This study investigates the spatial distribution and plasticity of Fibroblast Activation Protein (FAP)+ Cancer-Associated Fibroblasts (CAF) in breast cancer (BC). By analyzing single-cell transcriptomics data and performing functional assays, the authors identify 10 spatially organized cellular niches, called EcoCellTypes (ECT), which are differentially localized within tumors. These ECTs are composed of specific FAP+ CAF clusters and immune populations, and they play a crucial role in tumor progression and immune evasion. The Detox-iCAF cluster, which is associated with detoxification pathways, transitions into Wound-myCAF and ECM-myCAF clusters through DPP4- and YAP-dependent mechanisms, driven by cancer cells. ECM-myCAF and TGFβ-myCAF clusters polarize TREM2+ macrophages and regulatory NK and T cells, inducing an immunosuppressive environment. The study also reveals that FAP+ CAF subpopulations accumulate differently according to the invasive status of BC and predict invasive recurrence of ductal carcinoma in situ (DCIS), providing a potential biomarker for identifying low-risk DCIS patients eligible for therapeutic de-escalation.This study investigates the spatial distribution and plasticity of Fibroblast Activation Protein (FAP)+ Cancer-Associated Fibroblasts (CAF) in breast cancer (BC). By analyzing single-cell transcriptomics data and performing functional assays, the authors identify 10 spatially organized cellular niches, called EcoCellTypes (ECT), which are differentially localized within tumors. These ECTs are composed of specific FAP+ CAF clusters and immune populations, and they play a crucial role in tumor progression and immune evasion. The Detox-iCAF cluster, which is associated with detoxification pathways, transitions into Wound-myCAF and ECM-myCAF clusters through DPP4- and YAP-dependent mechanisms, driven by cancer cells. ECM-myCAF and TGFβ-myCAF clusters polarize TREM2+ macrophages and regulatory NK and T cells, inducing an immunosuppressive environment. The study also reveals that FAP+ CAF subpopulations accumulate differently according to the invasive status of BC and predict invasive recurrence of ductal carcinoma in situ (DCIS), providing a potential biomarker for identifying low-risk DCIS patients eligible for therapeutic de-escalation.