Abstract A001: AscitesPredict: Validating anti-cancer therapeutic efficacy in gastroesophageal cancer using rapid ex vivo high-dimensional morphology profiling

Abstract Overcoming cancer drug resistance is a major unmet clinical need as it remains challenging to accurately predict therapeutic sensitivity from baseline tumor and immune profiling. While biopsies from refractory disease are challenging to routinely obtain for research, upwards of 40% of gastroesophageal and 70% ovarian patients develop peritoneal carcinomatosis and ascites, a plentiful, heterogenous and molecularly representative sample format that is easily accessible since routine drains occur. However, in standard clinical practice this material is only subjected to cytology to detect malignant cells and then discarded. Bulk profiling technologies are challenging to deploy due to the highly variable single cell heterogeneity of ascites. To address these challenges, we have developed the first generation of AscitesPredict, an ‘ex vivo tumor biosensor’ technology using single cell technologies to measure cell identities and therapeutic sensitivities during a 5 day period in which viability is preserved via media supplementation with patient-matched ascites fluid. AscitesPredict utilizes high dimensional single cell image-based morphological profiling and machine learning applied to brightfield microscopy. Thus far, we have profiled over 35 GEA ascites samples, made initial assessments of technological reproducibility, and have optimized initial workflows against a diversity panel of 20 therapeutics. Our ultimate goal is to harden the technology to facilitate deployment by a wide diversity of researchers and clinicians for two driving use cases: (1) supporting preclinical drug development and (2) eventually as a functional diagnostic. Citation Format: Jesse S Boehm, Mushriq Al-Jazrawe, Csaba Molnar, Beth Cimini, Samuel Klempner, Niklas Rindtorff, Steven Blum. AscitesPredict: Validating anti-cancer therapeutic efficacy in gastroesophageal cancer using rapid ex vivo high-dimensional morphology profiling [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Functional and Genomic Precision Medicine in Cancer: Different Perspectives, Common Goals; 2025 Mar 11-13; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(5 Suppl):Abstract nr A001.