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

Paper Details

Abstracts, authors, affiliations, and disclosures.

Monday, December 14, 2026
06:00 p.m. - 08:00 p.m.
MCCNO - Hall F
Access, Representation, Population Health, Technology and Procedures, Imaging, Molecular Testing, Omics Technologies, Serologic Tests, Bioinformatics, Computational Biology, Emerging Technologies, Profiling, Treatment Considerations, Combination Therapy, Therapy Sequence, Biological Therapies, Antibody Therapy, Transplantation, Bispecific Antibody Therapy, Chimeric Antigen Receptor, (CAR)-T Cell Therapies, Immunotherapy, Adverse Events, Artificial Intelligence (AI), Computer Vision, Deep Learning, Machine Learning, Biological Processes, Apoptosis, CHIP, Genomics, Immune Mechanism, Diseases, Lymphoid Malignancies, Plasma Cell Disorders, Myeloid Malignancies, Measurable Residual Disease, Research, Fundamental Science, Translational Research, Clinical Research, Health Outcomes Research, Real-World Evidence, Study Population, Human, Adult
PAPER

5841 Single-cell multiome profiling identifies epigenetic reprogramming of apoptotic balance as a marker of early therapeutic resistance

Arwa Bohra, MD, Amelia Mazzone, Ph.D.2, Jenna Fernandez, Ph.D.3, Wazim Mohammed Ismail, Ph.D.4, S. Vincent Rajkumar, MD5, Alexandre Gaspar-Maia, PhD6, Shaji Kumar, MD7, Moritz Binder, MD, MPH8
1 Mayo Clinic, Rochester, United States
2 Mayo Clinic, Rochester, United States
3 Mayo Clinic, Rochester, United States
4 Mayo Clinic, Rochester, United States
5 Mayo Clinic, Rochester, United States
6 Mayo Clinic, Rochester, United States
7 Mayo Clinic, Rochester, United States
8 Mayo Clinic, Rochester, United States

Background: Suboptimal response to front-line therapy is a marker of early functional high-risk myeloma. Pre-treatment identification of quadruplet response is desirable to enable early interception of high-risk multiple myeloma.

Methods: The Mayo Clinic discovery cohort included 12 patients with newly diagnosed multiple myeloma: 7 resistant patients, defined as achieving a partial response or worse, including 6 treated with daratumumab, ixazomib, lenalidomide, and dexamethasone (D-IRd) and 1 treated with bortezomib, lenalidomide, and dexamethasone (VRd), and 5 responsive patients, including 2 treated with D-IRd and 3 with VRd. We profiled 15,496 cells from these patients. After quality control, doublet removal, and iterative plasma cell purification, 7,891 plasma cells (3,956 resistant, 3,935 responsive) were retained for transcriptomic analysis; 3,026 cells passed ATAC-seq quality control. Differential accessibility (logistic regression + pseudo-bulk DESeq2) and differential expression (Wilcoxon + pseudo-bulk DESeq2) were assessed between cells from resistant and responsive patients, with motif enrichment (JASPAR2020) and chromVAR motif activity. RNA-ATAC concordant, resistance-associated genes identified in the discovery cohort were tested for prognostic association in the MMRF CoMMpass validation cohort (n=754) via Cox models, then carried into four cross-validated survival models to build a therapeutic resistance signature.

Results: In the discovery cohort, high-confidence concordant chromatin regions (71 peaks) mapped resistant-open accessibility near CEBPB, MECOM, and TGM2, and responder-open accessibility near RUNX1 and KLF13. Resistant cells showed enrichment of AP-1/JUN, NF-κB/REL, and HIF1A motifs, most pronounced in the VRd subgroup, while responder cells retained ETS/IRF and PRDM1-associated chromatin activity. Concordant single-cell/pseudo-bulk differential expression identified 87 high-confidence genes (34 resistant-associated, including PMAIP1/NOXA and CCND1; 53 responder-associated, including BCL2L1 and RUNX1), with relative loss of plasma-cell differentiation markers (IRF4, XBP1, MZB1) and gain of proliferative/stress-adapted programs (CCND1, KLF4, NFKB2, RELA) in resistant cells. Pathway analysis identified PD-L1 regulation, SPOP-mediated PD-L1 degradation, and cyclin/S-phase pathways as resistance-enriched programs. In the validation cohort, resistance-associated genes STK39 (HR 1.2) and YBX3 (1.1) predicted inferior survival, while response-associated genes BCL2L1 (0.8) and ABCA7 (0.9) predicted superior survival. Random survival forest (RSF) and elastic net, trained and cross-validated in the MMRF cohort, achieved similar discrimination for predicting resistance-associated risk (C-index 0.666±0.034 and 0.666±0.049), outperforming stepwise Cox (0.631±0.025). The final 20-gene RSF model was selected for its parsimony, stratifying median overall survival at 102.5 months in the low-risk versus 21.0 months in the high-risk tertile. Top-ranked genes by variable importance included TUBA1B (HR 1.5), CD38 (0.7), LBR (1.3), SIPA1L3 (0.8), EZH2 (1.5), MAP7 (0.8), KIF21B (1.2), ITPRIP (0.7), PAM (0.7), CDKN2A (1.5), and CHEK1 (1.6), with lower CD38 expression predicting worse outcomes. Across resistance status in the Mayo discovery cohort and RSF-predicted risk tertiles in the MMRF cohort, the BCL2:BCL2L1 expression ratio was significantly higher in resistant cells and RSF high-risk tertiles (both p<0.01), reflecting reduced BCL-XL–associated cell survival signaling relative to BCL-2. The MCL1:PMAIP1 ratio decreased with resistance and higher RSF-predicted risk, driven by compensatory NOXA induction rather than reduced MCL-1 dependence, supported by a modestly lower BCL2:MCL1 ratio in resistant cells at single-cell resolution.

Conclusions: Adverse gene expression at the time of diagnosis can identify patients at increased risk for the early development of therapeutic resistance to the contemporary quadruplet regimens. Altered BCL2 signaling emerged as one possible underlying mechanisms and further mechanistic studies are warranted to identify therapeutic vulnerabilities in this type of functional high-risk myeloma.
Arwa Bohra, MD: No relevant conflicts of interest to declare.