Abstract 4875: Modeling GBM-vasculature interaction on chip

Abstract Background: The tumor microenvironment (TME) of glioblastoma (GBM) is highly complex and dynamic, playing a pivotal role in tumor progression and therapeutic resistance. Comprising diverse components including blood vessels, immune cells, and the blood-brain barrier, the heterogeneous TME forms distinct niches that critically shape tumor behavior. Recent TME-based classification has delineated three subtypes TMELow, TMEMed, and TMEHigh distinguished by varying vascular and immune cell compositions. Methods: This study focuses on elucidating the vascular interactions within the GBM-TME interface. To achieve this, we developed a brain-specific 3D vascular network by integrating primary human brain-derived endothelial and stromal cells within an extracellular matrix (ECM) on a microfluidic platform. Following the establishment of the brain vascular network model, GBM cell lines were incorporated to examine tumor-vasculature interactions in a 3D context. GBM integration was performed via two distinct strategies: co-seeding with vascular-forming cells within the ECM at the start of culture, or grafting tumor cells onto a pre-established vascular bed culture to facilitate invasion assays. Results: The resulting perfusable model recapitulates key features of the brain vasculature, including the expression of adherens junction markers CD31 and VE-cadherin, tight junction marker Claudin-5, and the strategic alignment of stromal cells along vascular structures. Preliminary results from the integration of GBM in the vascular network model indicate that, irrespective of the seeding method, GBM cells preferentially localize along vascular structures, suggesting an intrinsic tropism toward the vasculature. Notably, co-cultures of GBM with vascular networks exhibited significantly higher proliferative and invasive capacity compared to GBM monocultures, highlighting the pro-tumorigenic influence of the vascular microenvironment. Conclusion: The brain vascular network model demonstrates physiological relevance, robustness, and scalability making it suitable for medium- to high-throughput applications. This multifaceted platform offers a unique opportunity to functionally model and stratify GBM TME subtypes in vitro, with the potential to uncover novel therapeutic targets within the vascular and immune niches of the tumor ecosystem. Citation Format: Promise Emeh, Alan Fetah, Jade Admiraal, Marleen Bokkers, Kevin Jimenez-Cowell, Will Allen, Nienke Wevers, Martine Lamfers, Clemens M. Dirven, Todd Burton, Karla Queiroz. Modeling GBM-vasculature interaction on chip abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 4875.