Abstract Endometrial carcinoma (EC) is the most common gynecologic cancer diagnosed in developed countries, with incidence on the rise globally. Over the past decade, EC outcomes have worsened in the United States, despite the fact that early stage, well-differentiated cancers are often cured by simple hysterectomy. Work by the Cancer Genome Atlas Consortium (TCGA) and others has stratified EC patients by four genomically-defined subtypes: POLE ultramutated, microsatellite instability high (MSI-H) hypermutated, copy-number low (CNV-L), and copy-number high (CNV-H). These subtypes not only correlate with clinical outcome but also, in specific instances, response to targeted agents such as immune checkpoint inhibitors (ICI). By integrating genomic and proteomic data, our recent study with the Clinical Proteomic Tumor Analysis Consortium (CPTAC) built upon findings from the TCGA, specifically identifying immunologic subtypes which differentiate tumors based on their overall mutation burdens and antigen processing machinery (APM) profiles. Despite this progress, effective treatments for advanced stage, metastatic and/or recurrent disease are often lacking. Thus, a better understanding of how EC subtypes can be efficiently identified and targeted to improve their outcomes is urgently needed. Here, we present the results of a comprehensive proteogenomic analysis of a new prospectively curated set of 138 EC tumors and 20 specimens enriched for normal endometrium. We also evaluated the utility of a novel machine learning algorithm for parsing clinically-relevant genomic features based solely on images from Hematoxylin and Eosin (H&E) stained cross-sections. Our integrative multi-omic analyses examined the diverse dimensions of molecular pathways to better describe diagnostic and therapeutic targets. From our targeted proteomic assay, we describe a model that uses two peptides to accurately predict tumor APM status. Using gene expression and clinical data, we observed an association between lower Myc activity and Metformin usage in our patients with diabetes. Linking genomics, proteomics, and phospho-proteomics, we uncovered an upregulation of AKT-pT308 in tumors with PIK3R1 in-frame indels. Multi-omic clustering revealed a subset of CNV-L tumors enriched for ß-catenin hotspot mutations, which have elevated Wnt signaling due to impaired ß-catenin degradation signals. Lastly, we observe a subset of MSI-H tumors with a gain of chromosome 1q that show lowered immune infiltration. Analysis of this independent cohort, incorporating published EC tumor and cell line cohorts, has not only confirmed findings from our recent exploratory study but also provided new biological insights relevant to potential therapeutic strategies. These findings can be further investigated to guide patient stratification for improved precision treatment of EC. Citation Format: Lizabeth Katsnelson, Yongchao Dou, Marina A. Gritsenko, Matthew L. Anderson, Karin D. Rodland, Bing Zhang, Tao Liu, David Fenyö, the Clinical Proteomic Tumor Analysis Consortium. Proteogenomic insights suggest druggable pathways in endometrial cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 955.