AUTHORS
Stephanie Dara Curtis, PhD, Bo Hu2, Dante DeAscanis3, Kendall Condon4, Shilpaa Mukundan5, Julia Timis6, Dalen Chan7, Marc Casillas8, Michael Angelo9, Shayla Goller-Mann10, Kumiko Nagashima11, Jonathan Olsen12, Romain Georges, PhD13, Celia Fontanillo, PhD14, Scott Wood15, Zhongying Mo16, Antonia Lopez-Girona, PhD17
AFFILIATIONS
1 Bristol Myers Squibb, San Diego, California, United States
2 Bristol Myers Squibb, San Diego, California, United States
3 Bristol Myers Squibb, San Diego, California, United States
4 Bristol Myers Squibb, Lawrenceville, New Jersey, United States
5 Bristol Myers Squibb, Lawrenceville, New Jersey, United States
6 Bristol Myers Squibb, San Diego, California, United States
7 Bristol Myers Squibb, San Diego, California, United States
8 Bristol Myers Squibb, San Diego, California, United States
9 Bristol Myers Squibb, San Diego, California, United States
10 Bristol Myers Squibb, San Diego, California, United States
11 Bristol Myers Squibb, San Diego, California, United States
12 Bristol Myers Squibb, Lawrenceville, New Jersey, United States
13 Bristol Myers Squibb, San Diego, California, United States
14 Bristol Myers Squibb, San Diego, California, United States
15 Bristol Myers Squibb, San Diego, California, United States
16 Bristol Myers Squibb, San Diego, California, United States
17 Bristol Myers Squibb, San Diego, California, United States
ABSTRACT
Background:
Therapeutic development in hematologic malignancies remains challenging due to heterogeneity of the disease biology, which complicates target validation, biomarker discovery, and clinical prioritization. BMS Heme PRISM (HP; Profiling Relative Inhibition Simultaneously in Mixtures) is a scalable, high-throughput functional screening platform comprising >130 barcoded hematologic cancer cell lines across acute myeloid leukemia (AML), multiple myeloma (MM), diffuse large B-cell lymphoma (DLBCL), and other malignancies. Each model expresses Cas9 and a unique 26-bp DNA barcode, enabling multiplexed quantification of treatment response by next-generation sequencing in pooled culture assays. By integrating response profiling across genetically diverse models, HP identifies molecular determinants of sensitivity and resistance and nominates clinically relevant patient subsets for therapeutic development.
Results:
Benchmarking of standard-of-care agents and internal discovery programs generated functional response maps that revealed therapeutic vulnerabilities and differentiation opportunities across disease contexts. To evaluate clinically relevant combination strategies, HP was adapted to model frontline azacitidine (Aza) plus venetoclax (Ven) therapy in AML. Following Aza priming and anchored Aza/Ven co-treatment, the pan-RAS(ON) inhibitor daraxonrasib was profiled across the 130-CL panel. Of the 39 included AML models, HP capture of on-target efficacy was evidenced by substantial dropout of 9 constituent RAS-mutant CLs to <50% viability. HP identified RAS-mutant, MCL1-dependent, venetoclax-resistant AML as the subset most likely to benefit from this daraxonrasib combination, supporting a biomarker-guided strategy to overcome Ven resistance and inform patient-enrichment approaches.
HP was also extended to immune-engaging modalities through evaluation of T-cell engagers (TCEs). Barcoded B-cell malignancy models spanning a broad range of CD19 antigen expression were co-cultured with CD3+ cytotoxic T cells and treated with a bispecific CD19 TCE. Barcoded K562 cells with stable overexpression of CD19 were also included in the cell line pool to serve as positive controls of CD19-dependent viability defects; the K562-CD19 exhibited strong sensitivity (AUC <30%) while the parental line with low CD19 expression showed a limited effect on viability (AUC >80%) in response to TCE treatment. Similarly, cell line depletion across the rest of the HP CL pool was also tightly correlated to CD19 expression levels. The HP ability to capture specific TCE efficacy in this assay supports its utility for identifying antigen-dependent response patterns and guiding rational development of next-generation immune-engaging therapies.
Conclusions:
BMS Heme PRISM provides a scalable translational platform to connect preclinical functional profiling with clinically actionable development strategies in hematologic malignancies. Spanning small-molecule, combination, and immune-engaging modalities, HP enables identification of response and resistance biomarkers, mechanistic insights, and patient subsets most likely to benefit from novel therapies. These capabilities support more informed clinical prioritization, patient-selection strategies, and rational combination development, positioning HP as a powerful platform to accelerate precision medicine for patients with hematologic cancers.
Keywords: Hematologic Cancers; Multiplexed Cell Line Viability Screening; Drug Discovery; Combination Therapy; T-cell Engagers
DISCLOSURES
Stephanie Dara Curtis, PhD: Bristol Myers Squibb - Current equity holder in publicly-traded company; Bristol Myers Squibb - Current Employment