Abstract 2803: Ex vivo micro-tumor testing platform to characterize and predict patient-specific immunotherapy drug responses

Abstract Background. Immunotherapies have shown significant clinical benefit, but upfront identification of patients with therapy-sensitive cancers remains a major challenge. We have previously demonstrated that our ex vivo micro-tumor platform, which preserves the tumor microenvironment (TME), is predictive for clinical responses to platinum-based therapy (Koedoot et al., npj precision oncology, 2025). Here, we show that the platform also enables patient-specific profiling of drug sensitivity and immune activity for immune checkpoint inhibitors (ICIs) and bispecific T cell engagers (BiTEs). Methods. Fresh tumor tissues were collected from colorectal cancer (CRC, N=6), non-small cell lung cancer (NSCLC, N=23) and ovarian cancer (OC, N=85) patients with primary or recurrent disease. Micro-tumors including the TME were isolated from resections, ascites or pleural fluid, embedded in hydrogel and exposed to chemotherapeutics, six ICIs and/or CD3xMUC16 BiTE ubamatamab. Drug responses were quantified by extracting morphological features from high-throughput 3D imaging. Treatment sensitivity was defined by statistically significant immune cell expansion and micro-tumor killing. Sensitivity profiles were correlated with immune cell populations. Supernatants collected after treatment were analyzed for secreted cytokines, chemokines and cytotoxic mediators using multiplex immune biomarker profiling. PD-L1 tumor proportion scoring (TPS) reported on original patient tumor tissue was used (NSCLC). Results. Significant ex vivo ICI responses were observed in 50% of OC, 27% of early NSCLC, and 33% of metastatic NSCLC samples, with reproducible results across serial samples. Most OC samples were positive for T-cell, macrophage and NK-cell markers but not for B-cell markers. Stronger immune responses were observed in OC samples with high CD3 (p=0.01) or CD68 (p=0.003) expression. Importantly, immune sensitivity correlated with PD-L1 TPS (p=0.13, NSCLC). Ubamatamab induced significant immune expansion in 42% (8/19) of OC samples, half of which also showed significant tumor killing. When compared to non-responders, ubamatamab responders with effective tumor killing exhibited robust and simultaneous upregulation of T cell effector cytokines (IFN-γ, IL-2), cytotoxic mediators (Granzyme B, Perforin) and pro-inflammatory cytokines (CXCL9, CXCL11). Conclusions. This study reports classification, characterization and prediction of autologous patient-specific sensitivity to ICIs and BiTEs by characterizing immune populations and reporting correlation with known clinical markers (PD-L1). Our ex vivo micro-tumor platform enables functional profiling of patient-specific immunotherapy responses, supporting biomarker discovery and selection of patients for emerging immunotherapies. The next step is to correlate the ex vivo sensitivity to actual clinical outcome. Citation Format: Felix M. Behr, Esmee Koedoot, Timothy J. Sijsenaar, Léa Le Large, Cor D. de Kroon, Anne van Altena, Farbod Khoraminia, Laurie C. Steinbusch, Christi M. Steendam, Dieudonné J. van der Meer, Willemijn Vader, Nelleke Ottevanger, Judith R. Kroep. Ex vivo micro-tumor testing platform to characterize and predict patient-specific immunotherapy drug responses 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 2803.