Combination chemotherapy is central to acute myeloid leukemia (AML) treatment, yet no functional assay is available in clinical practice to guide regimen selection for individual patients. We developed and optimized an ex vivo drug combination assay to quantify synergy between clinically relevant AML drug pairs using primary patient samples. The workflow was validated in HL-60 cells, where synergy metrics—Combination Index, Combination Sensitivity Score (CSS), and the Quadratic Phenotypic Optimization Platform (QPOP)—revealed drug- and concentration-dependent variability, with cytarabine + mitoxantrone exhibiting the strongest synergy. Applying the protocol to bone marrow aspirates from 15 AML patients, we observed substantial inter-patient heterogeneity and intra-patient variability in synergy across five drug pairs. Cytarabine-based combinations achieved the highest synergy in 60% of patients, while venetoclax-based combinations ranked highest in 40%. Synergy rankings differed between CSS and QPOP, underscoring the value of multi-metric assessment. This approach demonstrated reproducible performance and identified patient-specific top-ranked regimens, including cases where synergy was comparable between mechanistically distinct combinations. Our findings establish a methodological framework for integrating multi-metric synergy profiling into functional precision oncology pipelines, with potential to improve rational selection of AML combination therapies.
