Designing patient-oriented combination therapies for acute myeloid leukemia based on efficacy/toxicity integration...

Area of Science:

• Hematology
• Computational Biology
• Pharmacology

Background:

• Acute myeloid leukemia (AML) is an aggressive cancer often requiring combination drug therapy.
• Current computational methods for drug combinations primarily focus on synergy, neglecting efficacy and toxicity.
• Personalized treatment strategies are essential for improving AML patient outcomes.

Purpose of the Study:

• To develop a computational model for identifying effective and low-toxicity drug combinations for AML.
• To explore drug-response heterogeneity using a bipartite network approach.
• To validate novel drug combinations in preclinical models.

Main Methods:

• Constructed a bipartite network integrating patient tumor samples and drug-response data.
• Projected the network onto drugs to create a drug similarity network.
• Utilized community detection and pathway analysis to identify drug clusters targeting distinct biological processes.
• Selected top-efficacy, low-toxicity drugs from each cluster for validation via cell viability assays.

Main Results:

• Identified distinct drug clusters targeting different biological pathways in AML.
• Ruxolitinib-ulixertinib and sapanisertib-LY3009120 demonstrated significant synergistic effects on AML blast cells.
• These combinations exhibited superior efficacy with minimal toxicity compared to single agents.
• Drug combinations showed promising results in preclinical cell viability assays.

Conclusions:

• The developed computational network approach effectively identifies synergistic drug combinations for AML.
• Ruxolitinib-ulixertinib and sapanisertib-LY3009120 represent promising candidates for novel AML combination therapies.
• These findings support the development of personalized AML treatment strategies.
• The identified combinations have the potential to complement standard first-line AML treatments.