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

504 A Mecom defined transcriptional state portends poor prognosis across genetic subgroups of AML

Justin Wheat, MD, PhD, Justice Ameyi, M.S.2, Deedra Nicolet, MS3, Ann-Kathrin Eisfeld, MD4, Ross L. Levine, Prof. Dr.5
1 Memorial Sloan Kettering Cancer Center, New York, New York, United States
2 The Ohio State University Comprehensive Cancer Center, Columbus, Ohio, United States
3 The Ohio State University Comprehensive Cancer Center, Columbus, Ohio, United States
4 The Ohio State University Comprehensive Cancer Center, Columbus, Ohio, United States
5 Memorial Sloan Kettering Cancer Center, New York, New York, United States

Introduction: AML with 3q26.1 rearrangements (MECOMr) has an extremely poor prognosis and is driven by overexpression of the hematopoietic stem cell (HSC) transcription factor MECOM, which impairs differentiation and promotes therapy resistance. Whether MECOM expression has prognostic significance beyond MECOMr AML, and whether MECOM-regulated transcriptional programs define a common leukemia stem cell (LSC) state across AML subtypes, remains unknown. We sought to (1) determine the prognostic significance of MECOM expression at diagnosis across adult AML and (2) define a shared MECOM-associated LSC transcriptional network.

Methods/Results: Analysis of the BEAT-AML cohort (n=342; PMID:33106665) identified a bimodal pattern of MECOM expression, with 14% of patients classified as MECOM expressors. MECOM expression increased with ELN-2022 risk (26.8% Adverse, 9.7% Intermediate, 0% Favorable; p=1.06×10⁻⁹) and was detected across multiple genetic subtypes (ELN2022), including MECOMr (100%), KMT2Ar (47.5%), TP53-mutant (25%), AML with MDS related changes (AML-MRC) (21%), AML with MDS related cytogenetic changes (AML-MRGC) (11%), AML-NOS (9.3%), and NPM1c/FLT3-ITD+ (2.3%). MECOM expression independently predicted inferior overall survival (HR 1.79, 95% CI 1.22–2.63, p=0.003; median OS 9.3 vs 17.9 months) and lower remission rates (CR+CRi 48.5% vs 67%, p=0.035), independent of genotype and ELN risk.

Differential expression analysis identified 131 genes associated with MECOM expression. LASSO Cox regression generated a 24-gene TARGET-score that predicted inferior overall survival across AML subtypes (HR 2.43, 95% CI 1.77–3.33, p<0.0001). External validation in the ALLIANCE cohort (n=873) confirmed its prognostic value (HR 2.05, 95% CI 1.74–2.40, p<0.001). Backward AIC selection further refined the model to a 14-gene score that independently predicted overall survival in both ALLIANCE (HR 2.05, p<0.001) and BEAT-AML (HR 1.79, p<0.001). The 14-gene score retained prognostic significance within NPM1c/FLT3-ITD+, KMT2Ar, TP53-mutant, and AML-MRC subsets.

To relate the TARGET-score to cellular heterogeneity, we repurposed the Xenium spatial transcriptomic platform to perform high sensitivity single cell transcriptomics on leukemia tumor arrays we created of diagnostic marrow samples from 29 patients (78,160 cells). MECOM-positive cells were identified across all genetic subtypes, comprising 30–90% of cells in MECOMr AML, <10% in favorable-risk AML, and variable proportions in TP53-mutant, AML-MRC, AML-MRGC, and KMT2Ar disease. Leiden clustering identified 20 cellular states, including four HSC-like populations. Three HSC-like clusters exhibited high expression of MECOM, MSI2, CD74, BCL11A, and ZFP36L2 ("MECOM-type HSPCs"), whereas a fourth HSC-like cluster expressed low MECOM and was enriched for TAL1, LYL1, GATA2, and MYB. Individual leukemias preferentially expanded one HSPC state. Gamma-distribution modeling identified a parsimonious five-gene single-cell signature (scMECOM-score: MECOM, MSI2, FAM30A, CD74, ADGRG1) that robustly distinguished this cell state across the cohort (ΔBIC=11,290; Ashman's D=1.95) and within individual patients (23/29 with D>1.9). The scMECOM program was present across AML but was enriched in adverse-risk disease (>50% of cells in MECOMr, >20% in TP53-mutant AML, and >25% in AML-MRC and AML-MRGC). Pseudo-bulk TARGET-scores correlated significantly with scMECOM-scores (Spearman ρ=0.568, p=0.0013), suggesting that the bulk transcriptional signature reflects the abundance of a shared MECOM-driven LSC state.

Conclusions: MECOM expression identifies a high-risk subset of AML independent of genotype and ELN classification. We developed and externally validated a MECOM-associated TARGET-score that provides prognostic information across genetically diverse AML. Integration with single-cell spatial transcriptomics demonstrates that this signature reflects a conserved MECOM-dominant HSC/LSC program enriched in adverse-risk AML, supporting a shared stem-cell state across multiple AML subtypes. Ongoing studies are defining the regulatory circuitry and therapeutic vulnerabilities of this program. As clinical transcriptomic profiling becomes more widely adopted, the TARGET-score may complement genetic risk stratification, particularly in heterogeneous high-risk AML subgroups.

Justin Wheat, MD, PhD: No relevant conflicts of interest to declare.