RP1: Molecular Determinants of GIST Pathogenesis Abstract Gastrointestinal stromal tumor (GIST) is one of the most prevalent sarcoma subtypes and mesenchymal neoplasms of the GI tract. Clinically, advanced unresectable and/or metastatic GISTs are rarely curable despite their initial responses to tyrosine kinase inhibitors. Current risk models (e.g., Miettinen, NIH modified), although useful to predict patients at high risk of recurrence following surgery, often underestimate recurrence risk in cases of intermediate size or mitotic count and non-gastric location, and no current risk stratification schemes consider molecular biomarkers. Discovering, validating, and incorporating molecular determinants of aggressive and metastatic clinical behavior in GIST clinical care are critical knowledge gaps. Our recent pan-cancer genomic characterization of metastatic vs. primary tumors and GIST-specific genomic and clinical analysis of a MSK discovery cohort of >400 untreated patients revealed that biallelic loss of function mutations in MAX or MGA (each 5%), and MYC amplifications (0.5%), were enriched in untreated metastatic GIST compared to primary localized GIST. Further, MAX/MGA/MYC genetic alterations and arm-level deletion of 1p were associated with worse relapse free survival (p<0.001). MAX is the obligative heterodimerization transcriptional activator of MYC and modulates MYC regulation of cellular function. MAX also interacts with MGA, a core component of the PRC1.6 complex that regulates cellular differentiation. We hypothesize that genetic perturbations of MAX/MGA/MYC and arm-level copy number alterations (e.g., -1p) in GIST promote cellular plasticity, tumorigenesis, and aggressive metastatic behavior through functional inactivation of the PRC1.6 complex and may represent independent molecular risk biomarkers for GIST relapse and facilitate discovery of rational pathway targets for metastatic GIST. Here, we propose preclinical and clinical investigations, leveraging sophisticated and novel multi-omics (e.g., transcriptome, cistrome, epigenome, single cell (sc) and/or single nuclei (sn) RNA-seq, scATAC-seq coupled with a CellTagging system for clonal tracking) and machine learning approaches to 1) elucidate the molecular mechanisms of MAX/MGA/MYC perturbations involved in enhanced cellular plasticity, migration, and metastatic behavior in GIST, and 2) identify and validate molecular biomarkers predictive of recurrence risk and adjuvant therapy determination in primary GIST after resection, using discovery and multiple validation GIST patient cohorts to develop a combinatorial pathologic-genomic nomogram to guide adjuvant imatinib therapy decisions. These studies could change clinical practice in GIST adjuvant treatment and identify novel therapeutic strategies to target metastases in GIST management.