Abstract 7262: Organoid-based prediction of immune checkpoint inhibitor response in gastric cancer: A correlation with clinical outcomes

Abstract Research and development in immune checkpoint inhibitor (ICI) therapy are rapidly advancing across various cancer types. However, accurately predicting treatment responses remains a significant challenge due to factors such as patient-specific immune system variability, tumor heterogeneity, and genetic differences. Current model systems for evaluating ICI efficacy also face limitations. While organoids replicate patient-specific tumor characteristics, their lack of immune components hinders their ability to assess the complexity of immunotherapy. To address these challenges, we developed a co-culture platform combining tumor organoids with immune cells to evaluate ICI drug efficacy and predict patient-specific therapeutic responses. We collected ascites and pleural effusion samples from 88 gastric cancer patients, obtaining 109 samples and successfully establishing 63 organoids, achieving an establishment rate of 80.7%. Immunohistochemical staining and next-generation sequencing (NGS) confirmed the organoids’ molecular and histological similarity to the corresponding patient tumors. Peripheral blood mononuclear cells (PBMCs) were isolated from matched blood samples and activated using cytokines such as IL-2. Patient-derived immune cells were then co-cultured with tumor organoids and treated with ICI drugs, including nivolumab, pembrolizumab, and atezolizumab, as well as chemotherapy drugs, to evaluate their anticancer efficacy. Our results showed immune cells localizing around the organoids and inducing concentration-dependent tumor cell death following ICI treatment. Drug response evaluations were conducted on 25 different organoids, and the findings were compared with the clinical responses of the corresponding patients. Among 17 patients who underwent follow-up after anticancer treatments, including immunotherapy, the co-culture platform demonstrated high predictive performance, with an accuracy of 94.12%, sensitivity of 83.33%, and specificity of 100%. Notably, sequential analysis of samples from the same patients revealed that in vitro predictions of resistance to nivolumab corresponded with clinical progression as assessed by the Response Evaluation Criteria in Solid Tumors (RECIST). These findings highlight the robustness of our co-culture platform in predicting ICI responses and underscore its potential to guide personalized immunotherapy strategies for gastric cancer patients. Related patents and rights belong to Podo Therapeutics. Citation Format: Hyoyoung Kim, Shin Young Chang, Hye Ryeong Jun, Un-Jung Yun, Jung Eun Kim, Hyunki Kim, Tae Il Kim, Minkyu Jung. Organoid-based prediction of immune checkpoint inhibitor response in gastric cancer: A correlation with clinical outcomes [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 7262.