants have been reported to be associated with development from the meninges/soft tissue/head and neck, higher cellularity, and recurrent tumors [1, 3, 5, 7]. In addition to the fusion gene variants, TP53 and TERT promoter mutations have been found to offer clinicopathological value in SFT; aberrant p53 overexpression and/or TP53 mutations have been demonstrated in dedifferentiated areas of SFT, and TERT promoter mutation has been shown to be associated with worse prognosis [4, 7, 9,10,11,12,13].

Salivary gland tissue outside the major and minor salivary glands is considered as ectopic salivary glands (ESGs), with frequent sites being the mandible, ear, and mylohyoid muscle [14]. Intracranial ESGs are rare and mainly found in the sellar region. Only 3 cases of salivary gland tissue presented in the CPA have been reported [15,16,17]. Neoplasms involving ESGs are uncommon; outside the CNS, 35 cases have been reported, and they all were salivary gland-related tumors (Table S1) [18,19,20,21,22,23,24,25,26,27,28,29,30]. As intracranial tumors involving ESGs, in addition to 2 salivary gland–related tumors, 2 SFTs in the CPA have been reported; however, SFT with ESGs has not been reported outside the CNS (Table S1) [15,16,17, 31]. Genetic analysis was not performed for the 2 reported SFT with ESGs.

In this study, we investigated a series of 3 cases of meningeal SFT with ESGs, including 1 previously reported case, for clinicopathological and molecular features.

Materials and methods

Tumor samples

Three cases were collected for this study. Case 1 was previously reported [17]. Case 2 was from the consultation files of 1 of the authors (J.H). Case 3 was from the pathology archives of the Japan Brain Tumor Reference Center. Sections for histologic and genetic analyses were prepared from formalin-fixed paraffin-embedded (FFPE) tissue specimens. This study was conducted in accordance with the ethical committees of Gunma University (HS2016-075).

Conventional histological analysis

Three micrometer thick tissue sections were cut and stained with hematoxylin and eosin. Immunohistochemical staining was performed on FFPE tissue sections. Primary antibodies directed against the following antigens were applied: CD34 (NU-4A1; 1:100; Nichirei, Tokyo, Japan), STAT6 (S-20; 1:300; Santa Cruz, Dallas, TX, USA), p53 (DO-7; 1:50; Leica Microsystems, Wetzlar, Germany), and Ki-67 (MIB-1; 1:100; Dako, Glostrup, Denmark). For coloration, a commercially available biotin-streptavidin immunoperoxidase kit (Histofine, Nichirei, Tokyo, Japan) and diaminobenzidine were employed.

Detection of NAB2-STAT6 fusion transcripts

Reverse transcription polymerase chain reaction (RT-PCR) was performed, as previously described [32]. Briefly, total RNA was extracted from FFPE samples using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and was reverse-transcribed using Superscript III reverse transcriptase (Invitrogen) to prepare the first-strand complementary DNA. NAB2-STAT6 fusion assays were performed using primers described previously [32]. Each PCR product (5 μL) was loaded onto 2% agarose gel with ethidium bromide and visualized under UV illumination. The PCR products were also evaluated by direct sequence analysis using the Big Dye Terminator method (version 1.1; Applied Biosystems, Foster City, CA, USA) to confirm the breakpoints of the fusion transcripts.

DNA extraction and direct DNA sequencing for TP53 and TERT promoter mutations

Genomic DNA was extracted from the FFPE tissue sections, as previously described [33], and was amplified and sequenced using previously described primer sets [34, 35]. PCR products were analyzed by agarose gel electrophoresis, and were then sequenced on a 3130xl Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) with the Big Dye Terminator v.1.1 Cycle Sequencing Kit (Applied Biosystems) following standard procedures.

Statistical analysis

Extensive review of the English-language literature was conducted to identify previously reported cases of meningeal SFTs without ESG components with known NAB2-STAT6 fusion variants. Fisher’s exact test was performed to analyze the correlation between the fusion variants and presence of ESGs.

Results

Clinical findings

Relevant clinical data are summarized in Table 1. All 3 cases were adult females (27 to 41 years old) and presented with an extra-axial mass in the CPA. Tumors were cystic in cases 1 and 2, which showed high-signal intensity on T2-weighted magnetic resonance images (Fig. 1a) with capsule and septa showing contrast enhancement (Fig. 1b). In case 3, the recurrent tumor was solid, which showed iso- to high-signal intensity on T2-weighted images (Fig. 1c) and homogeneous enhancement on gadolinium-enhanced T1-weighted images (Fig. 1d).

Table 1 Case list with clinical features
Full size table
Fig. 1
figure 1

Magnetic resonance images of case 2 at initial onset (a, b) and case 3 at first recurrence (c, d). a, b A lobulated mass in the right cerebellopontine angle compresses right cerebellum and middle cerebellar peduncle. The mass shows very high signal intensity on T2-weighted images (a), suggesting a multilocular cystic mass, and the capsule and septa show contrast enhancement on gadolinium-enhanced T1-weighted images (b). c, d A lobulated mass in the left cerebellopontine angle compresses middle cerebellar peduncle. The mass shows mildly high signal intensity on T2-weighted images (c) and shows homogeneous enhancement on gadolinium-enhanced T1-weighted images (d)

Full size image

All cases underwent gross total tumor resection. In case 2, the tumor recurred locally 8 (first recurrence, R1), 18 (R2), and 19 (R3) years after the initial surgery, and excisions were performed. In case 3, although the tumor had been initially controlled by the first surgery and a gamma knife treatment at 7 years, its recurrence was more aggressive, and the patient received additional excisions at 13 (R1), 14 (R2), and 16 (R3) years after initial surgery. Multiple dissemination throughout posterior fossa and cervical spinal cord had developed at 18 years, and 2 more surgeries (R4 and R5) and 4 stereotactic radiosurgeries were performed in the next 4 years. The patient died 22 years after the initial surgery at her age of 57 years.

Histopathological findings

Representative histological images are shown in Fig. 2, and the histological features are summarized in Table 2. Specimens from R3 and R5 of case 3 were not available.

Fig. 2
figure 2

Microscopic appearance. a, b, c The lesion of case 1 demonstrates a mixture of tumor tissue and ectopic salivary gland (ESG) components. The tumor is composed of spindle to ovoid cells with eosinophilic cytoplasm separated by hyalinized collagenous stroma. d. R2 of case 2 shows a proliferation of small round or short spindle cells with high cellularity. ESGs are mixed in the tumor. e, f R3 of case 2 is composed of atypical mesenchymal cells with pleomorphic nuclei (e). ESGs remained in the tumor (f). g P of case 3 is composed of short spindle to oval shaped cells with high cellularity. ESGs are observed in the tumor. h, i R4 of case 3 demonstrates brisk mitoses (h) and necrosis (i). P, primary; R2-R4, second-fourth recurrence. Original magnification: a, i × 40; b, d, e, f, g × 200; c, h × 400

Full size image
Table 2 Histological, immunohistochemical, and molecular features
Full size table

The lesion of case 1 demonstrated a mixture of tumor tissue and ESG components (Fig. 2a, b). The tumor was composed of spindle to ovoid cells with eosinophilic cytoplasm separated by hyalinized collagenous stroma and had variable cellularity (Fig. 2c). Necrosis and mitosis were absent, and this tumor was considered to be SFT, WHO grade 1. In addition, components of glandular structures were present in the tumor (Fig. 2a, b). The components consisted of serous and mucous acini and dilated ducts surrounded by myoepithelial cells. No cellular atypia was observed in these glandular ducts. These characteristic structures were consistent with ESGs.

The primary tumor (P) and R1–2 of case 2 showed a mixture of tumor tissue and ESG components (Fig. 2d). These tumors were focally composed of small round or short spindle cells with high cellularity with no necrosis or mitosis, and were considered to be SFT, WHO grade 1. ESGs were mixed in these tumors. Meanwhile, R3 of case 2 was composed of atypical mesenchymal cells with pleomorphic nuclei (Fig. 2e). Necrosis was absent. The mitotic rate was up to 6 per 10 high-power fields (HPFs), corresponding to SFT, WHO grade 2, and atypical mitoses were observed. These histological features were judged as dedifferentiation. ESGs remained in the tumor (Fig. 2f).

P, R1–2, and R4 of case 3 were composed of short spindle to oval shaped cells with high cellularity (Fig. 2g, h). In P and R1–2, mitotic counts ranged from 4 to 9 per 10 HPFs and necrosis was absent. These tumors were considered to be SFT, WHO grade 1 or 2. Meanwhile, brisk mitoses (21 per 10 HPFs) and necrosis were observed in R4 (Fig. 2h, i). This tumor was classified as SFT, WHO grade 3. ESGs were observed only in P (Fig. 2g).

The immunohistochemical findings are summarized in Table 2. A diffuse and strong expression of CD34 was demonstrated in case 1 and P and R1–2 of case 3. In R1 of case 2, a focal and strong expression of CD34 was demonstrated. Nuclear STAT6 positivity was observed in the spindle tumor cells but not in the ESGs in all specimens analyzed (Fig. 3a–c). The pleomorphic tumor cells in R3 of case 2 were also positive for STAT6. A diffuse and moderate expression of p53 was de monstrated only in R2 of case 2 (Fig. 3e), whereas the expression of p53 in the other specimens was negative or weakly/focally positive (Fig. 3d, f). In cases 2 and 3, the MIB-1 labeling index tended to increase along with tumor recurrence (Fig. 3g–i).

Fig. 3
figure 3

Immunohistochemistry. a, b, c STAT6 staining. Nuclear positivity is observed in the spindle tumor cells but not in the ectopic salivary glands in all 3 cases (a, case 1; b, R1 of case 2; c, P of case 3). d, e, f p53 staining. p53 is diffusely and moderately positive in R3 of case 2 (e). The expression of p53 was negative in case 1 (d) and focally/weakly positive in R4 of case 3 (f). g, h, i Ki-67 staining (g, case 1; h, R3 of case 2; i, R1 of case 3). P, primary; R1, R3, and R4, first, third, and fourth recurrence, respectively. Original magnification: d, e, f, g, h, i × 200; a, b, c × 400

Full size image

Genetic analysis

The results of genetic analysis are summarized in Table 2. RT-PCR for detecting NAB2-STAT6 was performed for all 3 cases (for cases 2 and 3, R1 and R2 were analyzed, respectively), and they all harbored NAB2 exon 4-STAT6 exon 2 (Fig. 4a–c).

Fig. 4
figure 4

Results of genetic analysis. a, b, c NAB2 exon 4-STAT6 exon 2 fusion was identified in all 3 cases (a, case 1; b, R1 of case 2; c, R2 of case 3). d TP53 mutation was detected only in R3 of case 2 (c.535C > T, H179Y)

Full size image

Direct DNA sequencing was performed for all samples available from all 3 cases except for P of case 2. Among all samples analyzed, TP53 mutation (c.535C > T, p.H179Y) was detected only in the sample that demonstrated dedifferentiation, R3 of case 2 (Fig. 4d). No TERT promoter mutation was found in any samples.

Statistical analysis

Review of the English literature yielded 78 cases of meningeal SFTs with NAB2 exon 4-STAT6 exon 2/3 and 185 cases of those with NAB2 exon 5/6/7-STAT6 exon 16/17/18 [6,7,8,9, 13, 32, 34, 37,38,39,40,41,42,43]. We explored the association of NAB2-STAT6 fusion variants versus the presence of ESGs, and the presence of ESGs was significantly associated with NAB2 exon 4-STAT6 exon 2/3 (p = 0.028).

Discussion

In the present study, we described clinicopathological and molecular features of 3 meningeal SFT with ESGs in the CPA. All 3 cases harbored NAB2 exon 4-STAT6 exon 2, which is a minor fusion variant in meningeal SFTs.

Immunohistochemically, although the spindle tumor cells were positive for STAT6, the ESG cells were negative in all 3 cases (Fig. 3a–c), supporting that the ESG cells were non-neoplastic components. That is also in line with the findings that the ESG cells demonstrated no cellular atypia and disappeared in the specimens of recurrence/dissemination of case 3. The histologies of our cases were similar to those of reported SFTs involving non-neoplastic epithelial components other than ESGs, such as eccrine glands, orthotopic minor salivary glands, kidney, Mullerian epithelium in the omentum, and prostate [44,45,46]. As discussed in the previous report of case 1, due to the lack of cellular atypia in the ESG cells, carcinosarcomas as a whole or carcinomas metastasizing in SFTs were not considered [17].

Although the number of cases investigated was small, the current study showed that the presence of ESGs was associated with NAB2 exon 4-STAT6 exon 2 in meningeal SFTs. This may reflect a difference in pathogenesis between ordinary meningeal SFTs and those with ESGs. In addition, SFTs developing in the orthotopic salivary glands are rare and 4 cases have been tested for NAB2-STAT6 fusion variants: 2 cases with NAB2 exon 6-STAT6 exon 16 and 2 cases with NAB2 exon 3-STAT6 exon 19, with the latter being a rare variant [1, 43, 47,48,49]. Therefore, SFTs in the orthotopic salivary glands and those with ESGs may show different distributions of NAB2-STAT6 fusion variants.

In case 3, ESG components were found only in the primary tumor (P) and disappeared in the recurrent and disseminated tumors (R1, R2, and R4) (Table 2, Fig.2 g–i). Considering this phenomenon, some of the meningeal SFTs with ESGs may have been diagnosed as ordinary SFTs without ESGs, as the salivary gland components disappeared like case 3 in the current study, probably because of the overrunning of ESGs by the growth of SFTs. Sampling errors, unawareness, and underestimation may also mislead the diagnosis.

We understand that the presence of ESGs contributes to the tumorigenesis of salivary gland–related tumors, such as Warthin’s tumor, pleomorphic adenoma, and mucoepidermal carcinoma; however, it is unknown whether the same is true for SFTs or whether they are only involved in the growth of tumors. If the former is true, we would expect to find SFT with ESGs in the sellar region or extracranial sites, which are more common for ESGs than CPA as mentioned before; however, there have so far not been any reports of this. Asayama et al.’s hypothesis about the developmental mechanism of SFTs with ESGs was that some salivary gland primordial cells are misplaced into dura mater during embryonic migration and then neoplastic transformation into SFTs is brought about by the result of differentiation [17]. Along with this theory, we speculated that ESGs particularly (or only) in the CPA, as well as orthotopic salivary glands, may possess precursor cells that give rise to SFT.

On the other hand, ESGs found in 3 cases of SFTs in the current study may be the pre-existing tissue that was only embedded in the growth of tumors. If so, the absence of reports of SFT with ESGs in the sellar region, except for those in extracranial sites, does not contradict the higher incidence of conventional SFTs in the CPA than in the sellar region [50, 51]. As for meningiomas, it is now widely accepted that differences in genetic mutations (NF2, KLF4/TRAF7, AKT1/TRAF7, SMO) are associated with tumor location [52, 53]. Between NAB2 exon 4-STAT6 exon 2/3 and NAB2 exon 5/6/7-STAT6 exon 16/17/18, functional differences associated with the presence or absence of CHD4-interacting domain in NAB2 which regulates activity of ERG1, an important regulator of fibrosis, were suggested [1]. Similar to meningiomas, it may be of interest to see whether the fusion gene variants of NAB2-STAT6 are associated with the location of meningeal SFTs, and whether those in the CPA predominantly have NAB2 exon 4-STAT6 exon 2/3 fusion. However, most of the literature of meningeal SFTs that analyzed NAB2-STAT6 variants did not describe the detailed site of lesion [6,7,8,9, 32, 37,38,39]; thus, further studies are needed to investigate this.

In conclusion, we report 3 meningeal SFT with ESGs harboring NAB2 exon 4-STAT6 exon 2 in the CPA; however, the etiological association between SFTs and the ESGs, the reason why ESGs only in the CPA have been found to be associated with the development of SFTs, and the association between the variant of NAB2 exon 4-STAT6 exon 2 fusion and SFT with ESGs remain to be elucidated. Further clinicopathological and genetic analyses on more cases are needed to clarify these issues.