aimed analytics logo

Revolutionizing AML Treatment: A Multi-omics Approach

Discover how a groundbreaking multi-omics classification system is redefining acute myeloid leukemia (AML) subtypes, offering new insights into prognosis and personalized treatment strategies — and learn something new every day!

Multi-omics Classification System Reveals Biological Differences in Acute Myeloid Leukemia and Clinical Implications

Acute myeloid leukemia (AML) is a heterogeneous group of diseases with varied genetic and clinical features, making the prognosis and treatment complex. Traditional classification systems using genetic mutations and chromosomal abnormalities fall short in capturing the intricate molecular interactions within AML. This limitation necessitates a multi-omics approach to integrate various molecular data types for a comprehensive understanding of AML.

Traditional classification systems primarily rely on genetic mutations and chromosomal abnormalities to predict outcomes and guide treatment. However, these systems do not fully capture the complexity of AML, particularly the interactions between different molecular features.

Multi-omics Approach to AML Classification

A recent study conducted by researchers at the Chinese Academy of Medical Sciences & Peking Union Medical College has introduced an innovative unsupervised AML multi-omic classification system (UAMOCS). This system integrates DNA methylation, transcriptomics, and mutation data to classify AML into three distinct subtypes: UAMOCS1, UAMOCS2, and UAMOCS3.

At a Glance:

  • Three Unique Subtypes: UAMOCS1 (poor survival, high chromosomal instability), UAMOCS2 (intermediate prognosis, immune suppression), UAMOCS3 (best prognosis, MYC pathway activation)

  • Drug Sensitivity: UAMOCS1 and UAMOCS2 show promise with a combination of Azacitidine and ABT-199, except for monocytic-type patients.

  • Immune Profiles: UAMOCS1 and UAMOCS2 are characterized by high immune cell infiltration, while UAMOCS3 has a barren immune landscape.

  • Prognostic Value: The UAMOCS system improves survival prediction and aligns with existing prognostic models.

Now, let’s dive a bit deeper into the complexities behind these findings.

Subtype Characteristics

UAMOCS1: Characterized by high LAG3 expression, chromosomal instability, myelodysplasia-related mutations, and poor survival outcomes. This subtype shows frequent AML-MR mutations and unstable chromosomal changes, correlating with a lower response rate to conventional therapies.

UAMOCS2: Exhibits a monocytic-like profile with intermediate prognosis, immune suppression, and activation of angiogenesis and hypoxia pathways. This subtype is marked by high infiltration of immunosuppressive cells including MDSCs and TITRs, and generally shows a demethylator phenotype.

UAMOCS3: Identified by CEBPA mutations and MYC pathway activation, this subtype has the best prognosis among the three, with favorable outcomes for patients.

Clinical Implications

The UAMOCS system provides significant prognostic insights, offering a new tool for patient stratification and potentially improving personalized treatment strategies. The study introduces 16 novel rapid classification signatures for clinical application and an R package ‘UAMOCS’ to re-classify patients based on transcriptomic profiles.

Drug Sensitivity Analysis: The study also highlights differences in ex-vivo drug responses among UAMOCS subtypes. Notably, patients in the UAMOCS1 subtype showed marked sensitivity to a combination of Azacitidine and ABT-199.

Validation and Reproducibility

The UAMOCS system was validated using transcriptomic data from four external datasets and a real-world cohort comprising 98 cases, named 'ihCAMs-AML'. The classifications were consistent across datasets, demonstrating the system's robustness.

Key Findings

  1. Prognostic Value: UAMOCS1 patients exhibited worse overall survival, while UAMOCS3 patients had better survival outcomes. This classification system aligns with existing prognostic models and improves the precision of prognostication.

  2. Immune Profiles: UAMOCS1 and UAMOCS2 subtypes were characterized by high immune cell infiltration and immune suppression, while UAMOCS3 showed a barren immune landscape. This finding underscores the relationship between immune status and disease progression in AML.

  3. Relapse Patterns in CBF-AML: Core binding factor (CBF) AML patients were divided into two groups within the UAMOCS system. This division reflects relapse patterns and highlights the relevance of UAMOCS in predicting treatment responses and outcomes.

Pathway Analysis

Pathway analysis revealed distinct dysregulations in immune responses and tumorigenesis pathways among the subtypes. For instance, UAMOCS2 was associated with VEGFA-mediated angiogenesis, and UAMOCS3 showed unique activation of the MYC signaling pathway.


The UAMOCS system represents a significant advancement in AML classification by incorporating multi-omics data, providing a more comprehensive understanding of AML's molecular landscape. The study's findings suggest that specific AML subtypes may benefit from tailored treatment regimens, such as the combination of Azacitidine and ABT-199 for certain subtypes.

The researchers noted:

Our study introduces 16 novel rapid classification signatures to expedite clinical application and patient stratification.
Song et al. (2024): Multi-omics classification system reveals biological differences of acute myeloid leukemia and clinical implications.

Future Directions

Further research with larger cohorts is needed to validate these findings and explore the broader applicability of the UAMOCS system. The development of the R package ‘UAMOCS’ facilitates its integration into clinical practice, potentially improving the management and treatment of AML.

By redefining the characteristics of AML subtypes, the UAMOCS system not only enhances prognostic accuracy but also guides the selection of effective therapeutic regimens, ultimately contributing to the advancement of personalized medicine in hematologic malignancies.