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

Paper Details

Abstracts, authors, affiliations, and disclosures.

Sunday, December 13, 2026
06:00 p.m. - 08:00 p.m.
MCCNO - Hall F
Technology and Procedures, Assays, Molecular Testing, Omics Technologies, Bioinformatics, Computational Biology, Emerging Technologies, Profiling, Treatment Considerations, Combination Therapy, Drug Development, Therapy Sequence, Biological Therapies, Antibody Therapy, Transplantation, Checkpoint Inhibitor, Immunotherapy, Non-Biological Therapies, Chemotherapy, Adverse Events, Artificial Intelligence (AI), Deep Learning, Machine Learning, Biological Processes, Apoptosis, Pathogenesis, Genomics, Hematopoiesis, Immune Mechanism, Metabolism, Molecular Biology, Multi-Systemic Interactions, Clinical Practice (Health Services and Quality), Practice Models, Value, Diseases, Myeloid Malignancies, Acute Myeloid Malignancies, Chronic Myeloid Malignancies, Measurable Residual Disease, Research, Fundamental Science, AML, APL, Translational Research, MDS, MPN, CMML, Clinical Research, Epidemiology, Health Outcomes Research, Health Disparities Research, Real-World Evidence, Survivorship, Study Population, Human, Adult
PAPER

3450 Disease stage- and risk-associated RNA editing signatures in acute myeloid leukemia and their utility for peripheral blood-based assessment

Tongjun Gu, PhD, Doan Bui2, Ji-Hyun Lee3
1 Versiti Blood Research Institute, Milwaukee, Wisconsin, United States
2 Versiti Blood Research Institute, Milwaukee, Wisconsin, United States
3 University of Florida, Gainesville, United States
Background
RNA editing, predominantly adenosine-to-inosine (A-to-I) editing mediated by ADAR enzymes, is a widespread post-transcriptional regulatory mechanism. However, its role in acute myeloid leukemia (AML) progression, relapse, and risk stratification remains incompletely understood, and whether clinically informative RNA editing signatures can be detected in peripheral blood (PB) remains unknown.
Aims
To systematically characterize RNA editing landscapes associated with disease stage, relapse, and ELN-2022 risk classification in AML, evaluate the reproducibility of RNA editing signatures across independent cohorts, and determine whether clinically informative RNA editing events can be detected in PB as a minimally invasive alternative to bone marrow (BM).
Methods
RNA-seq data from the ESAML cohort, including 59 paired diagnosis-relapse AML patients and eight age-matched healthy controls (126 samples after quality control; 93 BM and 33 PB), were analyzed using a stringent RNA editing discovery pipeline incorporating dual-aligner validation, extensive artifact filtering, and exclusion of known genomic variants. Recurrent editing sites detected in at least 5% of samples were retained. Site-specific differential editing was evaluated using mixed-effects beta-binomial regression to account for overdispersion, repeated measurements, demographic variables, technical covariates, and interferon-related transcriptional activity. Key findings were validated in independent AML cohorts (GAML, TCGA-LAML, and BEAT-AML). Global RNA editing burden and BM-PB concordance were also evaluated.
Results
We identified 166,323 high-confidence RNA editing sites mapping to 5,917 genes, with most sites located in intronic and 3′ untranslated regions. Overall, 1.2%–3.6% of evaluated editing sites were significantly associated with disease stage or ELN-2022 risk groups (FDR < 0.1).
Disease stage analyses revealed dynamic remodeling of RNA editing during AML progression. Editing signatures distinguished healthy controls, diagnosis, and relapse samples. Among shared disease-associated sites, 107 exhibited opposite editing patterns between healthy-versus-diagnosis and relapse-versus-diagnosis comparisons, indicating progressive reprogramming during disease evolution. Editing sites in MRPL42, MKRN1, CLK1, and CHST12 displayed progressive increases or decreases across disease stages.
Relapse-associated signatures demonstrated reproducibility across independent datasets. Forty-five relapse-associated editing sites were shared between ESAML and the independent GAML cohort, including four reproducible editing sites in TMEM165, COQ4, TIMM17A, and PLXDC2.
RNA editing profiles were also associated with ELN-2022 risk stratification. Intermediate- and adverse-risk AML exhibited substantial overlap in editing programs, with gene-level overlap approaching 40%–45% and more than 76% of shared sites demonstrating concordant effect directions. Cross-cohort validation identified a reproducible 3′UTR editing site in ABHD18 that was consistently associated with intermediate- and adverse-risk AML in both ESAML and TCGA-LAML PB samples.
Comparison of BM and PB demonstrated extensive conservation of RNA editing landscapes. Across ESAML and BEAT-AML cohorts, 77.5%–85.7% of editing sites and over 94% of edited genes were shared between tissues, whereas only 2.1%–2.3% of evaluated sites exhibited significant tissue-specific editing differences. Importantly, PB captured clinically informative editing events, supporting the use of minimally invasive blood sampling for AML risk assessment and longitudinal monitoring.
Higher global editing burden was associated with leukemic state, elevated white blood cell counts, adverse-risk disease, and refractory disease across independent cohorts, although site-specific editing signatures provided greater discriminatory power.
Conclusions
RNA editing landscapes undergo dynamic remodeling during AML progression and contain reproducible signatures associated with relapse and ELN-2022 risk. Multiple editing biomarkers were independently validated across AML cohorts. The extensive concordance between BM and PB supports the development of minimally invasive PB-based RNA editing biomarkers for AML risk stratification and longitudinal disease monitoring. Together, these findings establish RNA editing as a clinically relevant molecular layer in AML and provide a foundation for future RNA editing-based biomarker development.
Tongjun Gu, PhD: No relevant conflicts of interest to declare.