ImSAVAR’s overall approach involves iterative cycles based upon a set of workflows that move from case studies to models developed with samples from healthy tissues, or non-disease in vivo models, to models that include samples from disease and in vivo disease models. During the fifth year of the project, the consortium has accomplished significant progress:
WP1 focused on project management and innovation, successfully finalizing the IMI Annual Activity Report and gaining approval for the fourth periodic report. The Project Management Office (PMO) organized a study-a-thon and hosted a hybrid consortium meeting in November 2024. Regular meetings fostered innovation management, emphasizing patient sample usage, model validation, and community building around immune-related adverse outcome pathways (irAOPs). Stakeholder engagement led to drafting a white paper on irAOPs and enhanced dissemination efforts, culminating in increased social media engagement and a special issue of the Journal of Immunotoxicology.
WP2 achieved the harmonization of irAOPs for cytokine release syndrome (CRS), integrating this into the MINERVA platform. The consortium submitted two key manuscripts: one detailing the harmonized CRS irAOP and another on cytokine release assays in immunotherapy. Lund developed a BOEC:PBMC co-culture model to predict CRS risk in patients treated with checkpoint inhibitors. A cross-consortium study aims to correlate donor phenotypes with CRS susceptibility using in vitro cytokine release assays. The breast-cancer-on-chip model was refined to enhance CAR-T efficacy and safety testing, while upgraded vessel-on-chip models provided insights into inflammatory cell recruitment, vascular leakage and coagulopathies. Biomarkers linked to cytopenias were identified in CAR-T patients, with ongoing validation supported by extensive patient data collection. CPI studies indicated no hepatotoxicity in co-culture models, prompting further research into long-term effects.
WP3 refined the framework for investigating side effects of high- and low-dose IL-2 treatments. Regular meetings facilitated effective knowledge transfer and data discussions. Results from in vitro, ex vivo, in vivo, and microphysiological models were published in a special issue of the Journal of Immunotoxicology. The harmonization of irAOPs for IL-2-mediated skin rash, hepatotoxicity, and vascular leakage was completed, with efforts to transfer these findings to other drugs for agnostic irAOP development. Refined models, including co-cultures of tissue-resident and immune cells, generated wet-lab data on IL-2 side effects. Ongoing model optimization and biobank establishment for relevant samples are in progress, although sample availability for high-throughput immune safety assessments remains limited.
WP4 established a framework for integrating biomarker development into the irAOPs concept, essential for nonclinical models in immunomodulatory therapies. Existing OECD guidelines regarding biomarkers were found lacking, prompting ongoing efforts in this area. Key developments included the publication of an irAOP for CAR-T cells and a model for CRS. Collaborative work with EFPIA partners focused on creating standardized protocols for biomarker assessment. Progress on the CAR-T cell atlas advanced nonclinical therapy development. A cross-consortium study was initiated to ensure the safe transition of immunomodulatory drugs to human trials. The MINERVA platform was effectively utilized for mapping irAOPs, enhancing the understanding of immune-related adverse events. Significant strides were made in identifying biomarkers linked to CRS through patient data analysis, with ongoing refinement of guidelines for biomarker development.
WP5 refined the Data Catalogue and the DAISY system for documenting data access activities. Workshops, both online and in-person, clarified questions about Data Information Sheets (DISHs) and data uploads to UNILU. The contract with the hepatocyte biobank in Leipzig was extended to include hepatic immune cells for WP3. Additionally, well-characterized lung specimens for precision cut lung slices (PCLS) were provided, yielding samples from 13 disease-free and 17 end-stage fibrotic lungs, although the availability of suitable end-stage samples slightly decreased. A comprehensive Catalogue of Assays and Models was delivered, consolidating experimental tools for all stakeholders involved.
