TMOD-34. Results from a clinical feasibility study investigating an <i>ex vivo</i> drug screening platform that tests living, passage-zero brain tumor tissue

Abstract Neuro-oncology needs a fundamental shift in how cancer treatments are developed and tailored to individual patients. Despite widespread clinical use of genomic/molecular markers to guide treatment, these approaches have widely failed to predict a patient’s response to treatment. To improve personalized patient care, we have developed a functional precision medicine platform where passage-zero patient brain tumor tissues are directly treated with therapeutics ex vivo. Our Screening Live Cancer Explants (SLiCE) technology engrafts heterogeneous tumors atop living rat-derived organotypic brain slice cultures (OBSCs) and recapitulates clinically relevant tumor characteristics that include genomic fidelity, growth, invasion, and response to treatment. This assay includes built-in quality control metrics to ensure assay reproducibility and is amenable to tumor cryopreservation in multiple representative aliquots, enabling iterative testing of a single patient’s tumor to optimize combination therapies. Through our actively accruing clinical feasibility study (NCT05978557), we have successfully engrafted and tested 33-of-33 brain tumors and discovered that this approach preserves key elements of the tumor microenvironment, generates drug response readouts that reflect dynamic tumor–brain interactions, and avoids the physiological distortion seen in other models, enhancing fidelity. The SLiCE assay takes only four days to complete, and in this study, drug sensitivity profiles for every tumor were completed within 28 days after surgery, satisfying the study’s Primary Objective. Across the cohort, we generated 138 normalized Drug Sensitivity Scores (DSSs) from 337 drug–patient combinations, forming a reference library for future benchmarking. Analysis revealed that metastatic tumors exhibited the lowest average DSSs, followed by high-grade and low-grade tumors, suggesting a trend in drug sensitivity by tumor aggressiveness. Among glioblastoma patients treated with temozolomide, DSSs predicted clinical response with 95% accuracy (AUC=0.875, p=0.0175) and correlated with overall survival (R²=0.73). These results support SLiCE as a promising functional precision medicine platform to guide both preclinical development and clinical care.