Expanding the PRISM platform: Addition of new hematologic models and a 10-day assay improve drug sensitivity mapping for epigenetic targets

Despite major advances in targeted therapies, cancer remains a leading cause of death among pediatric, adolescent, and adult populations, emphasizing the need for continued therapeutic innovation. The PRISM assay enables large-scale evaluation of oncology agents by barcoding and pooling over 900 cancer cell lines with extensive genomic and functional characterization. This platform enables systematic analysis of drug potency, selectivity, and rapid identification of genomic correlates of sensitivity. To broaden the coverage of pediatric, hematologic, and rare tumor types in the PRISM assay and enhance our ability to investigate novel therapeutics for these disease areas, we expanded the PRISM cell line panel in collaboration with the Pediatric Cancer Dependencies Accelerator to include 109 additional lines, including 21 pediatric, 60 hematopoietic, and 14 previously unrepresented subtypes. We re-profiled compounds with predicted activity in new cell models and found that many compounds targeting heme-centric proteins such as BCL6, FLT3, IKZF3, EP300, EZH2, EED had stronger correlations with target dependency. Targeting of epigenetic drivers is an attractive therapeutic strategy for hematopoietic tumors, so we sought to compare clinically relevant epigenetic targeting compounds in PRISM. For example, mevrometostat more selectively targets EZH2-dependent cell lines compared to other EZH2 inhibitors profiled after 5 days of treatment. However, the observed potency of many epigenetic targeting compounds is somewhat limited in a 5-day assay, as the mechanistic nature of these compounds often requires a longer incubation period to enact their cytotoxic effects. We developed a 10-day PRISM assay for our hematopoietic cell collection and re-screened mevrometostat as part of an assay validation screen. We observed greater potency, selectivity, and stronger correlations with EZH2 dependency for mevrometostat in the 10-day assay compared to the 5-day assay. Future assay development will enable higher throughput screening of slow-acting compounds against our hematopoietic cell lines. Taken all together, these advances position PRISM to accelerate hematologic and pediatric cancer research by expanding disease-relevant models and enabling long-duration single-agent screening at scale. All data will be made publicly available on depmap.org, providing a resource for the cancer research community to explore novel therapeutic opportunities.