2026 ASH®
Annual Meeting & Exposition December 12-15, 2026 ▪ New Orleans, Louisiana

Paper Details

Abstracts, authors, affiliations, and disclosures.

Sunday, December 13, 2026
09:30 a.m. - 11:00 a.m.
MCCNO - Great Hall BC
Technology and Procedures, Assays, Molecular Testing, Omics Technologies, Bioinformatics, Computational Biology, Emerging Technologies, Profiling, Treatment Considerations, Drug Development, Artificial Intelligence (AI), Machine Learning, Biological Processes, Molecular Biology, Diseases, Lymphoid Malignancies, Lymphoid Leukemias, Myeloid Malignancies, Acute Myeloid Malignancies, Measurable Residual Disease, ALL, Research, AML, APL, Translational Research, Clinical Research, Real-World Evidence, Study Population, Human, Adult
PAPER

500 Genome-wide DNA methylation is predictive of outcome in KMT2A-rearranged acute myeloid leukemia

Jeevitha D'Souza, Postdoctoral fellow, Benjamin J. Huang, MD2, Anushka Walia, BS3, Hamid Bolouri4, Todd A. Alonzo, PhD5, Yi-Cheng Wang6, Rhonda E. Ries, MA7, Christopher E. Randolph8, Akhilesh Kaushal9, Abigail S. Lin10, Colin Smith11, Li Fan, PhD12, Wallace A. Bourgeois, MD13, Alan S. Gamis, MD14, Richard Aplenc I, MD, PhD15, Maximilian Schönung, PhD16, Daniel B. Lipka, MD17, Xiaotu Ma, PhD18, Katherine Tarlock I, MD19, Tim Triche, PhD20, Jason Farrar, MD21, Soheil Meshinchi, MD, PhD22
1 Department of Pediatrics, University of California San Francisco, San Francisco, California, USA, San Francisco, California, United States
2 Department of Pediatrics, University of California San Francisco, San Francisco, California, USA, San Francisco, California, United States, Helen Diller Family Comprehensive Cancer Center, University of California San Francisco, San Francisco, California, USA, San Francisco, United States
3 School of Medicine, University of California San Francisco, San Francisco, California, USA, San Francisco, California, United States
4 Benaroya Research Institute, Seattle, Washington, USA, Seattle, Washington, United States
5 Department of Population and Public Health Sciences, University of Southern California, Los Angeles, California, USA, Los Angeles,, United States
6 Children's Oncology Group, Monrovia, California, USA, Monrovia, United States
7 Fred Hutchinson Cancer Center, Seattle, Washington, USA, Seattle, Washington, United States
8 University of Arkansas for Medical Sciences & Arkansas Children's Research Institute, Little Rock, Arkansas, United States
9 University of Arkansas for Medical Sciences & Arkansas Children's Research Institute, Little Rock, Arkansas, United States
10 Department of Pediatrics, University of California San Francisco, San Francisco, California, USA, San Francisco, California, United States
11 Department of Pediatrics, University of California San Francisco, San Francisco, California, USA, San Francisco, California, United States
12 Department of Computational Biology, St. Jude Children’s Research Hospital, Memphis, Tennessee, USA, Memphis, Tennessee, United States
13 Department of Pediatric Oncology, Dana-Farber Cancer Institute and Boston Children's Hospital, Boston, Massachusetts, USA, Boston, Massachusetts,, United States
14 Children's Mercy Hospitals and Clinics, Kansas City, Missouri, USA, Kansas City, Missouri, United States
15 Department of Biostatistics, Epidemiology, and Informatics, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA, Philadelphia, United States, Division of Oncology, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, USA, Philadelphia, Pennsylvania, United States
16 Section of Translational Cancer Epigenomics, Division of Translational Medical Oncology, German Cancer Research Center and National Center for Tumor Diseases, Heidelberg, Germany, Heidelberg, Germany
17 Faculty of Medicine, Otto-von-Guericke-University, Magdeburg, Germany, Magdeburg, Germany, Section of Translational Cancer Epigenomics, Division of Translational Medical Oncology, German Cancer Research Center and National Center for Tumor Diseases, Heidelberg, Germany, Heidelberg, Germany
18 Department of Computational Biology, St. Jude Children’s Research Hospital, Memphis, Tennessee, USA, Memphis, Tennessee, United States
19 Division of Hematology and Oncology, Seattle Children’s Hospital, Seattle, Washington, USA, Seattle, Washington, United States
20 Van Andel Research Institute, Grand Rapids, Michigan, USA, Grand Rapids, Michigan, United States
21 University of Arkansas for Medical Sciences & Arkansas Children's Research Institute, Little Rock, Arkansas, United States
22 Division of Hematology and Oncology, Seattle Children’s Hospital, Seattle, Washington, USA, Seattle, Washington, United States, Fred Hutchinson Cancer Center, Seattle, Washington, USA, Seattle, Washington, United States

Introduction: Structural and genetic alterations have been comprehensively characterized for pediatric acute myeloid leukemia (AML), and integrating cytogenetic and genetic sequencing data now forms the basis for contemporary disease risk stratification. However, the prognostic significance of epigenetic changes, such as DNA methylation, remains uncertain. The current standard of care is to prescribe consolidative hematopoietic stem cell transplantation (HSCT) for patients with high-risk genetic features and to prescribe chemotherapy alone for patients with low-risk genetic features. In contrast, for patients with standard-risk genetic features, the end of induction response (i.e., MRD) is used to determine consolidation therapy (chemotherapy alone versus HSCT in CR1). However, improved risk stratification for standard-risk patients is needed since the vast majority have no detectable disease after induction and half of these patients still experience relapse.

Methods: To address this question, we performed genome-wide DNA methylation profiling on patients enrolled on the two most recent phase 3 clinical trials completed by the Children’s Oncology Group (AAML1031 and AAML0531), comprising a study population of 1,466 patients. We divided our study population into a training cohort (half of the AAML1031 study participants, n = 428) and two validation cohorts (the other half of the AAML1031 study participants, n = 428, and AAML0531 study participants, n = 610). To compare methylation signatures between AML and normal hematopoietic cell populations, we also performed genome-wide DNA methylation profiling on normal bone marrow (NBM) samples from 44 healthy individuals.

Results: First, from our training cohort, AML specific CpG sites were selected by comparing AML and NBM methylation profiles and selecting CpG sites that are differentially methylated. Next, we performed unsupervised hierarchical clustering on the 500 most variably methylated CpG sites. Importantly, such unsupervised analysis of DNA methylation has allowed identification of epigenetic subgroups in juvenile myelomonocytic leukemia (JMML), which are currently being tested as a biomarker for prospective risk stratification in a clinical trial (NCT05849662).

We repeated the above analysis for every major genetic subtype of pediatric AML and found that methylation profiles were significantly prognostic for patients with standard-risk KMT2A-rearranged (KMT2Ar) AML (by partner fusion), which comprise nearly 50% of standard-risk pediatric AML cases. Specifically, unsupervised hierarchical clustering on the 500 most variably methylated CpG sites for patients with KMT2A::MLLT3 AML (the most prevalent KMT2Ar fusion partner) resulted in two distinct groups based on methylation status, with patients within the hypermethylated group having significantly worse EFS and OS compared to the hypomethylated group: EFS 6.7% versus 42% (p = 0.005) and OS 8.9% versus 64.7% (p = 0.001). The methylation signature derived from our training cohort was then applied to our KMT2A::MLLT3 AML validation cohorts and we observed similar methylation based clustering and impact on outcomes. For our AAML1031 validation cohort: 5-year EFS 9.1% versus 44.8% (p = 0.007) and 5-year OS 18.2% versus 71.7% (p < 0.001). For our AAML0531 validation cohort: 5-year EFS 10% versus 59.5% (p = 0.011) and 5-year OS 20.0% versus 71.8% (p = 0.011). Similar results were observed for other standard-risk non-KMT2A::MLLT3 AML cases. We compared methylation profiling for standard-risk KMT2Ar AML to other biomarkers and found that methylation profiling outperforms MRD by flow cytometry, transcriptional signatures, and CD34 expression on the basis of receiver operating characteristic area under the curve (ROC AUC) analysis.

Conclusion: Based on our results, we are incorporating methylation profiling into the next COG AML frontline clinical trial. Since KMT2Ar AML comprise nearly 50% of standard-risk AML cases, we anticipate that methylation profiling will serve an important role in risk stratification. Future clinical trials will need to focus on prospective validation, assessing the impact of menin inhibitors, and determining whether hypomethylating agents have the potential to reverse the inferior outcomes associated with hypermethylation.

Jeevitha D'Souza, Postdoctoral fellow: No relevant conflicts of interest to declare.