Abstract
Background/Aim: Gliomas are primary malignancies of the central nervous system (CNS). High-grade gliomas are associated with poor prognosis and modest survival rates despite intensive multimodal treatment strategies. Targeting gene fusions is an emerging therapeutic approach for gliomas that allows application of personalized medicine principles. The aim of this study was to identify detectable fusion oncogenes that could serve as predictors of currently available or newly developed targeted therapeutics in cross-sectional samples from glioma patients using next-generation sequencing (NGS). Patients and Methods: A total of 637 patients with glial and glioneuronal tumours of the CNS who underwent tumour resection between 2017 and 2020 were enrolled. Detection of fusion transcripts in FFPE tumour tissue was performed by a TruSight Tumour 170 assay and two FusionPlex kits, Solid Tumour and Comprehensive Thyroid and Lung. Results: Oncogene fusions were identified in 33 patients. The most common fusion was the KIAA1549-BRAF fusion, detected in 13 patients, followed by FGFR fusions (FGFR1-TACC1, FGFR2-CTNNA3, FGFR3-TACC3, FGFR3-CKAP5, FGFR3-AMBRA1), identified in 10 patients. Other oncogene fusions were also infrequently diagnosed, including MET fusions (SRPK2-MET and PTPRZ1-MET) in 2 patients, C11orf95-RELA fusions in 2 patients, EGFR-SEPT14 fusion in 2 patients, and individual cases of SRGAP3-BRAF, RAF1-TRIM2, EWSR1-PALGL1 and TERT-ALK fusions. Conclusion: The introduction of NGS techniques provides additional information about tumour molecular alterations that can aid the multimodal management of glioma patients. Patients with gliomas positive for particular targetable gene fusions may benefit from experimental therapeutics, enhancing their quality of life and prolonging survival rates.
Gliomas are primary malignancies of the central nervous system (CNS) (1). High-grade gliomas are associated with poor prognosis and modest survival rates, despite intensive multimodal treatment strategies comprising tumour resection and subsequent chemoradiotherapy. Novel, innovative approaches are needed to advance glioma management and improve patient prognosis.
A growing number of genetic alterations were identified as clinically relevant diagnostic, prognostic, or predictive biomarkers with regard to CNS tumours. This resulted in the 2016 WHO classification system for CNS tumours, integrating the histopathological evaluation of tumour tissue and various molecular biomarkers to provide the accurate final tumour diagnosis and better determine patient prognosis (1). Due to the enormous effort in glioma basic research, other molecular biomarkers are continuously being discovered that further specify the classification of tumour entities. These efforts resulted into the most recent WHO classification of CNS tumours in 2021 (2). The ever-increasing number of glioma biomarkers might allow personalization of treatment algorithms using targeted anticancer therapeutics and immunotherapy that is a substantial focus in neurooncological research (3–9).
Among other genetic alterations, gene fusions are an emerging personalized therapeutic target in patients with various cancers (10, 11). The development of next-generation sequencing (NGS) allowed identification of many gene rearrangements encoding novel oncogenic fusions in gliomas (12). Gene fusions represent hybrids of two coding or regulatory DNA sequences between genes. They belong to the pathognomonic mutations resulting from genomic rearrangements such as translocations, deletions, duplications, or inversions (12). Although most gene fusions represent passenger mutations without direct oncogenic functions, some are strong driver alterations playing role in the initial steps of tumour development, while other fusions may play important roles in tumour progression (13, 14). Recently, potentially druggable gene fusions were identified in 4.4-11% of glioblastomas, 12% of anaplastic gliomas, and 1.5% to 8% of grade II astrocytomas (14, 15). Despite the relatively low-frequency rate of gene fusion in gliomas, they provide potential targets for therapy in selected patients with specific fusions.
The aim of this study was to identify detectable oncogene fusions in cross-sectional samples from glioma patients using NGS. This article also discusses early publications regarding anticancer therapeutics targeting oncogenic fusions in CNS gliomas.
Patients and Methods
Patients. A total of 637 patients with glial and glioneuronal tumours of the CNS who underwent neurosurgery between 2017 and 2020 were enrolled. Tumours were evaluated histopathologically and categorized according to the WHO classification system (1). The institutional review board and local ethics committee of the University Hospital in Pilsen approved the study (ethical code no. 01082013). All patients signed informed consent forms for the use of their tumour tissue samples in clinical research.
Next generation sequencing (NGS). Fusion transcript detection was conducted using a TruSight Tumour 170 assay (Illumina, San Diego, CA, USA) to identify molecular predictors of treatment. For NGS analysis, 2-3 formalin-fixed, paraffin-embedded sections (10 μm thick) were macrodissected to isolate tumor-rich regions.
The extraction of total nucleic acids was performed using an FFPE DNA kit (automated on RSC 48 Instrument, Promega, USA). Purified DNA and RNA were quantified by Qubit Broad Range DNA and RNA assays, respectively (Thermo Fisher Scientific, Waltham, MA, USA). The DNA quality was evaluated by the FFPE QC kit (Illumina, San Diego, CA, USA), and the RNA quality was evaluated by the Agilent RNA ScreenTape Assay (Agilent, Santa Clara, CA, USA). DNA samples with Cq <5 and RNA samples with DV200 ≥20 were used for further analysis. After DNA enzymatic fragmentation with a KAPA FragKit (KAPA Biosystems, Washington, MA, USA), DNA and RNA libraries were prepared with the TruSight Tumour 170 assay (Illumina) according to the manufacturer’s protocol. Sequencing was performed on a NextSeq 500 sequencer (Illumina) as per the manufacturer’s recommendations. TruSight Tumour 170 data were analyzed using BaseSpace Sequence Hub (Illumina). DNA variant filtering and annotation were performed using Variant Interpreter (Illumina), a cloud-based tool. A custom variant filter was set up including only variants with coding consequences and GnomAD database (16) frequency values less than 0.01. The remaining subset of variants was checked visually, and suspected artefactual variants were excluded.
In addition, fusion transcript detection was performed using FusionPlex kits (Solid Tumour and Comprehensive Thyroid and Lung) as previously described (17). Total nucleic acid was extracted from tumor samples using Agencourt FormaPure Kit (Beckman Coulter, Brea, CA, USA) following the corresponding protocol, with an overnight digestion and an additional 80°C incubation as described in the modification of the protocol required by ArcherDX (ArcherDX Inc., Boulder, CO, USA). RNA component of the total nucleic acid was quantified using the Qubit Broad Range RNA Assay Kit (Thermo Fisher Scientific). Input of 250 ng of formalin-fixed, paraffin-embedded RNA was used for library preparation of the FusionPlex kits. PreSeq RNA QC Assay using iTaq Universal SYBR Green Supermix (Biorad, Hercules, CA, USA) was performed on all samples following the Archer Fusion Plex Protocol for Illumina (ArcherDX Inc.). Final libraries were quantified using Library Quantification for Illumina Libraries kit assuming a 200-bp fragment length (KAPA, Wilmington, MA, USA). Libraries were sequenced on a NextSeq500 sequencer (Illumina). Analysis of sequencing results was performed using the Archer Analysis software (v5; ArcherDX Inc.).
Institutional Review Board Statement. The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the institutional review board and local ethics committee of the University Hospital in Pilsen. The ethical code is no. 01082013. Informed consent was obtained from all subjects involved in the study.
Results
Patients and gene fusion prevalence. A total of 637 patients with glial and glioneuronal tumours of the CNS were enrolled during the study period. Molecular genetic analysis was possible for 362 patients (57%; 362/637). Specific gene fusions were detected in 33 patients (5.2%; 33/637) (Table I). The main reasons that molecular genetic examination was not possible for some enrolled patients were poor sample quality, insufficient sample amount, and problems in the extraction process.
Patients with detected gene fusions, their histopathological diagnosis and opportunity for experimental targeted therapeutics.
Overview of detected gene fusions and corresponding tumor types. The most frequent fusion was KIAA1549-BRAF, which was detected in 13 patients (9 patients with pilocytic astrocytomas, 1 patient with anaplastic pilocytic astrocytoma, 1 patient with pilomyxoid astrocytoma, 1 patient with diffuse leptomeningeal glioneuronal tumour, and 1 patient with ganglioglioma). The second most frequent aberrations were FGFR fusions (FGFR1-TACC1, FGFR2-CTNNA3, FGFR3-TACC3, FGFR3-CKAP5, FGFR3-AMBRA1), that were observed in 10 patients (4 patients with glioblastoma, 2 patients with anaplastic astrocytoma, 1 patient with anaplastic oligodendroglioma, 1 patient with diffuse astrocytoma, and 1 patient with pilocytic astrocytoma). Other infrequent oncogene fusions were also diagnosed. Two patients with glioblastomas had tumours positive for MET fusions (PTPRZ1-MET and SRPK2-MET fusions), 2 anaplastic ependymomas were positive for C11orf95-RELA fusions, 2 glioblastomas were positive for EGFR-SEPT14 fusions, 2 pilocytic astrocytomas were positive for SRGAP3-BRAF and RAF1-TRIM2 fusions, 1 low-grade glial tumour not-otherwise specified was positive for the EWSR1-PALGL1 fusion, and 1 glioneuronal tumour containing neuropil-like islands was positive for the TERT-ALK fusion.
Treatment applied to patients with detected fusions. The standard treatment modalities among patients with detected gene fusions comprised neurosurgery (n=33), adjuvant chemotherapy and radiotherapy (n=14), radiotherapy alone (n=5) and chemotherapy alone (n=1). Thirteen patients received no adjuvant therapy after neurosurgery (mainly patients with pilocytic astrocytomas and ganglioglioma).
Enrolment in the pan-FGFR inhibitor clinical trial was discussed at the molecular tumour board for two patients with FGFR rearranged tumours. Similarly, a clinical trial with an ALK inhibitor was discussed for 1 patient with an ALK fusion-positive tumour. However, rapid disease progression or disagreement with the use of experimental therapeutics ultimately precluded the inclusion of these patients into clinical trials.
Discussion
Glial tumours represent a wide group of primary CNS malignancies that can be categorized into “low grade” and “high grade” glioma subgroups based on the aggressiveness of the disease and corresponding patient prognosis (1, 2). High-grade gliomas (such as anaplastic astrocytoma or glioblastoma multiforme - GBM) are the most malignant and are associated with poor prognosis, rapid clinical deterioration, and high mortality rates (7). The survival of GBM patients remains 12.1-14.6 months regardless of intensive multimodal treatments, and only 3-5% of patients survive long-term (18).
In this study, the most frequently detected fusion was KIAA1549-BRAF, which was found in 13 patients (mean age=12.6 years). The majority (9/13) of patients had pilocytic astrocytoma. An additional patient with pilocytic astrocytoma was diagnosed with an alternative BRAF fusion, SRGAP3-BRAF. A fusion oncogene involving KIAA1549 and BRAF arising from tandem duplication at locus 7q34 was observed by Jones et al. (19). KIAA1549-BRAF fusions were identified in 66% of pilocytic astrocytomas but were not found among high-grade gliomas. KIAA1549-BRAF fusions can be used as a diagnostic biomarker to aid the diagnosis of low-grade gliomas, particularly pilocytic astrocytomas that can show necrosis and microvascular proliferation similar to that of high-grade gliomas. These fusions lead to constitutive BRAF kinase activity that promotes glioma oncogenesis (20). Anticancer treatment that targets the mitogen-activated protein kinase (MAPK) pathway, such as BRAF or MEK inhibitors, may prevent the growth of KIAA1549-BRAF fusion-positive tumours (21). In particular, second-generation BRAF inhibitors, such as PLX-PB3, which specifically targets the fusion protein, may be beneficial if it is impossible to remove the whole tumour by radical surgery (22).
The second most common glioma gene fusions identified in this study involved the fibroblast growth factor receptor genes FGFR1, FGFR2 and FGFR3. FGFR fusions were detected in 10 patients (mean age 51.3 years), predominantly in high-grade gliomas (4 glioblastomas, 2 anaplastic astrocytomas and 1 anaplastic oligodendroglioma). The FGFR tyrosine kinase coding domains are most often found to be fused to the TACC1 or TACC3 transforming acidic coiled-coil (TACC) coding domains in FGFR-TACC fusions (23). The abnormal expression of FGFR3-TACC3 can activate various oncogenic cell signalling pathways, promoting glioma development (24). These aberrations are more specific to high-grade gliomas (1.2% to 8.3% of glioblastoma patients are positive for FGFR-TACC fusions) (12). Additionally, three alternative FGFR fusions were identified in this study (FGFR2-CTNNA3, FGFR3-CKAP5 and FGFR3-AMBRA1). The personalized medicine approach using inhibitors that target FGFR fusions may improve the prognosis of patients with glial tumours (25). Various targeted therapies have been evaluated in clinical trials for patients with FGFR-fused gliomas, such as erdafitinib (NCT04083976), ponatinib (NCT02478164), and infigratinib (NCT01975701).
Other oncogene fusions were also identified in this study that serve as potential druggable targets, such as EGFR (EGFR-SEPT14) and MET (SRPK2-MET, PTPRZ1-MET) fusions, identified in 4 glioblastoma patients. The EGFR-SEPT14 fusion gene is an in-frame fusion with C-terminal deletion of EGFR that is present in approximately 4% of glioblastomas (26). This fusion gene can activate the STAT3 signalling pathway and increase tumour proliferation and migration. Preclinical evidence of the benefit of EGFR inhibitors, such as lapatinib and erlotinib, was observed in EGFR-SEPT14 gliomas (12, 26, 27). However, results from clinical trials focused specifically on EGFR fusions in gliomas are generally lacking. MET fusions in adult glioma patients were identified in the Chinese Glioma Genome Atlas database, where the PTPRZ1-MET fusion was observed in 3% of glioblastomas (28). This fusion can increase the expression and phosphorylation of the MET oncoprotein. PTPRZ1-MET fusion was also linked to poor patient prognosis. MET fusions that can activate MAPK signalling, compromise cell-cycle regulation, and induce aggressive glial tumours have also been observed in up to 10% of paediatric glioblastomas (29). The MET and VEGFR2 inhibitor foretinib was found to be effective for the treatment of MET fusion-positive tumour growth in preclinical experiments. The ALK, MET and ROS1 inhibitor crizotinib was found to be effective in an 8-year-old glioblastoma patient with the PTPRZ1-MET fusion; however, this benefit was temporary (29). Clinical trials evaluating MET inhibitors, such as crizotinib (NCT02270034), bozitinib (NCT02978261), cabozatinib (NCT00704288), SGX523 (NCT00607399, NCT00606879) and others, for the treatment of glioma are ongoing.
One patient was diagnosed with a TERT-ALK fusion-positive glioneuronal tumour containing neuropil-like islands in this study. Targeted ALK inhibitors, such as ceritinib, alectinib, brigatinib, entrectinib or crizotinib, hypothetically benefit glioma patients. These therapeutics have been successfully used in non-small cell lung cancer patients with ALK fusions (30). However, the effect of ALK inhibitors on gliomas has yet to be proven in clinical trials (31).
The exact role of gene fusions in glioma development and progression and the effects of possible therapeutic interventions should be further evaluated in in vitro studies.
Substantial progress has been made in glioma genetics and epigenetics research, and the availability of inhibitors of aberrantly activated oncogenes has increased. The development of sequencing techniques has facilitated large multiplatform studies and revealed several aberrations in a variety of mutated genes and cell signal pathways involved in oncogenesis of low-grade (27) and high-grade gliomas (32–34). A growing number of genetic alterations that were identified and are applicable as clinically relevant diagnostic, prognostic or predictive biomarkers resulted in the 2016 WHO classification system for CNS tumours. For the first time, the so-called “integrated diagnostics” of CNS tumours was introduced into clinical practice, combining the histopathological evaluation of tumour tissue and various molecular biomarkers to accurately diagnose the final tumour and better determine patient prognosis (1). Due to the enormous effort in glioma basic research, other molecular biomarkers are continuously being discovered that further specify the classification of tumour entities. The international consortium cIMPACT-NOW (The Consortium to Inform Molecular and Practical Approaches to CNS Tumour Taxonomy – not official WHO) aims to bring new glioma biomarkers closer to use in clinical practice (35). The 7th update introduced novel biomarkers for ependymomas (36). Substantial progress in biomarker research and molecular genetics led to the most recent WHO classification of CNS tumours in 2021 (2).
The ever-increasing number of glioma biomarkers used in the clinical decision-making process allows application of personalized medicine principles for patients with these tumours. Various anticancer therapies targeting aberrantly activated oncogenes that may further enhance the treatment options for CNS tumours are in development and evaluation processes and may significantly improve patient prognoses, especially in high-grade glioma (5, 7, 37, 38).
Conclusion
More specific stratification of patients with CNS gliomas using advanced technologies could help adapt treatment strategies to individual molecular genetic profiles, as presented in this study. Gene fusions were revealed by genomic and transcriptomic analyses of gliomas. Despite the relatively low frequency rate, gene fusions are important for understanding the mechanisms of glioma tumorigenesis. Moreover, the findings of this study may provide potential therapeutic targets for specific patients with tumours positive for particular fusions (12, 14). Targeting gene fusions by various anticancer therapies is a successful treatment strategy for other solid malignancies and haematological malignancies, such as chronic myeloid leukaemia with the BCR-ABL1 fusion (39), non-small cell lung cancer positive for ALK, ROS1 or RET fusions (40), and cholangiocarcinoma with FGFR2 fusions (41). Tumour-agnostic therapy has also been used for malignancies with NTRK fusions (42). Therefore, gene fusions are emerging targets for next-generation personalized medicine strategies for cancer patients. Glioma patients with tumours positive for particular gene fusions could benefit from experimental therapeutics with improved quality of life and a prolonged survival time. However, these therapeutics must first be evaluated in large prospective multicentre clinical trials. The possibility of including patients in ongoing clinical trials that use novel targeted therapeutics needs to be considered in each case, especially if standard treatment options have been exhausted.
Even though gene fusions in gliomas are relatively rare, the potential for targeted therapy (if successfully evaluated in large clinical trials) emphasizes the need to incorporate gene fusion analysis into glioma diagnostic procedures. Screening for gene fusions may be more effective in specific scenarios, such as when such aberrations are expected (e.g., BRAF fusions in pilocytic astrocytomas) or when morphologically atypical gliomas are negative for traditional molecular biomarkers. However, in the context of further technological advances and the increased availability of NGS to clinical laboratories for routine tests, screening for gene fusions could be feasible for the majority of glioma patients.
Acknowledgements
This research was supported by Charles University Research Fund (Progres Q39), and by the Cooperatio Program, research area MED/DIAG, and by the grant of Ministry of Health of the Czech Republic – Conceptual Development of Research Organization (Faculty Hospital in Pilsen - FNPl, 00669806).
Footnotes
Authors’ Contributions
Conceptualization: JP, MS, VP, JM; Methodology: JP, MS, PM, NP, PP; Formal analysis: JP, MS, DS, MSB, PP, PK; Funding acquisition: JP, VP, PM, NP, MP, OT; Project administration: JP, MS, VP, MP; Resources: JP, MS, OT; Software: PM, NP; Supervision: JP, MS, PP, OT; Writing - original draft: JP, PM; Writing - review & editing: JP, MS.
Conflicts of Interest
The Authors declare no conflicts of interest.
- Received September 30, 2021.
- Revision received February 1, 2022.
- Accepted March 8, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.