开发一种整合自体成纤维细胞和肿瘤浸润淋巴细胞的高通量多模态患者来源平台用于结直肠癌药物发现

Background: Colorectal cancer (CRC) remains one of the leading causes of cancer-related mortality, driven by extensive molecular heterogeneity and the emergence of therapy-resistant subclones. Conventional in vitro models fail to reproduce the complexity of the tumor microenvironment (TME). To address this, we developed a high-throughput, multimodal patient-derived tumoroid (PDTO) platform incorporating autologous tumor-infiltrating lymphocytes (TILs) and / or cancer-associated fibroblasts (CAFs), enabling integrated pharmacologic, genomic, and immunologic analyses for translational drug discovery. Methods: Matched tumor and adjacent normal tissues were collected from CRC patients under the NHSGGC biorepository ethics approval (REC 22/WS/0020). PDTOs, CAFs, and TILs were isolated, expanded, and cryopreserved under optimised protocols that maintain molecular signature of the parental tumor. Each patient derived model was subjected to multi-omic characterisation, including whole-exome sequencing, and bulk RNA sequencing. Derived models were treated with standard-of-care drugs (SOC) to assess sensitivity/resistance, tumoroid survival and viability post treatment, which was analysed by high throughput fluorescent imaging, flow cytometry and multiplex cytokine profiling. Results: To date, five PDTO, CAF, TIL models, including mono-, dual-, and tri-culture formats have been successfully established from both primary and metastatic CRC samples encompassing diverse mutational backgrounds. Integrated mutational burden analysis together with transcriptional and clinical data revealed discrete transcriptional clusters associated with microsatellite status, immune infiltration, and drug response patterns. SOC regimens demonstrated expected cytotoxic profiles, confirming physiological relevance of the models. Initial experiments indicated that inclusion of immune and stromal component confirm that incorporating CAFs and TILs modifies pharmacologic response profiles. Conclusions: This multimodal CRC tumoroid platform recapitulates the epithelial, stromal, and immune components of the native TME. Its integration of clinical, genomic, and gene expression datasets enables high-content functional screening and predictive biomarker discovery. The platform supports translational drug development while reducing the need for large-scale expansion across heterogeneous patient cohorts by capturing most dominant CRC molecular subtypes and intratumoral complexity within each culture system.