Molecular precision medicine: Multi-omics-based stratification model for acute myeloid leukemia

Area of Science:

• Hematology
• Genomics
• Computational Biology

Background:

• Acute myeloid leukemia (AML) presents significant challenges in treatment efficacy, relapse, and drug resistance.
• Multi-omics data integration is crucial for understanding AML heterogeneity.

Purpose of the Study:

• To develop and validate a multi-omics stratification model for acute myeloid leukemia (AML).
• To compare prognosis, clinical features, gene mutations, immune microenvironment, and drug sensitivity across AML subgroups.

Main Methods:

• Utilized TCGA database datasets including RNA sequencing, DNA methylation, and somatic mutations for AML.
• Developed a multi-omics stratification model classifying AML patients into distinct clusters (CS).
• Validated the model using external datasets from the GEO database (mRNA and miRNA).

Main Results:

• Classified 126 AML patients into 4 clusters (CS) with varying prognoses and characteristics.
• CS4 exhibited the best prognosis, linked to specific gene mutations (WT1, FLT3, KIT) and sensitivity to HDAC/BCL-2 inhibitors.
• CS3 showed the worst prognosis, associated with mutations (RUNX1, DNMT3A, TP53), higher cytotoxic cells/Tregs, and potential sensitivity to mTOR inhibitors.
• CS1 and CS2 demonstrated differential prognoses, gene mutation profiles (FLT3, NPM1, DNMT3A), and drug sensitivities (cytarabine, RXR agonists, FLT3 inhibitors).
• The stratification model showed robust validation in external datasets.

Conclusions:

• The multi-omics stratification model effectively categorizes AML patients, offering insights into prognosis and therapeutic strategies.
• This approach supports the transition towards multi-drug combination clinical trials and multi-targeted precision medicine for AML.
• The study provides a framework for personalized treatment approaches in AML based on integrated omics data.