Prostate cancer is the second most common malignancy worldwide and the fifth leading cause of male cancer-related mortality. Approximately 15% of localized cases are classified as high-risk for biochemical recurrence. These patients frequently harbour occult metastases undetected by conventional imaging (CT and bone scintigraphy). Consequently, PSMA PET has revolutionized primary staging and is strongly recommended by EAU guidelines. In the Czech Republic, a transition toward a "PSMA-first" pathway is evident, where molecular imaging increasingly replaces conventional modalities. This "stage-migration" identifies a new cohort of miN1/M1 patients previously classified as having localized disease. However, PSMA PET is a resource-demanding modality and not yet universally available. To optimize its utility, identifying which combinations of routinely available parameters predict metastatic disease and characterizes this new cohort is essential. Furthermore, real-world evidence regarding early oncological outcomes within this PSMA-only framework is limited. This multicentre cohort study aims to define a multiparametric predictive model for PSMA-detected metastases and evaluate the real-world therapeutic trajectories and follow-up outcomes of this newly defined high-risk population. Integrating clinical, biological, and molecular data, seeks to refine patient selection and provide a longitudinal perspective on the "PSMA-first" diagnostic era.