ABSTRACT Malignant phyllodes tumors (MPT) are extremely rare primary breast cancers, which are globally understudied due the infrequency of the diagnosis and to the unfamiliarity with the aggressiveness of their biologic behavior. Because benign phyllodes tumors are much more common and have a very indolent course, MPT are often underappreciated. There are few reliable predictive markers for outcomes and unfortunately local recurrence (LR) and/or distant metastases occurs frequently for MPT (20%). Treatment is almost exclusively surgical; there is no systemic therapy outside of the metastatic setting. With no known effective chemotherapy and no approved targeted therapy options, metastatic progression, which occurs frequently, portends a dismal prognosis. Median survival for these young women (median age of 45) is just 7-15 months. This study will define clinically actionable opportunities for this rare, but frequently fatal, tumor, in a population currently not provided access to potential life-saving therapies due to severely limited data. This study specifically aims to define potentially targetable and actionable opportunities and to improve the ability to predict outcomes (both phenotypically and genotypically). The results may be immediately impactful by allowing known FDA-approved therapies to be offered to those known to be at highest risk. Our first aim is to define the repertoire of genomic alterations in borderline PT (BLPT) and MPT and to evaluate these for associations with clinical outcomes. This study will be conducted as an archival project (utilizing stored tissue blocks) and will include 100 MPT and 50 BLPT cases, with at least 25 cases having known LR and/or metastatic disease. Our second aim is to assess the reclassification and outcomes of phyllodes tumors applying the newly released College of American Pathologists (CAP) Phyllodes Cancer Protocol Template and to develop a predictive model and nomogram. Utilizing all available cases of BLPT and MPTs previously included in our group's US multi-center phyllodes tumor series (N=550), we will perform a centralized histopathologic re-review, based on digitized whole slide imaging and classification according to the new CAP reporting protocol. Without an accurate and reliable predictive model for LR and/or metastatic disease, we remain unable to stratify women for interventions. Our group's ability to perform molecular sequencing on a large number of BLPT and MPT will not only help define who is at risk, but who may immediately be eligible for systemic therapies, not currently available to women at high risk for an unforgiving clinical course.