Acute myeloid leukemia (AML) stands as one of the most aggressive blood cancers. Classification approaches determine individual treatment plans. For decades, this categorization relied on genetic mutations within leukemic cells, directing clinical decisions and drug development. However, gene mutations represent only part of the disease picture. The epigenome—the regulatory layer controlling which genes cells utilize—was long suspected to play an equally critical role in AML, yet the complete understanding remained elusive.

Researchers led by Professor Seishi Ogawa from Kyoto University’s Department of Pathology and Tumor Biology and Institute for the Advanced Study of Human Biology, along with Assistant Professor Yotaro Ochi and Professor Sören Lehmann from Karolinska Institute, conducted extensive epigenomic analysis of over 1,500 AML patient samples. The investigation demonstrated that AML divides into 16 distinct subgroups according to epigenomic characteristics. Each subgroup displays unique molecular composition, clinical outlook, and pharmaceutical sensitivities.

The findings reveal that “AML is not simply a ‘genomic disease’: its epigenomic architecture also determines disease behavior, prognosis, and treatment response,” establishing foundation for precision medicine incorporating epigenomic data.

Background

AML develops when blood-forming cell mutations interfere with standard production of mature blood components from hematopoietic stem cells. Immature blood cells instead accumulate and proliferate uncontrollably in bone marrow. Two decades of next-generation sequencing revealed numerous genetic mutations involved in AML, transforming understanding and targeted therapy development.

Beyond the genome, the epigenome—encompassing DNA chemical modifications and chromatin-packaging proteins—governs cellular function. Epigenetic alterations contribute to AML by influencing cell differentiation and proliferation. Previously, comprehensive epigenomic abnormality views across AML remained unavailable.

Key Findings

Researchers applied ATAC-seq methodology to 1,563 AML samples from Swedish and Japanese cohorts. The resulting eCHROMA AML dataset represents the largest chromatin accessibility resource for any cancer type. Chromatin analysis revealed 16 characteristic subgroups. Single-cell sequencing of over 280,000 cells from 36 patients confirmed each subgroup maintains distinctive, conserved chromatin profiles.

Each subgroup contained distinct mutation combinations, differentiation states, expression profiles, methylation patterns, and regulatory networks. Notably, numerous subgroups misaligned with existing genomic classifications, indicating genome analysis alone cannot capture complete AML diversity.

Epigenomic examination showed each subgroup organized around distinctive transcription-factor networks and super-enhancer structures with unique gene-regulatory programs. Clinically, incorporating chromatin data with existing genomic risk categories substantially enhanced prognostic accuracy across both cohorts.

The analysis uncovered unexpected drug sensitivities mutations alone missed: three subgroups responded to MEK inhibitors despite lacking typical RAS mutations guiding such drug use. Remarkably, one subgroup featuring frequent RUNX1 mutations and early B-cell precursor chromatin characteristics demonstrated high sensitivity to ABL inhibitors—traditionally employed for different blood cancers.

Looking Ahead

This inaugural large-scale investigation establishes that chromatin state, alongside gene mutations, proves essential for understanding leukemia behavior and individual case biological identity. Results could enhance diagnosis, prognostic assessment, and treatment selection within precision medicine frameworks. Supporting clinical adoption, researchers identified a compact 30-gene expression signature distinguishing chromatin-defined high-risk subgroups using standard sequencing.

The multi-omics database generated serves as foundational resource for cancer epigenomics broadly, facilitating new therapeutic target and disease mechanism discovery. The team intends developing simpler, lower-cost diagnostic methods and refining treatment approaches per subgroup, advancing toward routine clinical implementation.

---

Journal: Nature
DOI: 10.1038/s41586-026-10703-4
Article Title: Chromatin landscape and epigenetic heterogeneity of acute myeloid leukaemia
Publication Date: 8-Jul-2026

Source: EurekAlert / Kyoto University