ABSTRACT Although anti-Disialoganglioside (anti-GD2) therapy has significantly improved the survival rates of children with High-Risk Neuroblastoma (HR-NBL), its clinical utility is severely limited by its life-threatening side effects and variable response rates. Despite being standard of care for HR-NBL for over 10 years, there are currently no existing mechanisms to predict whether a child will respond to anti-GD2 therapy. The long-term goal is to identify predictive biomarkers for response to anti-GD2 therapy and establish a comprehensive understanding of the therapeutic response mechanism. Our overall objective is to 1) develop a predictive statistical model for anti-GD2 response using genomic and transcriptomic biomarkers and 2) experimentally characterize the mechanism underlying this model. The central hypothesis is that genomic changes drive tumor cell subpopulations with variable immune infiltration and mixed anti-GD2 responses. The rationale for this project is that identifying predictive biomarkers for anti-GD2 response in Neuroblastoma will improve patient treatment stratification and help identify strategies for increasing the effectiveness of anti-GD2 therapy. The central hypothesis will be tested by pursuing 3 specific aims: 1) Define the role of genomic changes in Neuroblastoma tumor subpopulations; 2) Characterize the role of tumor subpopulations in immune modulation and anti-GD2 response; and 3) Generate a predictive multivariate model for anti-GD2 response in Neuroblastoma. To assist with these aims, an institutional single cell expression dataset will be prepared for 20 Neuroblastoma patients. Diagnostic samples from anti-GD2 responders and non-responders will be sequenced. Under the first aim, cellular-level genomic changes will be quantified in Neuroblastoma subpopulations using publicly available and institutional single cell expression data. The second aim has 2 parts. For part one, spatial transcriptomics will be used to analyze 8 patients (4 responders; 4 non-responders) for well-defined intratumoral tissue states known as sub-tumor microenvironments. For part two, syngeneic mouse models will be used to assess the immunomodulatory role of the immune checkpoint related gene CD44 in Neuroblastoma. Finally, the third aim will develop a multivariate model comprising genomic, transcriptomic, and IHC-based features for anti-GD2 response prediction in HR-NBL. The model will be applicable to bulk sequencing cohorts and validated in two external cohorts. The research proposed in this application is innovative because it identifies novel genomic/transcriptomic biomarkers for anti-GD2 response in Neuroblastoma and seeks to characterize a novel mechanism that explains response. The proposed research is significant because it is expected to improve patient selection for anti-GD2 therapy and provide much needed insight into mechanisms underlying anti-GD2 response in Neuroblastoma. Ultimately, such knowledge has the potential to impro...