ID #483 Development of a platform approach to target fusion-driven central nervous system tumors
Monica Pomaville, Hyojeong Hwang, Derek Wong, Jinhua Wu, Jeffrey Schubert, Brandi Nelson, Alexis Boulter, Yiming Zhong, Minjie Luo, Lea Surrey, Chelsea Kotch, Cassie Kline, Peter Madsen, Marilyn Li, Kathy Liu, Jessica FosterAbstract
Background
Oncogenic fusions drive a substantial proportion of pediatric central nervous system (CNS) tumors. Although these fusions represent appealing therapeutic targets, directly targeting fusion oncogenes remains challenging due to the presence of diverse fusion partners, structural protein barriers, and the emergence of therapeutic resistance. We hypothesized that targeting fusion transcripts with antisense oligonucleotides (ASOs) could overcome these limitations.
Methods
To identify sequence features of CNS fusions, we analyzed CLIA-certified RNA-sequencing data of 1,690 CNS tumors diagnosed at Children’s Hospital of Philadelphia (patients aged 0-91 years, tumor grades 1-4). A DNA:RNA hybrid gapmer ASO therapeutic platform with modifications optimized to induce target knockdown was developed to design an ASO targeting MYB::QKI. ASO toxicity was evaluated in NOD scid gamma (NSG) and C57BL/6 mice via intracerebroventricular injection. A mouse model was generated by injecting NIH-3T3 cells containing the MYB::QKI fusion directly to the pons via stereotactic intracranial injection.
Results
Precise fusion sequences were identified and categorized by histologic diagnosis, molecular classification, and sequence conservation to identify therapeutically actionable targets. Fusions were identified in 21% of patients across all CNS histologic subtypes. Fusions were most prevalent in glial tumors (27%, 328/1213) and contained both conserved and disparate breakpoints among shared fusion partners. The MYB::QKI fusion contained conserved breakpoints and given the absence of available targeted therapies, it was uniquely suited for direct fusion-transcript targeting using ASOs. A representative MYB::QKI ASO reduced the mean fluorescent fusion signal by 83% (p < 0.05) in cells engineered to express fluorescently labelled MYB::QKI, and decreased transcript expression by 63% (p < 0.05) versus non-targeting ASO by RT-qPCR. ASOs did not induce toxicity when administered to NSG or C57BL/6 mice. A MYB::QKI-containing NIH-3T3 cell line injected into the pons achieved persistent tumor growth over 75 days and will be used for future in-vivo ASO testing.
Conclusion
Pediatric CNS tumor fusion oncogenes comprise diverse partners spanning multiple histologic diagnoses. Directly targeting fusions such as MYB::QKI via an ASO platform represents a promising strategy to address the therapeutic gap for patients with fusion-driven malignancies.