Abstract 1089: Developing subtype-specific pediatric cancer molecular targets by aggregating diverse genomic data resources to the pediatric cancer (PeCan) knowledge base portal

Abstract Knowledge about molecular targets for pediatric cancer has accelerated exponentially in recent years thanks to the increased application of multi-omics profiling in both research and clinical settings. The efficacy of genomic-based clinical interventions may depend on whether observed genomic abnormalities are fundamental to the pathogenesis of a specific cancer subtype. At present such information is limited in pediatric cancer due to the rapid evolution of subtype discovery and classification as well as disease heterogeneity owing to the presence of many rare cancer types. To develop a comprehensive resource of molecular targets in pediatric cancer subtypes, we have integrated somatic variants in pediatric cancer samples (n=∼5, 500) currently hosted in PeCan Portal (http://pecan.stjude.coud) with those (n=∼1, 400) profiled by NCI’s Childhood Cancer Data Initiative (CCDI) molecular characterization program. To leverage the recently developed WHO CNS5 classification of CNS tumors, we reclassified 2, 167 relevant samples based on molecular drivers and histology as a major update to our pediatric cancer ontology definitions, which have recently become a shared standard in the CCDI community. To determine driver genes in disease pathways relevant to specific cancer types, we analyzed the pathogenicity of SNVs and indels, CNVs, SVs and gene fusions and integrated the findings with driver genes reported in 176 publications. A total of 469 distinct driver genes were mapped to 40 pathways across the 22 major pediatric cancer types, which include 3 types of hematological malignancies (HM), 10 types of CNS tumor and 9 types of solid tumor (ST). Pathways with driver genes disrupted across the broad spectrum of cancer types include cell cycle, chromatin remodeling, PI-3K, RAS signaling and telomere maintenance. By contrast the vast majority (>90%) of the transcription factors (TF) are lineage-specific— among the 128 TF genes identified as driver genes, only 11 affect multiple lineages. These include MYC/MYCN which affect HM, ST and CNS tumors; ERG, EWSR1, FLI1 and WT1 which affect HM and ST; and KLF4, MED12, MN1, MYB and TCF4 which affect CNS and HM. These results highlight the importance of developing an infrastructure for exploring mutation prevalence in driver genes and pathways at the cancer subtype level, which will be supported in the upcoming release of the PeCan portal. This, coupled with new features for dynamic exploration of mutational signatures, global expression maps and gene-level transcription across the cancer subtype classification tree, will greatly expand our knowledge of molecular targets and vulnerabilities in pediatric cancer, paving the way for novel therapeutic advancements Citation Format: David B. Finkelstein, Delaram Rahbarinia, Ramzi Alsallaq, Stephanie Sandor, Michael Macias, Jian Wang, Sue Qiu, Brian Curran, Michael Edmonson, Andrew Thrasher, Jobin Sunny, James Madson, Scott Foy, Brent Orr, Mark R. Wilkinson, Clay McLeod, Michael Rusch, Xiaotu Ma, Jinghui Zhang. Developing subtype-specific pediatric cancer molecular targets by aggregating diverse genomic data resources to the pediatric cancer (PeCan) knowledge base portal [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 1089.