Abstract Metaplastic Breast Carcinomas (MBC) are rare and aggressive breast cancers with mesenchymal or squamous components. While histologically heterogeneous, MBC are typically triple negative: they lack expression of ER, PR and HER2. Importantly, approximately 40% of MBC harbor oncogenic gain-of-function mutations in PIK3CA, the gene encoding for the catalytic subunit of PI3K. These alterations are targetable by an array of inhibitors of the PI3K-AKT-mTORC1 pathway. However, few animal models of MBC are available, and it is currently difficult to study MBC biology and response to therapy. Here, we describe the generation of a new mouse model of MBC driven by the PIK3CA H1047R mutation. We histologically and biologically characterize this model and assess potential therapeutic strategies. We used genetically-engineered mouse models to express the human PIK3CA H1047R mutation under the control of the MMTV promoter, which is mostly expressed in the mammary epithelium. Tumors developed in mice in about 1 year. We histologically characterized these tumors and generated cell lines, organoids, and tumor grafts in immunocompromised mice. The resulting model, termed MMTV-102, was evaluated for the response to PI3K-AKT-mTOR pathway inhibitors, both in vitro and in vivo. The MMTV-PIK3CAH1047R tumor was a high grade poorly differentiated MBC negative for ER, PR, and HER2. It presented squamous components, keratin depositions, mitotic figures, and vascular invasion and stained positive for the basal marker CK14. In addition, due to the presence of the PIK3CA H1047R mutation, it stained positive for phosphorylated AKT, PRAS40, and S6RP, indicating constitutive activation of the PI3K-AKT-mTORC1 pathway. MBC derived cell lines and organoids were confirmed to be triple negative and could form tumors in mice. In addition, these cellular models were sensitive to PI3Kα, pan-PI3K, and mTORC1 inhibitors in vitro and in vivo, but the mTORC1 inhibitor everolimus exhibited the strongest effect. When resistance to everolimus eventually developed in vivo, we established cell lines from these resistant tumors. In addition, we generated an everolimus resistant cell line following chronic drug exposure in vitro (MMTV-102 Par res). Both cellular models were tested for their sensitivity to the next-generation mTORC1 inhibitor RMC-5552 by proliferation and clonogenic assays. We show that our mouse model can originate MBC with an active PI3K pathway. These tumors are amenable to transplantation into immunocompromised mice and to generate cell lines and organoids, allowing us to develop a simple, versatile, and rapid model of MBC. Tumors derived from our model are sensitive to PI3Kα and mTORC1 inhibitors, but eventually develop resistance to these agents. The use of next-generation mTORC1 inhibitors can overcome resistance in some cases, providing a rationale for their implementation in the treatment of patients with MBC. Citation Format: Paola Roa, Valentina Foglizzo, Michelle R. de Marchena, Shraddha Chandthakuri, Emiliano Cocco, Pau Castel. A PIK3CA-driven mouse model of metaplastic breast carcinoma to assess potential therapeutic strategies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1452.