Introduction
Emerging technological advancements have played a crucial role in delineating a classification system for tumors affecting the central nervous system (CNS). The fifth edition of the World Health Organization (WHO) classification of CNS tumors, revised in 2021, has been notably influenced by the integration of molecular results derived from DNA and RNA-based molecular methodologies, along with DNA methylation profiling [
13]. These techniques have proven to be essential in characterizing CNS neoplasms and revealing crucial driver events, including oncogenic gene fusions. With the application of genome-wide and non-targeted methodologies for fusion detection, novel fusion events are being discovered in routine clinical diagnostic settings.
Pediatric low-grade gliomas (pLGG) and glioneuronal tumors (GNT) comprise over 30% of pediatric CNS tumors [
3]. Within this category, GNTs pose a considerable diagnostic challenge because they lack consistent distinguishing histological characteristics. Several histological subtypes are acknowledged, yet in clinical practice, their differentiation is often challenging [
25,
26].
A substantial number of pLGG/GNTs are associated with oncogenic fusion events. The most commonly observed fusions in pLGG/GNTs involve
BRAF,
FGFR1,
MYB, and
MYBL1, which result in up-regulation of the RAS-mitogen-activated protein kinase (RAS/MAPK) and PI3K pathways [
15,
19‐
21]. Receptor tyrosine kinase (RTK) fusions, such as those involving
MET,
ALK,
ROS1, and
NTRK, drive a group of infantile hemispheric gliomas, but are generally rare in pLGG/GNTs, accounting for less than 5% of cases [
4,
9,
21].
Mesenchymal–epithelial transition factor (
MET) encodes an RTK which activates MAPK, PI3K/AKT, SRC, and STAT pathways to promote cell proliferation, invasion, and angiogenesis [
12,
17,
23].
MET fusions, activating mutations, exon 14 skipping, and amplifications, leading to MET overexpression have been identified in a variety of human cancers [
14,
28]. In the context of CNS tumors,
MET fusions, with different 5’ partner genes, have been predominantly observed in high-grade gliomas, with a notable prevalence in infantile high-grade gliomas in the pediatric setting. However,
MET fusions have not been commonly associated with pLGG/GNTs, and only several cases exist that describe their presence in low-grade GNTs [
4,
8,
24].
We present two novel cases of pediatric glioneuronal tumors with a
CLIP2::
MET fusion detected by whole transcriptome sequencing (RNAseq), along with their clinical, pathologic, and molecular findings. While the
CLIP2::
MET fusion has been previously reported in three instances, including an adult glioneuronal tumor [
8], a case of spontaneous regression of a congenital high-grade glioma [
18], and at least two cases of infantile hemispheric high-grade glioma [
1,
6,
9], this fusion has not been described in pediatric GNTs to date.
Discussion and conclusions
In CNS tumors,
MET fusions with different 5’ partner genes, have been predominantly observed in high-grade gliomas and have been reported to demonstrate aggressive biological behavior [
16,
29,
30]. In the pediatric setting, there is a notable prevalence of
MET fusions in infantile high-grade gliomas [
9,
11,
27]. However,
MET fusions have not been commonly associated with low-grade gliomas or, even more rarely, glioneuronal tumors. In a study of 1,000 low-grade gliomas by Ryall et al., RTK fusions were identified in < 5% of the cases evaluated, and
MET fusions in less than 1% of the cases [
21]. To date, only several cases of
MET fusions have been reported in low-grade GNTs [
4,
8,
24].
Histologically, both of our cases demonstrated low-grade features, including the absence of mitotic activity and no evidence of microvascular proliferation. There was evidence of glial and neuronal differentiation based on morphology and immunohistochemical stains. High-grade features were not observed in our cases, whereas CLIP2::MET fusion cases previously described in the literature showed high cellularity, brisk mitotic activity, and microvascular proliferation [
9,
18]. Only a single glioneuronal tumor reported by Chowdhury et al. showed similar low-grade features to our cases, characterized by a low Ki-67 labeling index [
8].
The
CLIP2::
MET fusion was identified by RNAseq in both of our cases. Neither was initially detected by OncoKids, as expected, since RNA sequencing with OncoKids is a targeted approach, and novel fusions may be missed by these targeted methods. Incorporating a genome-wide analysis approach into routine clinical diagnostics is imperative for the identification of such fusions. While the exact breakpoints at the DNA level are unknown, the exon-exon pair of the fusion is identical between the two cases, with exon 11 of
CLIP2 fused with exon 15 of the
MET gene. The resulting fusion is in-frame, contains the protein kinase domain of MET, and is predicted to result in the upregulation of the MAPK pathway for tumorigenesis [
8]. By RNAseq gene expression analysis, the presence of the differentially expressed 3’
MET (exons 15–21) between 5’
MET (exons 1–14) further supported the oncogenic impact of this fusion.
Notably, methylation profiling of both tumors was consistently clustered with low-grade glial/glioneuronal tumors across classifiers, such that Case 1 was clustered with PA/GG and Case 2 clustered with DNT, although the specific subclass was indeterminate on multiple classifiers. Our cases, in conjunction with the previously reported cases with
CLIP2::
MET fusion, including IHGs, suggest a variable methylation pattern. Additional cases may be helpful in determining whether these
CLIP2::
MET fusion-positive tumors have a specific methylation pattern in different age groups, such as IHG in infants, LGG/GNT in pediatric patients, and variable histology in adults. Interestingly, the same methylation clustering pattern seen in Case 2 has also been previously identified in a DNT by Chowdhury et al., but in a 30-year-old female patient [
8].
Both the
CLIP2 and
MET genes are located in chromosome 7, with
CLIP2 being more proximal (7q11.23) and
MET more distal (7q31.2). The fusion is thus considered to result from intrachromosomal rearrangement(s). It is worth noting that Case 2 had an interstitial deletion on chromosome 7, with a breakpoint located within the
CLIP2 gene, suggesting potential involvement of
CLIP2 by CMA analysis. However, the absence of copy number variants involving
CLIP2 or
MET in chromosome 7 does not rule out the presence of the
CLIP2::
MET fusion, as seen in Case 1. Nonetheless, both cases had chromosome 7 copy number alterations, which may be secondary to the intrachromosomal rearrangement(s). Although information regarding copy number alterations in
CLIP2::
MET fusion-positive tumors is limited, due to the limited total number of cases, chromosomes 1, 7, and 22 copy number alterations appear to be common, as observed in both our cases and the GNT case reported by Chowdhury et al. [
8].
The clinical course in both of our cases appears to follow that of low-grade tumors. In Case 1, the tumor presented post-hemorrhage, and the hydrocephalus was treated with a ventricular-peritoneal shunt. The patient then underwent one gross-total resection. In Case 2, the patient had been followed by neurology for his chronic eye twitching, and once symptoms became worse, a mass was visualized on MRI, and a single near-gross total resection was accomplished. Both patients improved from baseline clinical status at the last follow-up, and they are being followed with serial imaging. Neither is undergoing adjuvant therapy.
Stucklin et al. reported two infantile high-grade hemispheric gliomas with a CLIP2::MET fusion. The first case underwent two resections and chemotherapy, while the second case underwent one resection and chemotherapy. These cases were evaluated retrospectively, limiting the clinical data gathered and investigated. At present, Chowdhury et al. is the only case in the literature with histologic similarity to our cases; however, the tumor arose in a 30-year-old female with dysphasia and right arm pain, whereas our cases arose in pediatric patients with varied symptomatology. The group’s MRI findings did show similarities to our cases, with solid and cystic compartments noted in the left parietal lobe, although they do not describe any hemorrhage in their case of adult GNT with CLIP2::MET fusion. The patient underwent gross-total resection, and no adjuvant treatment was administered. The patient has remained in remission for 7 years.
Interestingly, Riedmeier et al. present clinical parallels to our Case 1 in their case report on infantile high-grade glioma with a CLIP2::MET fusion. The patient presented with hydrocephalic complications and IVH. The mass on MRI presented with high-grade features, and a biopsy was completed, which demonstrated a congenital anaplastic astrocytoma and glioblastoma. Without further intervention, the mass underwent spontaneous regression by a 10-week follow-up.
Prior reports of CLIP2::MET fusion have displayed tumors in various locations, including the frontal, temporal, occipital, and parietal lobes. These investigations on CLIP2::MET fusions have discernible differences in both histological and clinical attributes. Chowdhury et al. and our study are the only published reports for GNT with CLIP2::MET fusion, and all occurred, specifically in the occipital and parietal-occipital lobes, and presenting with seizure, ocular, and/or motor dysfunction. Both of our cases show a favorable outcome with surgical resection and without adjuvant therapy; however, more cases are needed to more firmly establish clinical outcomes of patients with GNT harboring this fusion.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.