患者匹配的类器官-免疫共培养模型作为精准肿瘤学中免疫治疗应答的预测平台

Background: Immunotherapy (IO) is increasingly integrated into cancer treatment, with checkpoint inhibitors (anti-CTLA-4, anti-PD-1/PD-L1) improving progression-free and overall survival in subsets of patients. However, many across cancer types do not respond, highlighting the need for predictive preclinical models to test and tailor IO-based therapies. Traditional in vivo models and 2D cultures fail to replicate the complexity of the human tumor-immune microenvironment. To bridge this gap, we refined a patient-derived organoid (PDO) and immune cell co-culture platform for rapid drug screening of IO combinations, aligning with the NIH directive to reduce animal use. PDOs can be tested within 10-15 days, providing a personalized system to evaluate treatment response and tumor-immune interactions across solid tumors. Methods: We established a rapid (<7-day) PDO-peripheral blood mononuclear cell (PBMC) co-culture integrating short-term organoid generation, immune activation, and checkpoint blockade assays. PBMCs were isolated, cryopreserved, and reactivated prior to co-culture using optimized cytokine cocktails (IL-2, IL-7, IL-15, IL-21) and tested in multiple media (AIM-V, X-VIVO, HPLM, RPMI). Cell viability was measured using CellTiter-Glo 3D and apoptosis via Caspase-Glo 3/7 assays. Activated PBMCs were co-cultured with PDOs at 5:1 and 10:1 effector-to-target ratios under IO or combination chemotherapy. Results: We developed a rapid workflow to generate immune-competent PDOs from breast, bladder, and lung cancers co-cultured with autologous PBMCs. Mono- and combination therapies were tested across tumor types, including various IO agents, platins (cisplatin, carboplatin), taxanes, antibody-drug conjugates, and targeted therapies. Multiple functional parameters were measured, and an apoptotic effectiveness index (AEI) is being developed to rank treatment responses. Across all tumor types, IO combined with platinum or taxane agents produced the highest AEI, reflecting strong synergy in standard-of-care regimens and validating the model's predictive potential. AEI-based ranking identified differential apoptotic responses among tumor types, supporting the model's utility in predicting patient-specific responses. Conclusions: We established and validated a platform that rapidly assesses whether a patient's immune cells can be activated to kill their tumor organoids in the presence of specific immunotherapy or chemotherapy combinations. We anticipate identifying key biomarkers, including immune phenotypes and cytokine signatures, that distinguish immunotherapy-responsive from non-responsive PDO co-cultures across breast, bladder, and lung cancers. These findings lay the groundwork for predictive tools and future clinical correlations. AI disclosure: AI was used only for language editing; content was verified by the authors.