Evaluation of immune-checkpoint inhibitor-induced pneumotoxicity using a patient-derived alveolar organoid and PBMC co-culture model

Immune-checkpoint inhibitors (ICIs), such as Pembrolizumab (anti-PD-1) and Atezolizumab (anti-PD-L1), have revolutionized oncology. However, their use is frequently limited by immune-related adverse events (irAEs), with checkpoint inhibitor-pneumonitis (CIP) being one of the most clinically significant and potentially fatal toxicities. Current preclinical models, including traditional 2D cell cultures and animal models, often fail to accurately recapitulate the complex human alveolar microenvironment and immune responses. In this study, we established a human-derived alveolar organoid and peripheral blood mononuclear cell (PBMC) co-culture system to evaluate ICI-induced immunotoxicity in vitro. • Sample Collection: Alveolar tissues were obtained from 24 healthy donors, and PBMCs were isolated from 20 healthy volunteers. • Organoid Culture: Human alveolar epithelial cells were isolated and cultured in a 3D extracellular matrix to form mature alveolar organoids. • Co-culture System: To simulate the lung immune environment, alveolar organoids were co-cultured with autologous or allogeneic PBMCs.• ICI Treatment: The co-culture models were treated with clinical-grade Pembrolizumab and Atezolizumab at varying concentrations. • Toxicity Assessment: * Cell viability and structural integrity of the organoids were visualized using Calcein AM staining. • To further characterize the inflammatory response, cytokine profiling via ELISA/Luminex is currently being optimized to quantify the localized "cytokine storm" effect.The established alveolar organoids successfully maintained the expression of alveolar type II (AT2) markers (e.g., SFTPC). Upon co-culture with PBMCs and treatment with Pembrolizumab or Atezolizumab, a dose-dependent increase in organoid damage was observed. Calcein AM staining revealed significant morphological changes and a decrease in green fluorescence intensity in the ICI-treated groups compared to the control group, indicating loss of epithelial viability due to hyper-activated immune cells. While the definitive cytokine secretion data are currently being processed, the preliminary visual evidence of epithelial destruction suggests that the presence of ICIs triggers an aggressive T-cell mediated attack on healthy alveolar cells within this 3D architecture. Our findings demonstrate that the human alveolar organoid-PBMC co-culture model effectively replicates the cellular interactions involved in ICI-induced pneumotoxicity. The observed morphological degradation validated by Calcein AM staining suggests that this platform can serve as a high-fidelity surrogate for predicting irAEs. Once cytokine analysis is integrated, this model will provide a comprehensive tool for screening the safety profiles of next-generation immunotherapies and developing localized management strategies for pneumonitis.