Abstract / Summary
Abstract Background Pleural mesothelioma (PM) is a rare and aggressive cancer primarily caused by asbestos exposure. Although dual immune checkpoint inhibition (anti-PD-1/anti-CTLA-4) is now approved as first-line therapy, clinical benefit is limited to a small subset of patients, necessitating the need for alternative strategies. Oncolytic viruses represent a promising therapeutic approach as they selectively infect and lyse tumor cells and display immunogenic properties. In PM, we showed that the oncolytic activity of measles virus (MV) is primarily dependent on alterations in the type I interferon (IFN-I) pathway. However, to date no in vitro models have yet been developed to mimic a sophisticated tumor microenvironment (TME), and accurately evaluate MV efficacy in PM. Methods We designed a humanized 3D “vascularized mesothelioma-on-chip” (VMOC) model using microfluidic chips that mimic a TME. VMOC are generated using primary endothelial cells, lung fibroblasts and patient-derived PM cells. We characterized the functionality of our model by confocal imaging, perfusion assay and single-cell RNA sequencing. The impact of MV on the VMOC was assessed by bulk transcriptomics, multiplex ELISA and deconvolution strategies, and compared to clinical data in patients with PM. Results We first proved VMOC functionality with an appropriate endothelial barrier integrity, perfusability, and vascular responsiveness to external stimuli. We demonstrated significant heterogeneity in the VMOC, with the presence of different mesenchymal, cancer and endothelial subpopulations. Interestingly, the addition of PM cells to the VMOC reduced both the IFN-I pathway in cells from the TME, and the abundance of a novel endothelial cell subset characterized by an IFN-I signaling signature. Following MV administration via the endothelial network, we detected PM cell infection and death, as well as strong activation of the IFN-I pathway with the production of multiple inflammatory mediators. Interestingly, the proteomic and transcriptomic inflammatory response induced by MV in VMOCs was similar to that induced in PM patients treated with an oncolytic virus. Conclusions The VMOC model we developed enables in vitro investigation of MV infection and TME reprogramming, paving the way for a deeper understanding of the TME role in OV responses and the development of novel strategies to improve clinical efficacy.