Non-functioning pituitary adenoma (NFPA) is a common subtype of pituitary tumor and is a major cause of morbidity due to mass effects and tumor recurrence. Although surgical resection is the primary treatment modality, a substantial proportion (~50%) of patients experience tumor regrowth, while others remain stable for many years following surgery. Currently, there are no reliable molecular biomarkers that can accurately predict tumor progression or guide postoperative management. As a result, patients are often subjected to prolonged radiological surveillance placing a burden on both patients and healthcare systems. Biomarkers are needed to identify patients at risk of progression who may benefit from early re-operation, radiotherapy and/or medical treatment. In parallel, understanding the molecular determinants of tumor progression is crucial for the development of novel therapeutic approaches, which is limited in non-functioning pituitary adenomas. The primary hypothesis of this project is that integrative analysis of molecular information across epigenetic, proteomic and transcriptomic layers together with patient’s clinical data will enable the identification of robust predictive biomarkers, define key regulatory mechanisms, and support the development of clinically applicable model for early risk stratification ultimately improving patient management. The specific aims of this project are: Aim I: To identify biomarkers predictive of tumor progression using integrative multi-omics (epigenetic, proteomic and transcriptomic) analysis of tumor tissue. Aim II: To construct multi-omics regulatory networks integrating DNA methylation, transcriptomics, and proteomics data to identify coherent regulatory networks and master regulators of tumor progression. Aim III: To identify circulating RNA biomarkers in plasma that enable non-invasive and early detection of tumor progression. Aim IV: To develop machine learning-based predictive model integrating molecular biomarkers and clinical variables including age, tumor size, invasion, surgical outcome and images from routine MRI to provide individualized risk estimates. Aim V: To validate the identified biomarkers and predictive model to demonstrate reproducibility and clinical utility supporting their clinical translation.