Background: Gene fusions have emerged as critical molecular events in the pathogenesis of central nervous system (CNS) tumors, with significant implications for both diagnosis and therapy. While their utility in tumor classification is well establishe...
Background: Gene fusions have emerged as critical molecular events in the pathogenesis of central nervous system (CNS) tumors, with significant implications for both diagnosis and therapy. While their utility in tumor classification is well established, the full clinical potential of fusion-driven CNS tumors is still being elucidated. Here, we present a comprehensive institutional experience from Seoul National University Hospital (SNUH), focusing on the prevalence, actionability, and clinical outcomes associated with gene fusions in CNS tumors.
Methods: A retrospective analysis was conducted on 809 CNS tumors diagnosed at SNUH between 2017 and 2024. All cases underwent molecular profiling through next-generation sequencing (NGS) using FiRST Brain Tumor Panel, which integrates DNA and RNA sequencing to maximize fusion detection. A total of 217 fusion-positive cases were identified. Fusion-positive cases were further evaluated for therapeutic actionability according to European Society for Medical Oncology Scale for Clinical Actionability of molecular Targets (ESMO’s ESCAT) criteria. Clinical outcomes were recorded for patients who received fusion-targeted therapies.
Results: Out of 809 CNS tumors, 217 (26.8%) were identified as fusion-positive, with 153 (18.9%) meeting ESCAT tier III or higher for clinical actionability. Fusion prevalence was markedly higher in pediatric tumors (53.3%) compared to adults (11.0%, p < 0.001), and pediatric low-grade gliomas (LGGs) were particularly enriched for BRAF fusions (44%). Among gliomas, fusion detection was significantly higher in IDH-wildtype tumors (20.7%) compared to IDH-mutant tumors (1.5%). Among 14 patients treated with fusion-directed therapies, pediatric cases (n=5) demonstrated the most favorable responses: 4 out of 5 achieved stable disease, with a median overall survival of 48 months, including durable control using TRK inhibitors in NTRK-fused tumors. In contrast, adult GBMs responded poorly to target therapy. Only one ROS1-fused GBM achieved long-term survival (60 months), while others progressed within 16 to 33 months. NTRK2 was the predominant NTRK fusion type in CNS tumors (82.4% of NTRK fusions), contrasting with the dominance of NTRK3 in other solid tumors. FGFR3 and MET fusions were common in adult GBMs (32% and 18%, respectively), but showed limited response to monotherapy. One out of five multifusion cases showed a poor match upon methylation clustering, but because this was within the reported error rate of methylation clustering, the effect could not be attributed to a confounding effect by the presence of multiple concurrent fusions.
Conclusion: This study establishes the diagnostic and therapeutic utility of using NGS in CNS tumor fusion detection, with actionable fusions identified in 18.9% of SNUH cases. Pediatric LGGs (53.3% fusion prevalence) and GBMs (20.7%) showed the highest clinical relevance, supporting routine NGS screening. While pediatric patients achieved durable responses to targeted therapies like TRK inhibitors, adult GBMs exhibited limited monotherapy efficacy, underscoring the need for combinational treatment strategies. Our findings highlight institutional fusion profiling and biomarker-driven trials to expand effective targeted treatments.