Clonal hematopoiesis of indeterminate potential (CHIP) is a precursor condition characterized by the expansion of mutant hematopoietic stem and progenitor cell (HSPC) clones that increases the risk of hematologic malignancies. Although genome-wide association studies have identified multiple non-coding loci associated with CHIP susceptibility, their mechanisms remain unclear. We hypothesized that CHIP risk variants alter enhancer activity in HSPCs. To test this, we screened 1,374 non-coding variants from 51 CHIP-associated loci using a Massively Parallel Reporter Assay (MPRA) in the CD34+ fraction of MUTZ-3 cells. We identified 87 regulatory variants across 32 loci. Targeted genome editing in hematopoietic cells and complementary reporter assays in primary human HSPCs validated enhancer activity for variants regulating NKD2, FLT3, and MSI2. Functional studies demonstrated that increased MSI2 expression, modeling the effect of the CHIP risk allele, promotes clonal expansion of TET2-deficient HSPCs, providing a mechanistic link between inherited non-coding variation and CHIP clonal expansion.
Acute myeloid leukemia (AML) is prone to relapse driven by therapy-persistent residual cells. To discover specific vulnerabilities in this population, we performed genome-wide CRISPR interference screens in leukemia cells treated with multiple agents. KHSRP was the top hit, whose depletion sensitized AML cells to therapy and substantially prolonged survival in treated AML-bearing mice. Analysis of in vivo residual disease after venetoclax/azacitidine treatment identified downregulation of the vitamin C and uric acid transporter SLC23A1, which mediated resistance to multiple therapies. KHSRP depletion restored SLC23A1 expression by preventing its ZC3H4-mediated nuclear mRNA degradation. KHSRP depletion therefore enhanced the synergistic cytotoxicity of vitamin C and uric acid, particularly in therapy-persistent leukemia cells. Re-expression of TET2 overrode the chemosensitizing effect of KHSRP depletion in TET2-mutant leukemia, suggesting that KHSRP-linked phenotypes were related to vitamin C and uric acid-mediated TET activation. These findings nominate targeting KHSRP to enhance treatment efficacy and selectively eradicate residual AML.
The pathobiology of multiple myeloma is influenced by cells of the bone marrow, but whether tumor support is organized in a spatially-defined tumor microenvironment (TME) remains unclear. Using spatial transcriptomics and imaging mass cytometry, we show that myeloma cells initially exist as scattered cells throughout the marrow, yet with disease evolution condense into a spatially-defined TME. Dense tumor nodules contain a cellular ecosystem distinct from the surrounding marrow, enriched for macrophages, conventional dendritic cells, and CD8⁺ T cells, as well as endothelial cells and THY1⁺ mesenchymal stromal cells. Conversely, neutrophil lineage cells are absent from this TME but interact with scattered myeloma cells outside of dense nodules. Presence of the dense tumor architecture at diagnosis confers a worse prognosis, emphasizing the relevance of spatial organization of the bone marrow. Together, these findings define a discrete spatial and cellular myeloma TME, that provides spatially-guided insights into patient stratification and disease pathobiology.
Oncogenic KRAS and NRAS mutations are common in hematologic malignancies, but their signaling in this context remains less well characterized than in carcinomas. Using multi-omics screens in multiple myeloma, we sought to identify regulators of RAS activity. We found that the phosphatase PP1C dephosphorylated conserved RAS residue T148, permitting LZTR1-dependent proteasomal degradation. LZTR1 was ineffective against KRAS A146 gain-of-function mutations, which lie adjacent to T148 and are enriched in hematologic cancers, such as diffuse large B cell lymphoma and acute myeloid leukemia. Remarkably, KRAS protein stability was four-fold lower in hematologic versus carcinoma cells, revealing a unique therapeutic opportunity targeting RAS protein stability. PAK1 and PAK2 shielded RAS from LZTR1-dependent degradation by phosphorylating T148, and inhibiting PAK1/2 activity improved RAS-directed therapy. Collectively, these findings reveal a regulatory circuit governing RAS stability that is preferentially active in blood cancers and potentially druggable.
It remains uncertain whether lower CD19 expression is clinically relevant for outcomes of CD19 CAR-T therapy of large B-cell lymphoma (LBCL). We conducted an integrative analysis of tumor CD19 levels with centralized quantitative assessment by flow cytometry (FC), immunohistochemistry and RNA-sequencing in a LBCL cohort (n=301). Pre-treatment CD19 by FC correlated with 1-year progression-free survival following axicabtagene ciloleucel or lisocabtagene maraleucel (16%, 53% and 64% for CD19low, CD19intermediate and CD19high, respectively; p=0.005). After CAR-T relapse, proportion of CD19low tumors increased (from 17% to 35%). Concordant associations were observed by immunohistochemistry and RNA-sequencing. Transcriptomic profiling linked lower CD19 to inflammatory pathways, validated in an external LBCL cohort (n=1,017). Genetic loss of the CD19 locus was observed in 8% of pre-CAR-T and 11% of post-CAR-T tumors (p=0.82), while no CD19 coding mutations were identified. Overall, CD19 expression is a clinically meaningful determinant of CAR-T outcomes, supporting quantitative assessment to improve risk stratification.
Chimeric antigen receptor (CAR) technology has revolutionized B-cell malignancy treatment by enabling T cells to effectively recognize and target lineage-specific surface antigens. However, CAR T cells show limited efficacy against myeloid neoplasms and solid tumors due to challenges in identifying suitable surface targets. In this study, we present a CAR targeting the intracellular WT1 oncoprotein, cross-presented by surface HLA class II (HLA-II) alleles. WT1-CAR T cells, derived from an antibody raised solely against a WT1 peptide, recognized the WT1330-348 peptide promiscuously presented by 18 out of 20 tested HLA-II alleles, overcoming traditional HLA restrictions. WT1-CAR T cells specifically recognized leukemic cells in a WT1- and HLA-II-dependent manner and mediated an antitumor response in vitro and in vivo. This approach broadens CAR-targetable antigens beyond traditional HLA restrictions and offers a promising therapeutic option to a wide and genetically diverse patient population. SIGNIFICANCE:Leveraging the promiscuous binding of HLA-II-peptide complexes, we developed a CAR T-cell approach targeting an intracellular oncoprotein WT1 presented across diverse HLA-II families. Our study establishes a framework for CAR therapies against intracellular antigens, extending potential CAR T-cell applications to new cancer types and patient populations.
In this issue of Blood Cancer Discovery, Merz and colleagues study more than 600 recipients of chimeric antigen receptor T-cell (CAR T) and/or bispecific antibody therapy in multiple myeloma with an emphasis on patients who received both modalities sequentially. They find a clear progression-free survival advantage when CAR T therapy was used first; importantly, because outcomes with both modalities were similar when following progression after the converse modality, this advantage was driven by CAR T therapy's up-front benefits. See related article by Merz et al, p. 697.
Relapsed or refractory CNS lymphoma (R/R CNSL) has limited treatment options. This multicenter retrospective study enrolled 84 consecutive R/R CNSL patients (median age 59 years; 53 PCNSL, 31 SCNSL) to evaluate efficacy and safety profiles of glofitamab therapy. With a median of 2.5 prior lines of therapy, the objective response rates (ORR) and complete response rates (CRR) for PCNSL were 88% (CRR 59%) with glofitamab monotherapy (n=32) and 100% (CRR 81%) with combination therapy (n=21), respectively; for SCNSL, ORR and CRR were 88% (CRR 75%) with glofitamab monotherapy (n=8) and 91% (CRR 65%) with combination therapy (n=23). At 14.7 months follow-up, median progression-free survival was 19.5 months for PCNSL and 13.5 months for SCNSL. Cytokine release syndrome occurred in 40% (all grade 1-2) of patients, and ICANS in 8%. Glofitamab-based therapy demonstrated substantial activity with manageable toxicity in this study, offering a promising treatment paradigm for R/R CNSL patients.
Chimeric antigen receptor (CAR) T-cell therapy is increasingly utilized with expanding indications beyond hematologic malignancies. Here, we review existing models developed for predicting toxicities in the CAR T-cell setting and identify both strengths and challenges emerging with their application. Predictive modeling approaches offer potential to guide risk stratification and inform clinical decision-making, but small sample sizes, overfitting, and poor data quality have limited model reproducibility and widespread adoption. As utilization of CAR T-cell therapy broadens, identifying additional biomarkers, developing context-specific models, standardizing guidelines for emerging toxicities, and leveraging federated learning to promote collaborative data sharing will be critical. SIGNIFICANCE:Predictive models integrating biomarkers and clinical variables are increasingly used to forecast potential toxicities after CAR T-cell therapy. However, due to heterogeneity in patient populations and cellular therapy products, the rapidly evolving nature of the field, and continued advancements in management of inflammatory toxicities, modeling in CAR T-cell therapy faces significant challenges. This comprehensive review of existing/emerging models serves to delineate components of developing predictive models including discrimination, calibration, biomarker integration, validation, and mitigation of overfitting while highlighting strengths, opportunities for improvement, and future directions applicable to CAR T-cell therapy.
In this issue of Blood Cancer Discovery, Yang and colleagues present a comprehensive repository of patient-derived xenograft models of aggressive large B-cell lymphoma. The publicly available resource links disease specimens from individuals treated with diverse modern therapies (including chimeric antigen receptor T cells) to their high-risk clinical treatment histories and the molecular and transcriptional characteristics of their tumors, enabling the study of tumor cell-intrinsic mechanisms of therapeutic resistance and susceptibility to novel therapies. See related article by Yang et al., p. 829.
Vasu and colleagues report one of the first clinical evaluations of rapidly manufactured trispecific chimeric antigen receptor T cells targeting CD19, CD20, and CD22. The lymphoma efficacy signal and favorable observed safety profile support further investigation, whereas the correlative analyses suggest that baseline T-cell fitness and early post-infusion functional states may remain important determinants of outcome. See related article by Vasu et al., p. 711.
Abstract In Classical Hodgkin lymphoma (cHL), while most patients are cured with front-line chemotherapy, 10-20% of patients will relapse or develop refractory (R/R) disease. Clinical trials investigating CAR-T cells targeting CD30+ Hodgkin Reed-Sternberg cells (HRS) showed a 57% overall response rate, but 1-yr progression free survival was only 39% indicating that a majority of the patients will need additional treatment. The malignant HRS cells represent 1-5% of cells with most of the surrounding tumor-microenvironment (TME) comprised of infiltrating immune cells: T cells, B cells, dendritic cells (DCs), myeloid derived suppressive cells (MDSCs), and tumor associated macrophages (TAMs). Patients with enrichment of immunosuppressive TAMs in the TME have worse outcomes and shorter survival. We hypothesize that engineering CAR-T cell therapy to target the HRS cells and the immunosuppressive TAMs in the TME will result in deeper and more durable responses for R/R cHL. CD74 is a transmembrane protein that facilitates antigen presentation and promotes pro-survival signaling in B cells. We have previously developed and optimized a CAR targeting CD74 with signaling domains of 4-1BB and CD3ζ. Methods: This study utilizes 5 cHL cell lines, patient samples, and a cell line derived xenograft (CDX) of KM-H2. A tissue microarray was used to evaluate CD74 expression by immunohistochemistry and multiplex immunofluorescence (n=45). Spectral flow cytometry was used to confirm CD74 expression in 6 patient tumor biopsies and 4 bone marrow samples, with comparisons to healthy donor peripheral blood and bone marrow. Cytotoxic cytokines were measured by ELISA and specific lysis of target cells was measured by AK release and flow cytometry assays. Results: We show that CD74 is expressed by 4 of 5 cHL cell lines and by HRS cells in 45 patient samples by TMA regardless of the cHL subtype. When co-cultured with CD74-antigen+ cell lines, CD74-CART cells proliferate and release significantly more TNF-α and IFN-γ, correlating to specific lysis of target cells (p<0.0001). Preliminary results from a KM-H2 engrafted CDX, treated with 5 million CD74-CART cells or control cells intravenously, showed significant decrease in tumor volume and bioluminescence of tumor cells (p<0.001 and p<0.05). Flow cytometry confirmed CD74-CART cell expansion in peripheral blood and homing to SQ tumors. Analyzing the cHL TME by spectral flow cytometry confirmed elevated CD74 on HRS cells, TAMs, DCs, B cells, and monocytic-MDSCs; of which expressed significantly more CD74 than matched cell types in patient or healthy bone marrow and peripheral blood (p<0.0001). CD74-CART cells co-cultured with patient samples, show specific lysis of B cells (29%), immunosuppressive TAMs (73%), M-MDSC (58%), and HRS cells (33%) at 24 hrs. Future experiments to model the TME will incorporate cHL engrafted CD34-humanized NSG mice with comparisons made to CD30-CARTs. Conclusion: CD74 is expressed on HRS cells and immunosuppressive TAMs in the TME. CAR-T cells directed at CD74 show therapeutic potential for the treatment of cHL. Citation Format: Shelby L. Sloan, Danny Kim, Nile Rizvi, Betsy Pray, Shirsha Koirala, Shamama Nishat, Ian Hout, Brian Barta, Xiang Cheng, Walter Hanel, Lapo Alinari, Wing Keung Chan. CD74 as a novel chimeric antigen receptor (CAR)-T cell target for cHL and the TME [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr A003.
Abstract Pharmacological targeting of Bruton’s tyrosine kinase (BTK) has become one of the most successful therapeutic strategies introduced in recent years for patients with B-cell lymphoid malignancies. Depending on the country and indication, six BTK inhibitors — ibrutinib, zanubrutinib, acalabrutinib, orelabrutinib, tirabrutinib, and pirtobrutinib — have already been approved for the treatment of mature B-cell neoplasms, including chronic lymphocytic leukemia, mantle cell lymphoma, follicular lymphoma, marginal zone lymphoma, Waldenström’s macroglobulinemia, and primary central nervous system lymphoma. More recently, BTK degraders have emerged as a new generation of BTK-targeting agents. These bifunctional molecules induce BTK degradation by exploiting the endogenous ubiquitin–proteasome system. Also referred to as PROTACs (proteolysis-targeting chimeras), and in some cases as CDACs (chimeric degradation-activating compounds), BTK degraders are composed of three functional elements: a BTK-binding moiety, an E3 ubiquitin ligase–recruiting moiety, most commonly engaging CRBN or VHL, and a linker connecting the two. This strategy belongs to the broader field of proximity pharmacology, an increasingly important area in anticancer drug development. Several BTK degraders are currently undergoing clinical evaluation, including BGB-16673, bexobrutideg/NX-5948, zelebrudomide/NX-2127, AC676, ABBV-101, HSK29116, and HZ-Q1070, with BGB-16673 and bexobrutideg already being evaluated in phase 3 trials. The extensive clinical use of BTK inhibitors has revealed a wide spectrum of resistance mechanisms. In chronic lymphocytic leukemia, resistance is frequently associated with acquired mutations in BTK or PLCG2, whereas in other diseases, such as mantle cell lymphoma and marginal zone lymphoma, resistance more often involves signaling rewiring, pathway bypass, and broader cellular reprogramming. Importantly, individual BTK mutations may confer resistance to multiple inhibitors, while others appear to be preferentially selected by specific compounds, reflecting differences in drug–target binding and pharmacological properties. Clinical data on resistance to BTK degraders remain limited. Available evidence indicates that the BTK A428D mutation can confer resistance not only to BTK inhibitors, but also to several BTK degraders, including BGB-16673, zelebrudomide, ABBV-101, and bexobrutideg. In addition, mechanisms that bypass BTK-dependent signaling are expected to reduce the efficacy of BTK degraders, as they do for BTK inhibitors. Resistance may also arise through alterations affecting the degradation machinery itself, including genetic or epigenetic lesions involving components of the ubiquitin–proteasome system or the recruited E3 ligase complex. This presentation will review the emerging landscape of resistance to BTK degraders, integrating lessons learned from BTK inhibitors with the distinct vulnerabilities and potential escape mechanisms associated with targeted protein degradation. Citation Format: Francesco Bertoni. Resistance to BTK degraders [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr IA006.
Abstract The germinal center (GC) reaction is a tightly coordinated multicellular system in which B cells, T follicular helper (TFH) cells, follicular dendritic cells (FDCs), and macrophages orchestrate selection, affinity maturation, and clonal elimination. GC B cells give rise to the majority of non-Hodgkin lymphomas, yet how oncogenic epigenetic mutations subvert this system, not merely by conferring cell-intrinsic advantages, but by dismantling the GC architecture itself, remains poorly understood. Using intravital imaging in genetic mouse models, single-cell behavioral and transcriptional profiling, and spatial proteomics of human lymphoma biopsies, we show that oncogenic gain-of-function mutation in the histone methyltransferase EZH2 initiates a progressive, systems-level corruption of GC homeostasis that remodels every cellular compartment of the microenvironment. The epigenetic lesion first reprograms GC B cell behavior in vivo, increasing motility and morphological plasticity while redirecting migration toward FDC-rich light zone subregions. Mutant B cells maintain normal FDC engagement but shortened interactions and reduced surface engagement with TFH cells, requiring prior FDC contact before TFH interaction and impairing dark zone recycling. Transcriptionally, these cells upregulate metabolic and antigen-presentation programs while suppressing cell-death pathways, early signatures of survival fitness. This behavioral reprogramming cascades into broader niche corruption. Epigenetically altered B cells increase contacts with macrophages, converting them from active phagocytes into lymphoma-supporting cells with stellate morphology, loss of phagocytic identity, and an SPP1+MIF+BAFF+ program. Reprogrammed macrophages fail to perform trogocytosis of FDC membranes, driving FDC network expansion and increased immune complex deposition, while delivering pro-survival signals back to B cells via MIF, BAFF, and ICAM1-LFA1 engagement, completing a feedforward circuit reinforced at the lymphoma stage by BCL2. The result is a GC in which affinity selection thresholds are lowered, B cell survival is uncoupled from T cell help, and clonal elimination is suppressed. Macrophage depletion suppresses both pre-malignant expansion and established lymphoma, and spatial proteomics of human DLBCL and FL confirms enhanced macrophage–B cell proximity and dominant ICAM1-LFA1 signaling in epigenetically driven tumors. Therapeutically, macrophage reprogramming explains the failure of anti-CD47 monotherapy: the phagocytic machinery is silenced before the "don't eat me" signal becomes relevant. EZH2 inhibition restores phagocytic function, re-sensitizing tumors to anti-CD47 blockade in vivo. These findings reframe GC-derived lymphomagenesis as an emergent systems failure in which a single epigenetic lesion sequentially dismantles multicellular homeostasis, demonstrating that therapeutic design must target the ordered intercellular dependencies through which the aberrant niche is constructed. Citation Format: Wendy Béguelin. Epigenetic regulation of the lymphoma microenvironment: from single-cell behavioral reprogramming to systems-level niche corruption [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr IA008.
Abstract Background. Olfactory receptors (ORs) are a large subgroup of chemosensors within the G protein-coupled receptor (GPCR) family and have been found ectopically throughout the human body, where they fulfil a range of functions and are associated with the pathogenesis of numerous diseases. While certain ORs have been linked to cancer development, evidence regarding their role in hematological diseases remains scarce. OR13A1 is part of the double-hit expression signature (HGBL-DH/TH) in diffuse large B-cell lymphoma (DLBCL) (Ennishi et al. 2019), is more expressed in germinal center B-cell (GCB) DLBCL where it sustains lymphoma cells’ growth (Sartori et al, 2022). Here, we quantified the ORs expression in lymphoma clinical specimens and performed functional experiments in cell lines. Methods. Gene expression data for clinical samples were obtained from the public dataset phs001444.v2.p1. Patients were stratified according to OR expression groups. Kaplan–Meier survival analysis was performed to evaluate differences in outcome between groups, and LIMMA was used to identify OR-associated gene expression signatures. OR’s expression was first evaluated using a previously generated total RNA-Seq dataset from 47 B-cell lymphoma cell lines and then was assessed by qPCR across five human cell lines: two GCB-DLBCL (OCI-Ly7 and VAL), two activated B-cell-like (ABC)-DLBCL (U2932 and HBL1), and one mantle cell lymphoma (MCL) (JEKO1). OR2B6 silencing was performed in HBL1 and JEKO1 cells using doxycycline-inducible shRNAs, and viability was monitored via live imaging. Results. OR13A1, OR2B6, and OR51E2 were expressed in at least 50% of patients, but only OR13A1 and OR2B6 were detected in RNASeq data from lymphoma cell lines. The lack of OR51E2 in cell lines was suggestive of its expression by cells of the tumor microenvironment (TME), such as macrophages, as recently reported in prostate cancer (Marelli et al, 2025). As expected, OR13A1 was associated with HGBL (DH/TH)-related genes and germinal center B cell-like signatures. OR2B6 was associated with ABC-DLBCL genes and down-regulated HGBL (DH/TH) genes. By qPCR OR13A1 was predominantly expressed in GCB-DLBCL (2-ΔCt = 0.015), moderately in MCL (2-ΔCt = 0.007) and almost absent in ABC-DLBCL (2-ΔCt < 0.0004). In contrast, OR2B6 was mostly expressed in ABC-DLBCL (2-ΔCt = 0.004) and MCL (2-ΔCt = 0.002), with negligible expression in GCB-DLBCL (2-ΔCt < 0.0004). While OR13A1 had already been identified in our lab as an essential gene in GCB-DLBCL, OR2B6 was validated as essential in HBL1 and JEKO1 cell lines derived from ABC and MCL, respectively. Conclusion. These findings suggest that three chemosensors from the OR family may play a role in lymphoma development: OR13A1 (GCB-DLBCL), OR2B6 (ABC-DLBCL and MCL), expressed by lymphoma cells, and OR51E2, expressed by the TME. Further research is needed to identify their ligands, understand their biological mechanisms, and evaluate their therapeutic potential. Citation Format: Giulio Sartori, Sara Sabatini, Alessandro Ghiringhelli, Natalie Amberg, Luciano Cascione, Jacopo Sgrignani, Hiroaki Matsunami, Andrea Cavalli, Francesco Bertoni. OR2B6 is a chemosensor associated with activated B-cell like diffuse large B cell and mantle cell lymphoma [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr A014.
Abstract Background: Glofitamab (Glofit) is a CD20xCD3 bispecific antibody approved for the treatment of patients (pts) with R/R LBCL. Polatuzumab vedotin (Pola) is a CD79b antibody-drug conjugate also approved for the treatment of LBCL. Preliminary data from an open-label, multicenter, Phase Ib study demonstrate that Glofit-Pola has a manageable safety profile and encouraging efficacy in pts with R/R LBCL after >1 prior lines of therapy (Hutchings et al. 2025). Here, we present an exploratory biomarker analysis to determine the biological contribution of adding Pola to Glofit. Methods: R/R LBCL (DLBCL NOS, HGBCL, trFL) pts treated with Glofit-Pola enrolled in NCT03533283 or Glofit monotherapy (mono) pts treated with a target dose (30mg) or efficacious dose (>10mg) enrolled in NCT03075696 were included in this analysis. In pts with available biomarker samples, peripheral T cells were evaluated by flow cytometry. Bulk RNA sequencing (RNAseq) from tumor biopsies (Glofit-Pola n=88; Glofit mono n=147) was analyzed using gene set variation analysis for tumor intrinsic pathways, and xCell for immune and stromal cell types. LymphoMAPs and dark zone signature (DZsig) were derived as previously described (Li et al. 2025, Ennishi et al. 2018). Inverse probability weighting was applied when comparing biomarker-evaluable populations from Glofit+Pola and Glofit mono studies. Results: In pts who received the Glofit-Pola combination, high complete response (CR) rates were observed across molecular subtypes ABC (73.1%), DZsig+ (53.6%) and GCB DZsig- (57.7%). Similar CR rates were also observed across LymphoMAP archetypes Lymph Node (LN, 70.4%), T exhausted (Tex, 59.1%) and Fibroblast/Macrophage (FMAC, 59.0%).To gain insights into the biological contribution of Pola to the Glofit MoA, we compared the association between outcomes and gene signatures related to intrinsic tumor features and tumor microenvironment in pts treated with Glofit-Pola vs Glofit mono, in a weighted and matched population (3L+ LBCL). High proliferation signatures (including E2F and MYC targets) were associated with shorter PFS in Glofit mono (E2F HR 1.58, 95% CI: 1.1-2.3) and longer PFS in Glofit-Pola (E2F HR 0.62, 95% CI: 0.3-1.3). Immune cell deconvolution showed high relative abundance of immune cells (CD4, CD8 T cells) associated with longer PFS in Glofit mono (CD4 HR 0.46, 95% CI: 0.3-0.8), however this association was not observed with Glofit-Pola (CD4 HR 1.60, 95% CI: 0.6-4.0). Additionally, comparison of T-cell pharmacodynamics in the periphery showed increased activation (HLA-DR+ CD4 T cells) by cycle 4 of treatment and decreased exhaustion markers (PD1+ CD8 T cells), suggest enhanced activation of T cells by the Glofit-Pola combination compared to Glofit mono. Conclusion: Biomarker analyses in pts with R/R LBCL treated with Glofit-Pola showed improved clinical efficacy in high-risk molecular subgroups. Exploratory results suggest that pts treated with Glofit-Pola may be less dependent on immune contexture and tumor proliferation status for outcomes than those treated with Glofit mono. Citation Format: Wilfred Leung, Malgorzata Nowicka, Anton Belousov, Gila Sellam, Martine Kallemeijn, Anna Sureda, Linda Lundberg, Estefania Mulvihil, Martin Hutchings, Giuseppe Gritti, Alessia Bottos. Uncovering the biological contribution of polatuzumab vedotin when combined with glofitamab in R/R large B-cell lymphoma [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr A026.
Abstract Background: Canine nodal marginal zone lymphoma (MZL) is the second most common histologic diagnosis of diffuse B-cell lymphomas in dogs, but a molecular profile that can distinguish this condition from the more common canine diffuse large B-cell lymphoma (DLBCL) has been elusive. Here, we use single-cell sequencing to define the transcriptomic signatures of canine MZL before and after oncolytic virotherapy followed by traditional chemotherapy. Objective: The aims of this study were to define the canine MZL transcriptome at single cell resolution and to evaluate its changes in the face of selective pressures introduced by therapy. Methods: Longitudinal biopsies and fine needle aspirates were collected from a dog with MZL enrolled in an immunotherapy trial. Peripheral blood B-cells from healthy dogs served as reference controls. Single-cell sequencing was done using the 10x Genomics Chromium GEM X platform (v3.1), and data was aligned to the canFam4_Y genome. Quality control, clustering, and cell annotation were performed using Seurat in R (v4.4.0). Differential gene expression analysis (DGEA) was performed on 500 randomized MZL B-cells and 500 healthy canine B-cells. InferCNV was deployed on the longitudinal MZL samples for prediction of tumor B-cell CNV profiles. Results: When compared to normal peripheral blood B-cells, the tumor cells were characterized by downregulation of quiescence-associated genes, including KLF2. Copy number analysis uncovered loss on canine chromosome 20 (CFA20) encoding KLF2, gain on CFA13, and gains on CFA32 and CFA36. While an anti-viral T-cell response was identified after VSV treatment, the B-cell subpopulations retained stable transcriptomes, with a disappearance of proliferating B-cells after chemotherapy. Conclusion: This is the first report describing the molecular properties of canine MZL at single cell resolution. Clonal diversity was apparent in the tumor, with copy number loss on CFA20, presumably leading to loss of KLF2. The transcriptional features of the tumor cells remained stable even in the face of a T-cell response to an oncolytic virus. Taken together, the results document silencing of KLF2 as a key phenotypic homology between canine and human MZL, supporting the potential for comparative approaches to enhance our understanding of the significance of KLF2 silencing in this disease. Citation Format: Elise Weir, Grace Walker, Jaime F. Modiano, Aaron L. Sarver, Shruthi Naik, Amy Treeful. Reduced KLF2 expression in canine marginal zone lymphoma mirrors human disease [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr A043.
Abstract Human peripheral T cell lymphoma, not otherwise specified (PTCL-NOS) is a rare and aggressive subset of non-Hodgkin’s lymphoma with poor survival. Pet dogs develop a similar neoplasm, termed peripheral T cell lymphoma (PTCL), that shares histology, antigen expression and gene expression profiles with human PTCL-NOS. These dogs are an attractive model for the study of PTCL-NOS because they develop spontaneous disease in an immune competent host, live in the same environment as humans, have relatively short lifespans allowing for lifelong follow up and are treated with the same anthracycline based chemotherapeutics (CHOP) with outcomes similar to their human counterparts. In addition, human PTCL-NOS has been tentatively subdivided based on the expression of the transcription factors TBX21 and GATA3. According to previous RNA sequencing data from our lab, canine PTCL resembles the GATA3 subtype of PTCL-NOS. The disease in pet dogs is more frequent than seen in people; it is the second most common form of lymphoma in this species. The purpose of this study was to characterize the mutational profile of canine peripheral T cell lymphoma and establish a pre-clinical model of disease using patient canine derived xenograft models (PDX). We established the mutational profile of canine PTCL using whole exome sequencing (WES). DNA was isolated from 95 canine samples previously used for RNA sequencing. This DNA was then used for WES and analyzed for cancer-related mutations using GATK best practices. Top mutated genes included FAT3, KMT2D and mTOR. These mutations are involved in tumor suppression, epigenetic modification, and cell growth and proliferation respectively. The mTOR mutations occur within the catalytic domain of the protein and are consistent with previous RNAseq data from our lab showing upregulation of the PI3K-AKT-mTOR pathway. Several mutations (14/26) align with known human mutations in location and type. To establish an in vivo pre-clinical model of PTCL, primary canine tumor cells were injected into NSG mice intraperitoneally. Mice were monitored for tumor development through serial weights, abdominal palpation and observation. At the first sign of disease progression, mice were euthanized and a comprehensive necropsy was performed to identify gross and microscopic disease. Engraftment of primary cells was successful in 9/23 (39.1%) mice. After first passage, engraftment was successful in 4/6 (67%) mice. Across passages, the histologic and immunophenotypic characteristics of the tumors remained consistent. These results show that the PDX model was successfully established, and characteristics were maintained through multiple passages. Overall, these data show that pet dogs are an excellent model to study PTCL-NOS and that PDX models can be successfully established using canine samples. Future directions include testing novel therapies in PDX mice before moving them to pre-clinical trials in pet dogs. Citation Format: Sara Cook, Eileen Owens, Kenzie Olsen, Tobey Autele-Acera, Adam Harris, Anne Avery. Peripheral T cell lymphoma not otherwise specified: Characterization of pet dogs as a model organism [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr A021.
Abstract Chronic B-cell receptor (BCR) signaling is a central driver of aggressive B-cell lymphomas, reinforcing oncogenic survival and proliferation programs; yet the mechanisms by which these signaling outputs are sustained at the level of protein synthesis remain to be fully elucidated. We hypothesized that selective mRNA translation mediated by eIF4A1 acts as a key regulator of BCR-driven oncogenic programs and represents a therapeutically actionable vulnerability. The RNA helicase eIF4A1 is a core component of cap-dependent translation initiation that facilitates ribosome loading on transcripts with structured 5′ untranslated regions. Analysis of lymphoma datasets and primary tumor samples demonstrates that eIF4A1 expression is elevated in diffuse large B-cell lymphoma relative to normal B cells and is associated with adverse clinical outcomes, supporting its disease relevance. Consistent with this expression pattern and its role in regulating translation of structured mRNAs, eIF4A1 activity sustains the expression of key oncogenic drivers, including MYC and survival-associated programs linked to BCR/NF-κB signaling, positioning it as a link between oncogenic signaling and protein synthesis. Supporting this role, functional genomic analyses further indicate that lymphoma cells rely on eIF4A1 activity to maintain oncogenic protein expression. To exploit this therapeutic vulnerability, we developed RocA-independent small-molecule inhibitors of eIF4A1, RBF197, and RBF208 that impair helicase-mediated unwinding of structured mRNAs. Pharmacologic inhibition of eIF4A1 suppresses lymphoma cell proliferation and clonogenicity, accompanied by reduced expression of oncogenic drivers including MYC, BCL2, and CCND1, while maintaining relative selectivity for malignant B cells. In orthotopic lymphoma models, RBF197 achieves target engagement and induces significant tumor growth inhibition with favorable pharmacokinetic properties. To further link these observations to oncogenic signaling networks, we evaluated independent models of B-cell activation and lymphoma, where BCR engagement is associated with enhanced translational output and enrichment of eIF4A1-responsive programs, whereas inhibition of eIF4A1 produces reciprocal suppression of these gene signatures. Combinatorial profiling across nodes of the BCR signaling cascade reveals pathway-selective interactions with eIF4A1 inhibition, indicating functional convergence between oncogenic signaling and translational control. Complementary perturbation studies further support functional interplay, whereas BCR activation augments protein translation and inhibition of BCR-associated signaling attenuates eIF4A1-dependent translational output. Ongoing ribosome profiling studies aim to refine these eIF4A1-dependent translational programs. Collectively, these findings position eIF4A1 as a key regulator linking oncogenic signaling to protein synthesis in aggressive lymphoma and highlight the BCR-translation axis as a therapeutically actionable vulnerability. Citation Format: Forum Kayastha, Bandish Kapadia, Anirban Roychowdhury, Nahid M. Nanaji, Won Sok . Lee, Ronald B. Gartenhaus. Targeting eIF4A1 reveals crosstalk between BCR signaling and translation in aggressive B-cell lymphoma [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr A028.
Abstract Background Plasmablastic lymphoma (PBL) and related Diffuse Large B-cell Lymphoma (DLBCL) variants with plasmablastic differentiation are aggressive malignancies characterized by high BCMA (TNFRSF17) expression, yet BCMA-targeted therapies remain largely unexplored in this setting despite poor outcomes with standard approaches. These tumors frequently present with extramedullary disease, poor vascularity, and profound immunosuppression – features that reduce the efficacy of Fc-dependent biologics and underscore the need for Fc-independent targeting approaches. Aim We aimed to identify small BCMA-binding peptide ligands (≈1–5 kDa) for theranostic applications in plasmablastic lymphomas. Our strategy focused on targeting the native BCMA receptor on intact tumor cells to preserve its physiological conformation, including its cysteine-rich ectodomain and membrane-associated lipid environment, which is critical for genuine epitope recognition and binding interactions. Methods We performed phage display screening to identify BCMA-binding peptides using large combinatorial peptide libraries displayed on M13 phage, through three iterative rounds of positive selection on BCMA-positive NCI-H929 cells and negative selection on induced BCMA-negative NCI-H929 cells. We subsequently performed in silico characterization of the binding mechanisms of selected peptides, including protein-peptide docking analyses using the HADDOCK server and binding free energy estimation with the PRODIGY tool. Results We identified three peptide scaffolds capable of binding BCMA in its native form on BCMA-positive cells. Peptide pBCMA_7 was selected because of its clonal prevalence and its binding affinity toward BCMA-positive tumor cells. Docking and free energy studies suggest a configuration in which the peptide associates with a surface-accessible region of BCMA, forming extensive interactions through both backbone and side-chain atoms. Molecular dynamics simulations further indicate a stable and consistent binding mode within an exposed pocket of the receptor. Notably, pBCMA_7 exhibited strong binding (-11.2 kcal/mol) across a range of plasmablastic lymphoma cell lines (DB, Pfeiffer, SU-DHL-2, U-2932, Karpas 422) as well as multiple myeloma cell lines (NCI-H929, U266 and IM9). Conclusion We report the identification of BCMA-binding peptide ligands with demonstrated activity in plasmablastic lymphomas and related large B-cell lymphomas with plasmablastic differentiation. These peptide scaffolds constitute a versatile theranostic toolkit supporting Fc-independent patient stratification, therapy monitoring and delivery of imaging and therapeutic agents. By establishing phage display against native receptors as a tractable strategy for ligand discovery in an unexplored oncological setting, this work opens a wide therapeutic space for BCMA-directed precision targeting in aggressive lymphomas. Citation Format: Elisabetta Pingitore, Khushboo Fatima, Selena Mimmi, Valentina Crapella, Anna Maria Zimbo, Antonio Lupia, Alessia Onali, Domenico Maisano, Francesco Bertoni, Alessandra Scagliola, Federica Melle, Maria Carmela Vegliante, Doriana Gramegna, Michele Guida, Sabino Ciavarella, Enrico Iaccino. Unlocking BCMA in aggressive lymphomas: A native receptor-targeted peptide for theranostic applications [abstract]. In: Proceedings of the Fifth AACR International Meeting on Advances in Malignant Lymphoma: From Discovery to Clinical Impact; 2026 Jun 24-27; Philadelphia, PA. Philadelphia (PA): AACR; Blood Cancer Discov 2026;7(3_Suppl):Abstract nr A034.
