Grant Number:	1R01CA312142-01 Interpret this number
Primary Investigator:	Song, Yuanquan
Organization:	Children'S Hosp Of Philadelphia
Project Title:	Multi-Modal Data Integration and Functional Screening to Discover Druggable Targets in Pediatric Low-Grade Glioma
Fiscal Year:	2026

PROJECT SUMMARY Low-grade gliomas are the most common central nervous system tumor among children, accounting for ~1/3 of pediatric brain tumors. Pediatric LGG (pLGG) is a heterogeneous disease with different molecular subtypes that have distinct genetic and molecular profiles, warranting subtype-specific targeted treatments. We bring together expertise across imaging, bioinformatics, and cancer biology teams that have previously established a model of risk in pLGG based on clinicoradiomic features. This proposal is poised to discover novel molecular events in pLGG that can subsequently be tested for markers of infiltration and cell proliferation in pLGG fly and organoid models. Our central hypothesis posits that clinicoradiomic risk groups harbor unique molecular events spanning across the genome, transcriptome, and epigenome that influence pLGG progression, and the effects of these molecular events can be preclinically tested and validated in fly and organoid models. In aim 1, we will leverage our pretrained clinicoradiomic risk model of progression to classify newly acquired pLGG tumors of the Children’s Brain Tumor Network into low, medium, and high-risk groups, forming the impetus for the identification of novel multi-omic targets. Leveraging preclinical drug screening data from the Childhood Cancer Model Atlas, we will train a preclinical multi-omic predictive model of drug response to identify potentially efficacious small molecules specific to each risk group. With the aim of addressing pLGG molecular heterogeneity and of implementing a preclinical precision medicine approach, we will then expand on this analysis with multi-omic clustering of clinicoradiomic risk groups to informatively subdivide risk groups and identify highly specific molecular aberrations that may better inform treatment. In aim 2, we will functionally interrogate the molecular targets in fly pLGG models. We will screen over 50 conserved candidate genes in our fly models to determine their potential roles in pLGG initiation, proliferation, and infiltration. We will focus on a specific group of genes to gain further mechanistic understanding of the metabolic vulnerability in various pLGG subtypes. In aim 3, to further determine the translational potential of our targets, we will verify their function in mediating tumor proliferation and invasion using tumor organoids derived from pLGG patients. We will further explore metabolic reprogramming in the treatment of pLGG. Although patient-derived organoids and other human-relevant NAMs are valuable validation platforms, the current pLGG model landscape remains constrained by scarce, difficult-to-maintain, and subtype-biased cell lines/organoids that are not well suited for rapid functional screening (PMC10628935). Building on the well-established use of fly tumor models (PMC9393232) to interrogate conserved oncogenic pathways in vivo, our fly pLGG pipeline is absolutely essential for high-throughput screening of conserved candidates, followed by testing positive hits in patient tumor organoids. Our holistic strategy, from patients to bioinformatics to molecular and functional interrogation and back to patients, has the potential to create a new paradigm for the standard of care of pLGG, improving outcomes and quality of life by discovering new therapeutic vulnerabilities.





None