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, Molecular Testing, Omics Technologies, Bioinformatics, Computational Biology, Emerging Technologies, Treatment Considerations, Combination Therapy, Biological Therapies, Bispecific Antibody Therapy, (CAR)-T Cell Therapies, Biological Processes, Pathogenesis, CHIP, Genomics, Molecular Biology, Clinical Practice (Health Services and Quality), Clinical Guidelines, Diseases, Lymphoid Malignancies, Lymphoid Leukemias, Measurable Residual Disease, ALL, Research, Fundamental Science, Translational Research, Clinical Research, Epidemiology, Real-World Evidence, Study Population, Human, Pediatric, Neonatal
PAPER

3324 CpG methylation analysis identifies myeloid-primed and stem-like subtypes of adult ETP-ALL with distinct clinical outcomes

Eitan Kugler, MD PhD, Ken Furudate2, Curtis Lachowiez, MD3, Eulàlia Genescà4, Lianqun Qiu, MD, PhD5, Sanam Loghavi, MD6, Farhad Ravandi, MD7, Nitin Jain, MD8, Gautam Borthakur, MD9, Nicholas Short, MD10, Sherry Pierce, BSN, RN11, Rebecca Garris12, Guillermo Garcia-Manero, MD13, Andrew Futreal, PhD14, Josep-Maria Ribera, PhD15, Hagop Kantarjian, MD16, Elias Jabbour, MD17, Koichi Takahashi, MD,PhD18
1 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
2 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
3 Oregon Health and Science University, Portland, Oregon, United States
4 Josep Carreras Leukemia Research Institute, Badalona, Barcelona, Spain
5 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
6 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
7 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
8 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
9 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
10 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
11 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
12 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
13 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
14 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
15 ICO-Hospital Germans Trias i Pujol, Josep Carreras Leukemia Research Institute (IJC), Universitat Autònoma de Barcelona, Barcelona, Spain
16 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
17 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States
18 The University of Texas MD Anderson Cancer Center, Houston, Texas, United States

T-cell acute lymphoblastic leukemia (T-ALL) is classified by immunophenotype (IP) into cortical, mature, and early subtypes, including early T-cell precursor (ETP) and near-ETP. IP incompletely captures T-ALL heterogeneity, particularly within ETP-ALL, a genetically diverse subtype arising from early thymic progenitors with residual lymphoid and myeloid potential and associated with chemoresistance and relapse. To characterize this heterogeneity, we integrated genome-wide CpG methylation, RNA sequencing, and targeted mutation profiling across adult T-ALL.

Bone marrow samples were collected from 102 adults with T-ALL between 2000 and 2018 (47 newly diagnosed [ND], 55 relapsed/refractory [R/R]). Multi-omics analysis integrating targeted gene sequencing of 297 genes (N = 101), CpG methylation profiling (N = 89) (Illumina EPIC array), and RNA sequencing (N = 68) were performed.  IP was thymic in 32 patients (31.4%), mature in 6 (5.9%), ETP in 22 (21.6%), ETP-like in 19 (18.6%), early in 8 (7.8%), and inconclusive in 15 (14.7%). 43% had diploid karyotype and 24% non-recurrent abnormalities.
ND patients mainly received hyper-CVAD–based regimens, whereas treatment regimens varied among R/R patients. Consensus non-negative matrix factorization of the top 5% most variable CpG probes identified four subgroups (C1–C4) across the full cohort. ND and R/R patients were represented in all four clusters, indicating that cluster assignment was independent of disease setting. C1 and C2 were enriched with mature/thymic cases, whereas C3 and C4 were enriched for ETP/near-ETP. To define biological differences between the two ETP-enriched clusters, C3 versus C4 analyses were restricted to ND ETP cases.

Global methylation was higher in C4-ETP than C3-ETP (mean beta 0.68 vs 0.63, p=0.029). Of 36,787 differentially methylated positions, 35,460 (96.4%) were hypermethylated in C4 and enriched at promoters (47%) of PRC2/H3K27me3-marked developmental genes. EpiTOC-derived mitotic age was also higher in C4-ETP (0.654 vs 0.389, p=0.029). Transcriptional profiling showed enrichment of embryonic and leukemic stem cell pathways in C4-ETP, whereas C3-ETP was enriched for myeloid differentiation and cell-cycle/DNA-replication pathways (padj<0.05). C4-ETP expressed a stem-like program marked by HOXA9, LYL1, MEF2C, and CD34. In contrast, C3-ETP expressed myeloid markers including CD13 and HLA-DR, had a higher frequency of FLT3 mutations (75%), and lacked NOTCH1 mutations (vs 38% cohort-wide). C4-ETP was associated with significantly worse clinical outcomes, with a median event-free survival (EFS) of 3.6 months versus not reached in C3-ETP (p=0.01). Together, C4-ETP represented a mitotically old, yet developmentally immature state characterized by hypermethylated developmental promoters and stem/progenitor-associated expression, in contrast to the myeloid-primed C3-ETP.

The marked upregulation of CD13 in C3-ETP relative to C4-ETP (log2FC=3.91; padj=0.001) prompted us to evaluate whether the C3 myeloid phenotype could be captured clinically by routine flow cytometry. To this end, we studied an expanded cohort of 42 ND ETP-ALL patients for their flow data. We developed a 2-marker myeloid score (CD13≥40%, HLA-DR≥10%), using maximally selected rank statistics followed by 1,000-resample bootstrapping. Patients with higher myeloid score were associated with higher CR/CRi rates (71%, 83%, and 94% for scores 0, 1, and 2; p=0.049). EFS improved with increasing myeloid score (median 8.2, 28.9 months, and not reached; p=0.013), with each point associated with a 61% reduction in the hazard of an event (HR 0.39; 95% CI, 0.20–0.75; p=0.005). The myeloid score retained its prognostic impact on EFS in an independent external cohort of 34 ND ETP patients treated on the PETHEMA ALL-HR-11/LAL2019 trials (HR 0.35 per point; 95% CI, 0.17–0.74; p=0.006). In summary, genome-wide CpG methylation profiling identifies two biologically and clinically distinct ETP-ALL subtypes: a myeloid-primed, relatively chemosensitive C3 subtype and a stem-like, chemoresistant C4 subtype. A clinically accessible 2-marker CD13/HLA-DR flow cytometry score derived from the C3 phenotype stratified EFS in both discovery and independent validation cohorts, supporting prospective validation as a tool for risk stratification in ETP-ALL.

Eitan Kugler, MD PhD: No relevant conflicts of interest to declare.