利用携带多种突变的患者来源肿瘤类器官进行药物筛查，以识别针对转移性癌症患者的有效治疗药物

Patient-derived primary tumors (PDPTs) are emerging as physiologically relevant and translational in vitro models for predicting human responses to cancer therapeutics. Compared to conventional 2D cell lines, PDPTs more accurately recapitulate the tumor microenvironment, enable high-throughput drug screening, and support precision medicine. However, widespread adoption has been limited by challenges in large-scale expansion without phenotypic changes. We developed a novel system for in vitro expansion and biobanking of colorectal cancer tumoroids derived from four metastatic donors with distinct mutations. Our approach integrates matrix-coated plates with microwell technology to scale up tumoroid growth. Two assay platforms were established: (a) Tumoroids generated by seeding 5000 single tumor cells as a stand-alone model, and (b) Complex 3D tissues created by co-seeding 500 tumor cells with fibroblasts, endothelial cells, and immune cells to mimic the tumor microenvironment (TME). Tumoroids were treated with five chemotherapeutic agents-Cisplatin, Doxorubicin, Oxaliplatin, Fluorouracil, and Cetuximab-at six concentrations over a 7-day period (three doses). For tumoroids grown in TME, microscopic and histological analyses confirmed uniform-sized tumoroids and glandular-like structures. Epithelial origin and fibroblast presence were validated by CK19+ and vimentin staining, respectively. Live/dead staining using calcein AM and PI revealed dose- and time-dependent drug responses in both platforms. Notably, Doxorubicin, previously unused in these patients, demonstrated significant activity-highlighting the assay's potential to identify drugs for precision therapy and enable patient-specific drug ranking. In summary, scalable PDPT expansion combined with physiologically relevant assay systems offers a cost-effective, predictive, and non-animal approach for metastatic cancer drug screening, advancing the concept of precision medicine.