Nimg-32. Multi-modal Analysis Uncovers Differentially Regulated Pathways Associated With Radio-phenotypes in Pediatric Low-grade Glioma
Year: 2023
Type: conference-abstract
Abstract: INTRODUCTION Pediatric low-grade glioma is comprised of a broad variety of histological subtypes with heterogeneous responses to standard-of-care treatments. Understanding the biological pathways underlying clinical radiographic appearances, or radio-phenotypes, may facilitate precision diagnostics prior to the initiation of any interventions, thereby, reducing the risk of co-morbidities while improving patient outcome. Here, we aim to comprehensively harness the high-throughput multi-modal data available through Children’s Brain Tumor Network (CBTN) to elucidate the pathways that form radio-phenotypic traits of pLGG. METHODS On a cohort of 165 patients with pLGG, 881 radiomic features were extracted from multiparametric MRI sequences. We applied principal component analysis (PCA) followed by K-Means clustering to the radiomic features along with clinical variables of age, gender, and tumor location to group the patients into distinct imaging subtypes. We used whole transcriptomic data from OpenTargets to compare differential expression and co-expression of network- and pathway-level properties via Gene Set Enrichment Analysis (GSEA), and Gene Sets Net Correlations Analysis (GSNCA) among the discovered imaging subtypes. RESULTS Three imaging subtypes of pLGG were identified, among which subtype1 tumors showed increased proliferative capacity and invasive potential. This was suggested by higher expression of cell cycle regulatory, extracellular matrix (ECM) remodeling, and cell migratory pathways in subtype1 than subtypes 2 and 3. Furthermore, subtype1 exhibited differential TNF/TNFR signaling compared to subtype2, potentially related to differential regulation of pro-tumorigenic immune pathways, and differential regulation of TGF-beta and Aurora Kinase A signaling compared to subtype3, associated with differences in proliferative pathways. Subtypes 1 and 2 exhibit higher pro-tumorigenic cytokine expression while subtype3 involves higher expression of neurotransmitter release and neuronal activity, which may be targetable. CONCLUSIONS Our multi-modal radiogenomic analysis unraveled molecular underpinnings that give rise to the formation of distinct imaging phenotypes, with the potential to inform personalized treatments for children with brain tumors.
Source: Neuro-Oncology
Language: English
Subfield: Radiology, Nuclear Medicine and Imaging
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W4388589236