Chimeric antigen receptor T (CART) cell therapy has demonstrated clinical efficacy in hematologic malignancies; however, primary or secondary treatment failure remains a major obstacle to durable responses. Defining the optimal cellular composition of CART products therefore represents a critical unmet need. Here, we show that CARTs targeting acute leukemias achieve maximal efficacy when composed exclusively of CD4+ T cells. Pure CD4+ CARTs exhibit superior anti-tumor activity and proliferation compared with CD8+-containing CART products. To elucidate the molecular basis of this functional divergence, we applied a combinatorial exploratory approach integrating bulk RNA sequencing and quantitative proteomics. Transcriptomic analyses revealed that pure CD4+ CARTs adopt a highly proliferative state characterized by a cytotoxic effector-like polarization. On the protein level, we are demonstrating coordinated loss of cell-cycle machinery and induction of apoptotic pathways in CD4+ CART cells co-cultured together with CD8+ CART cells, while pure CD4+ CART cultures maintained a proliferative, cytotoxic phenotype. Using mechanistically discriminative co-culture systems, we demonstrate that CD8+ CART-mediated impairment of CD4+ CART functionality is primarily driven by competitive access to shared antigen. Collectively, these findings identify antigen competition between CART subsets as a previously unrecognized mechanism limiting CD4+ CART efficacy and provide a framework for optimizing CART product composition to enhance therapeutic persistence and durability.

SnoRNAs are highly expressed in AML and have implications in leukemogenesis and leukemic maintenance. SnoRNAs can be further processed into snoRNA-derived RNAs (sdRNAs). The role of sdRNAs in AML and healthy hematopoiesis remains largely elusive. We characterized sdRNA and snoRNA levels in hematopoietic stem and progenitor cells (HSPCs), healthy WBCs, and 159 intensively treated AML patient samples at initial diagnosis. HSPCs, healthy WBCs, and AML blasts could be differentiated by their sdRNA expression pattern in a cell-type-specific manner. In AML, high sd3’-RNA/snoRNA-host gene ratios were associated with an inverse patient outcome. Particularly, in NPM1-mutated patients with favorable risk stratification and good initial therapy response, high sd3’-RNA ratios identified a subgroup with inferior outcome. High sd3’-RNA ratios were associated with altered oncogenic, inflammatory, and immune response signaling. Forced expression of single sdRNAs, such as sd3’-SNORD78, sd3’-SNORD76, and sd5’-SNORD93, enhanced clonogenic potential in AML and drove sdRNA-specific gene expression signatures in both AML and healthy HSPCs. Exemplarily, we propose and characterize NUDT21, an important regulator of alternative polyadenylation and oncogenic gene expression, as a downstream target of sd3’-SNORD78 in AML. Our data introduce sdRNAs as standalone regulatory effector molecules in healthy hematopoiesis and AML.

Bleximenib, a potent and selective menin inhibitor, in combination with intensive chemotherapy (IC) has previously exhibited an acceptable safety profile and efficacy in participants (pts) with newly diagnosed (ND) NPM1-mutated (NPM1m) or KMT2A-rearranged (KMT2Ar) acute myeloid leukemia (AML). We report updated safety and efficacy data (cut-off: July 2025) from this study.In the ALE1002 Phase 1b, multicenter, dose-finding study (NCT05453903), pts received a ‘7+3’ regimen of cytarabine 200 mg/m2/day and daunorubicin 60 mg/m2/day or idarubicin 12 mg/m2/day in combination with bleximenib 30–100 mg twice daily (from Day 4 of induction, including during count recovery). Pts who achieved a complete remission (CR) received consolidation therapy with up to 4 cycles of intermediate-dose cytarabine plus bleximenib. Those not proceeding to allogeneic hematopoietic stem cell transplant could continue bleximenib for up to 12 months. The safety dataset included all pts who received bleximenib 30–100 mg, and the intention-to-treat (ITT) efficacy dataset included those who received bleximenib 100 mg. Investigators assessed response criteria according to European LeukemiaNet (ELN) recommendations.The safety analysis set included 44 pts with ND AML (median age, 57.0 years [range, 19–71]; 52.3% female; 56.8% NPM1m, 43.2% KMT2A; 15.9% FLT3 co-mutations; ELN risk classification: 44.2% favorable, 27.9% intermediate, 27.9% adverse). The median duration of follow-up was 6.3 months (range, 1.31–22.51). All pts had ≥1 treatment-emergent adverse event (TEAE, all grades), the most common being thrombocytopenia (35/44; 79.5%), neutropenia (32/44; 72.7%), diarrhea (31/44; 70.5%), nausea (30/44; 68.2%), anemia, and febrile neutropenia (both 28/44; 63.6%). Most cytopenia TEAEs were Grade 3/4, consistent with an IC backbone. The 30- and 60-day mortality was 0/44 (0%) and 1/44 (2.3%), respectively. No differentiation syndrome was observed. Three TEAEs of QT prolongation were reported, all of which were Grade 1/2 and resolved without bleximenib interruption. Of 24 pts (NPM1m, n=15; KMT2Ar, n=9) in the ITT efficacy dataset, overall response rate (≥partial response) was 95.8%, composite CR (CR + CR with partial hematologic recovery [CRh] + CR with incomplete hematologic recovery rate) was 87.5%, and CR/CRh was 75%. Responses were similar across mutational subtypes. Median (range) time to CR was 28 days (21–36), similar to the median time to first response. Median duration of response was not reached. Among 37 pts achieving composite CR the median (range) time from Day 1 of induction to platelet count recovery was 32.0 days (22.0–82.0), and the median time to neutrophil count recovery was 30.0 days (21.0–71.0).In ND NPM1m or KMT2Ar AML, the safety profile of bleximenib + ‘7+3’ was consistent with a ‘7+3’ IC backbone, which, combined with promising early efficacy data, supports the planned Phase 3 study.

T cell-mediated immune surveillance is critical for cancer control, yet its role in bone marrow malignancies remains poorly understood. Here, we integrate TCR profiling, HLA immunopeptidomics, and functional screening to characterize tumor-reactive T cells in the bone marrow of patients with multiple myeloma (MM) and acute myeloid leukemia (AML). These cells are transcriptionally defined by a conserved effector program distinct from the exhausted phenotype of tumor-reactive T cells in solid cancers. Immunopeptidomic profiling reveals a partially shared antigen landscape enriched for noncanonical peptides driving convergent TCR responses. We develop TFiT (tumor-reactive features in T cells), a transcriptional classifier that identifies these cells and stratifies immunotherapy, but not chemotherapy, response across independent MM and AML cohorts, supporting its specificity for T cell-mediated tumor control. These findings reveal a latent but activatable anti-tumor T cell compartment in bone marrow malignancies and provide a framework for engaging endogenous immunity in MM and AML.

BACKGROUND:Chimeric antigen receptor (CAR) T-cell therapy has transformed the treatment of hematologic malignancies, but its success depends on obtaining sufficient CD3+ T-cell yields during leukapheresis. This can be difficult in heavily pretreated patients, who often show leukopenia and reduced T-cell fitness. METHODS:We analyzed 166 leukapheresis products from 154 patients undergoing manufacturing of CD19-directed CAR T-cells, including 146 from non-Hodgkin's lymphoma (NHL) and 20 from acute lymphoblastic leukemia (ALL). Collections were performed for commercial CAR T-cell products (axi-cel, tisa-cel, brexu-cel, liso-cel; n = 121) and the HD-CAR-1 trial with in-house manufacturing (heidagenlecleucel; n = 45). RESULTS:In 150/154 patients, a sufficient CD3+ T-cell yield was achieved by a single leukapheresis. Pre-apheresis lymphocyte count strongly predicted CD3+ T-cell yield (p< .001) and was associated with treatment response (p = .044). Impaired hematopoiesis, reflected by reduced nucleated cell count (p< .001), lymphocyte count (p <.001), and hematocrit (p = .017), was linked to poorer collection efficiency. Intensive prior therapies, including stem cell transplantation, reduced CAR T-cell expansion during manufacturing (p = .036). In patients with ALL, higher proportions of effector and CD8+ CAR T-cells in the product correlated with improved clinical outcomes (p = .04; p = .023). CONCLUSION:These findings highlight the importance of early leukapheresis, ideally before intensive treatments, to optimize T-cell yield, product quality, and therapeutic efficacy.

Anthracyclines are an essential component of induction therapy for acute myeloid leukemia (AML), but their optimal dosing and the associated risk for cardiotoxicity remain under debate. The DaunoDouble trial compared daunorubicin at 60 (Dauno60) versus 90 mg/m2 (Dauno90) in combination with cytarabine (100 mg/m2 for 7 days) in newly diagnosed AML patients aged 18-65 years. Cardiac function was assessed by left ventricular ejection fraction (LVEF) and cardiac biomarkers (high-sensitivity troponin T [hsTnT], NT-proBNP) before treatment and on Day 15 in 317 randomized patients. Median LVEF declined significantly from 65% [IQR 60%-68.5%] to 61% [IQR 58%-67.8%] across all patients (p < 0.01), without significant differences between treatment arms. NT-proBNP levels measured after induction therapy correlated negatively with LVEF at the same time point (ρ = -0.24, p = 0.02), but did not change significantly during induction-neither between Day 1 and 15 nor between treatment arms. HsTnT levels increased significantly from 5 [IQR 4-8] to 8 ng/L [IQR 6-14] across all patients (p < 0.01), with higher post-induction values in the Dauno90 group (9.5 ng/L [IQR 7-14]) compared to Dauno60 (7 ng/L [IQR 5-14]; p < 0.01). In exploratory subgroup analyses, post-induction hsTnT levels were also significantly higher in patients with obesity and arterial hypertension. These findings provide evidence of a dose-dependent cardiotoxic effect of daunorubicin, already apparent at standard induction doses, and underscore the importance of early cardiac monitoring. Long-term follow-up will be essential to determine the clinical significance of these early changes. Trial Registration: ClinicalTrials.gov identifier: NCT02140242.

Abstract Background Acute myeloid leukemia (AML) is predominantly a disease of older patients with a poor long-term survival. Approval of the combination of venetoclax (VEN) with azacitidine (AZA) or decitabine (DAC) in the European Union in 2021 for the treatment of patients with newly diagnosed AML who are ineligible for induction chemotherapy has become the treatment backbone for older, medically non-fit AML patients. Based on in vitro priming of AML cells to all- trans retinoic acid (ATRA) by DAC, we had previously conducted a randomized phase II trial (DECIDER, AMLSG 14 − 09) in 200 elderly, non-fit AML patients on the effect of ATRA as add-on to DAC. Median overall survival (OS) was 8.2 months with ATRA versus 5.1 months without ATRA (hazard ratio, 0.65; 95% CI, 0.48 to 0.89; P = 0.006). Notably, ATRA did not add toxicity, the combination was active also in patients with adverse genetics, and prolonged time to treatment resistance. Investigating a triple combination of DAC, VEN and ATRA was a logical next step. This randomized double-blind phase III trial compares the efficacy of ATRA versus placebo as add-on to the backbone treatment (DAC and VEN) with respect to OS, objective best response, quality of life and safety. The accompanying translational research will contribute to identify molecular markers for drug efficacy and better tailoring epigenetic therapy. Methods DECIDER-2 (AMLSG 32 − 21) is a prospective, randomized, double-blind, placebo-controlled, parallel group, multicenter trial. The primary study objective is to compare the efficacy of ATRA versus placebo as add-on to the backbone treatment (DAC and VEN) in terms of OS. The target population is adult AML patients unfit for standard induction chemotherapy. Patients are randomized (1:1) to receive backbone treatment in combination with either ATRA or placebo. Four interim safety analyses were planned with the involvement of the Data Monitoring Committee. Discussion This trial investigates whether the in vitro cooperativity of DAC, VEN, and ATRA prolongs OS in unfit AML patients compared to placebo, offering a novel, low-toxicity triplet combination with a promising risk-benefit ratio. Clinical trial number EU CT number: 2020-005495-36/2023-507461-26-00 (registered 18.11.2020); German clinical trials registry number: DRKS00023646 (registered 23.08.2022).

Cancer is characterized by complex interactions across genetic, cellular, and microenvironmental scales. However, a quantitative understanding of how these interactions shape clinical trajectories remains limited. Here, we present a multi-scale single-cell dataset from 184 treatment-naive acute myeloid leukemia (AML) patients spanning all major genetic subtypes, together with an analytical framework to dissect interactions across biological scales. We show that distinct clinical outcomes are encoded by specific cross-scale, cross-compartment interactions present at diagnosis: response to induction therapy is governed by interactions between genetic alterations and leukemic differentiation state; relapse following chemotherapy is associated with non-genetic programs linked to metabolism; and relapse after allogeneic stem cell transplantation is driven by interactions between the immune microenvironment and residual healthy hematopoiesis. Together, our study provides a framework to resolve intra- and inter-patient heterogeneity in cancer and supports a model in which clinical trajectories in AML emerge from defined interactions across biological scales.

Relapsed or refractory acute myeloid leukemia (R/R AML) remains a therapeutic challenge, with standard high‑dose cytarabine-based salvage regimens achieving complete remission (CR) rates of around 50%. In the phase 1/2 RELAX trial, adding venetoclax (VEN) to high‑dose cytarabine and mitoxantrone (HAM) was considered safe and resulted in CR or CR with incomplete hematologic recovery (CRi) in 75% of patients. Here, we performed a multicenter real‑world analysis of 128 R/R AML patients treated with HAM plus VEN (VEN 400 mg days 1-14, cytarabine 1000 mg/m² IV twice daily days 3-5, mitoxantrone 10 mg/m² IV days 5-7) between March 2023 and June 2025. The median age was 60 years (range, 20-74). Patients had a median of 1 prior therapy (range, 1-5), 38% had prior allogeneic hematopoietic cell transplantation (HCT), and 25% VEN exposure. Overall, 69% achieved CR/CRi; response rates were higher in patients with one prior therapy line (78%) and VEN-naïve patients (76%). Sixty-eight percent proceeded to HCT. After a median follow-up of 11.5 months, estimated 1‑year RFS and OS rates were 59% and 58%, respectively. In a prognostic classifier derived from pooled real-world and RELAX trial data (183 patients), TP53 mutation, complex karyotype, ≥2 prior therapy lines, and Eastern Cooperative Oncology Group performance status (ECOG PS) ≥2 predicted inferior outcomes. Patients without these features had a 1-year OS of 73%. HAM plus VEN retains high efficacy and encouraging survival outcomes in a real-world setting, supporting its use as an effective salvage and bridge-to-transplant strategy.

Multidisciplinary tumor boards integrate longitudinal treatment histories, molecular profiling and rapidly evolving evidence to guide decisions in hematological malignancies, yet access to this level of subspecialty deliberation is increasingly uneven. Here we develop HemaGuide, a locally deployable, modular large language model agent that converts unstructured clinical documents into structured case representations, autonomously routes cases to specialized decision modes ('guideline', 'advanced' and 'molecular') and grounds recommendations in disease-specific guideline flowcharts and a clinical decision memory of >2,000 real-world tumor board cases. In expert-blinded benchmarking on 45 high-complexity cases across six foundation models, HemaGuide substantially improved concordance with tumor board decisions. A systematic ablation study across 11 layers confirmed that performance gains were routing-type-dependent, with no single component sufficient across case types. Automated classification of 70 clinically relevant missense variants showed high concordance with expert standards; no oncogenic variant was downgraded to benign and the whole workflow was completed under real-time conditions on commodity hardware with a median latency of 39 s rather than the hours typically required for manual molecular board workflows. In a simulated practice study, agent-assisted resident physicians achieved near-senior concordance and partially outperformed senior physicians in their subspecialty. External validation on 555 independent cases from a second academic center yielded 81.8% concordance across 47 entities, and a prospective 1-month silent trial on 64 consecutive, unselected cases achieved 82.8% concordance. Hallucinations occurred in 2 of 664 evaluated cases (0.3%). Together these data provide evidence that locally deployable, case-grounded large language model agents can deliver auditable clinical decision support across hematological malignancies, with concordance maintained across institutions and under real-time conditions on commodity hardware.

ABSTRACT Molecular testing in hematology requires different assays for disease subgroup identification, risk stratification and selection of appropriate treatment regimens. Yet, molecular tests are not necessarily standardized between diagnostic laboratories, resulting in varying turnaround times and potentially divergent results. To resolve this issue and enable single-assay molecular testing, we have developed a hierarchical classification framework that combines epigenetic and genetic data from whole genome nanopore sequencing (WGNS) with machine learning to determine disease entities, epigenetic subgroups (epitypes) and genetic aberrations in hematopoietic neoplasms. We curated DNA methylation data from 5,420 samples and trained a classifier allowing entity-level diagnostics featuring 21 conditions, including healthy controls, acute and chronic myeloid and lymphoid neoplasms. This classifier was subsequently combined with entity-specific epitype classifiers predicting 44 therapeutically or prognostically relevant states, followed by integration of genetic data. Benchmarking of the combined (epi-)genetic testing strategy using WGNS confirmed high accuracy in the detection of diagnostic groups and risk stratification, and identified diagnosis-defining molecular alterations that were not reported by standard-of-care work-up.

Prognosis of adult patients with relapsed/refractory (r/r) or measurable residual disease (MRD)–positive B-precursor acute lymphoblastic leukemia (B-ALL) remains poor. Blinatumomab induces complete remissions (CR) and MRD negativity in a subset of patients, yet overall survival remains limited. As BCL2 overexpression contributes to leukemic cell survival and therapy resistance, combining Blinatumomab with the BCL2 inhibitor Venetoclax may enhance therapeutic efficacy. The GMALL-BLIVEN (NCT05182385) phase I multicenter trial evaluated the safety and feasibility of Venetoclax plus Blinatumomab in adults with CD19⁺, Philadelphia chromosome–negative r/r or MRD-positive B-ALL. Dose escalation followed a 3 + 3 design across three Venetoclax dose levels (DL-1: 400 mg; DL-2: 600 mg; DL-3: 800 mg) administered from day − 7 to day 42, with Blinatumomab given per label. The primary endpoint was determination of the maximum tolerated dose (MTD); key secondary endpoint was achievement of molecular complete remission (MOL-CR) by centralized IG/TR MRD assessment (sensitivity ≥ 10⁻⁴). Nine patients (median age 52 years, range 22–71) were enrolled across four German centers: four with r/r B-ALL and five with MRD-positive disease. Molecular subtypes included ZNF384-rearranged (n = 2), BCR::ABL1-like (n = 2), CEBP-rearranged (n = 1), hypodiploidy (n = 1), B-ALL NOS (n = 2), and KMT2A-rearranged (n = 1). No dose-limiting toxicities were observed, and the MTD was not reached. Treatment-emergent grade ≥ III toxicities included neutropenia, febrile neutropenia, cytokine release syndrome, and ICANS; no 30- or 60-day mortality occurred. Six patients completed two full cycles without treatment interruptions. The recommended phase II dose (RP2D) was established as Venetoclax 800 mg daily plus standard-dose Blinatumomab. All nine patients were evaluable for response. Among r/r B-ALL, responses included PR (n = 1), CR (n = 1), and PD (n = 2). Among MRD-positive patients, 4/5 achieved MOL-CR. Four patients were successfully bridged to allogeneic stem cell transplantation. After a median follow-up of 17.8 months, median overall survival was 15.0 months (r/r: 6.6 months; MRD-positive: 16.9 months). Non-responders were enriched for adverse molecular subtypes (BCR::ABL1-like and hypodiploid B-ALL). Venetoclax combined with Blinatumomab is safe and feasible in adults with r/r or MRD-positive B-ALL, without increased rates of CRS or neurotoxicity compared with Blinatumomab alone. High rates of MRD clearance were observed in MRD-positive patients. The phase II portion of GMALL-BLIVEN is ongoing at the RP2D of Venetoclax 800 mg daily.

Patients with relapsed or refractory acute myeloid leukaemia (r/r AML) harbouring FMS-like tyrosine kinase 3 (FLT3) mutations generally have poor prognosis and limited treatment options. The second-generation FLT3 inhibitor gilteritinib provides better disease control compared to standard relapse therapies and was approved based on two randomized trials. We analysed its efficacy and safety in a real-world cohort of r/r AML patients undergoing gilteritinib treatment including allogeneic haematopoietic stem cell transplantation (alloHSCT). Data were collected from all consecutive r/r AML patients with FLT3-internal tandem duplication and/or FLT3-tyrosine kinase domain mutations receiving gilteritinib at 25 centres. Response and survival were analysed using descriptive and correlative statistics. A total of 156 patients were analysed, 55.1% with relapse after previous alloHSCT, and 46.2% with prior FLT3 inhibitor exposure. The overall complete remission/incomplete remission (CR/Cri) rate was 45.3%, with a median overall survival (OS) of 10.0 months. Patients receiving gilteritinib only had a 1-year OS of 34%, while those with consolidating alloHSCT had a 1-year OS of 76%-92%, depending on gilteritinib maintenance post-transplant. Resuming gilteritinib post-transplant led to an 85.7% CR/CRi rate. Our real-world data confirm that gilteritinib is an effective single-agent therapy for r/r AML with FLT3 mutations, with best outcomes when used as a bridge to transplant and post-transplant maintenance.

The BCL-2 inhibitor venetoclax has transformed the treatment of acute myeloid leukemia (AML), but relapse due to resistance of leukemic stem cells (LSCs) remains a major challenge. By molecular and functional profiling of LSCs from >150 patients, we identify four LSC subtypes. These mirror distinct hematopoietic lineage stages, which determine the expression ratio between the venetoclax target BCL-2 and resistance-inducing proteins MCL-1 and BCL-xL (MAC-score). Longitudinal analyses reveal that venetoclax resistance mostly arises in LSCs through plasticity toward a megakaryocytic/erythroid-progenitor (MEP)-LSC state that switches survival dependency from BCL-2 to BCL-xL. In rare cases, mature monocytic/dendritic (MoDe)-LSCs, found within LAMP5+ monocytic AMLs, drive venetoclax resistance. LSC subtyping improves genetic risk stratification and provides subtype-specific therapies: venetoclax-resistant MEP-LSCs respond to BCL-xL inhibitors, whereas MoDe-LSCs are sensitive to MEK1/2 inhibition. Our findings reveal four distinct LSC types with unique vulnerabilities and propose biomarker-guided treatment strategies that complement genetic profiling to overcome venetoclax resistance.

Introduction Patients (pts) with core-binding factor (CBF) AML (defined by inv(16)(p13.1q22) or t(16;16)(p13.1;q22)/CBFB::MYH11 or t(8;21)(q22;q22.1)/RUNX1::RUNX1T1 gene fusions) are considered to have favorable risk disease when treated with intensive chemotherapy regimens. However, since these pts were excluded from pivotal trials evaluating venetoclax in combination with hypomethylating agents (HMA/Ven) or low-dose cytarabine (LDAC/Ven), little is known about the efficacy of these regimens in CBF AML pts ineligible for intensive treatment. Case series suggest that pts with CBFB::MYH11 respond well to HMA/Ven, whereas those with RUNX1::RUNX1T1 may be less responsive (Zhang et al., 2023, Shen et al., 2025). Given the lack of approved and effective therapies for older or frail pts with CBF AML besides HMA/Ven, this study aimed to assess its efficacy in this specific patient population. Methods Based on data from seven European study group registries (PETHEMA, DATAML, CELL, ALFA, NCRI, PALG, and SAL), we retrospectively analyzed adult pts with newly diagnosed CBF AML treated with HMA/Ven. CBF fusions were identified using conventional karyotyping, FISH, PCR and/or NGS. For response evaluation, ELN 2022 criteria were used, and best response within six HMA/Ven cycles is reported. The median overall survival (mOS) was estimated using the Kaplan-Meier method. Statistical analyses were performed using GraphPad Prism v.10.5.0. Results We identified 53 pts with newly diagnosed CBF AML (27 (51%) with CBFB::MYH11 and 26 (49%) with RUNX1::RUNX1T1) treated with HMA/Ven between September 2019 and July 2025. The median age of the entire cohort was 76 years (range, 30-86), and 21 pts (40%) were female. ECOG performance status was ≥2 in 17 pts (35%). Eight pts (20%) had AML post-cytotoxic therapy. According to the European LeukemiaNet (ELN) 2024 risk classification, 70%, 28%, and 2% of pts had favorable, intermediate, and adverse risk disease, respectively. The most common secondary genetic alterations were mutations in NRAS (31%), TET2 (30%), KIT (29%), KRAS (23%) and DNMT3A (14%). Except for a higher KRAS mutation prevalence in pts with CBFB::MYH11 (44% vs. 5%; p=0.051), there were no significant differences in baseline characteristics between CBFB::MYH11 or RUNX1::RUNX1T1 subgroups. None of the pts had previously been exposed to HMA or Ven, and most pts (95%) received Ven in combination with azacitidine. The median planned Ven schedule (cycle one) was 21 days (range, 7-28), and the median number of administered HMA/Ven cycles was 4 (range, 1-24). Two pts underwent subsequent allogeneic hemopoietic cell transplantation. The composite complete remission rate (CRc, CR + CRi) of the entire cohort was 69%, and was similar for pts with CBFB::MYH11 orRUNX1::RUNX1T1 fusions(67% vs. 72%,p=0.564). Pts with ELN 2024 intermediate risk mutations (i.e., FLT3-ITD, NRAS and/or KRAS) had a CRc rate similar to those without any of these variants (68% vs. 73%,p=0.752). Also, the CRc rate of pts with KIT mutations was comparable to wild-type pts (68% vs 79%,p=0.29). The 60-day mortality rate was 8%. After a median follow-up of 14 months, the mOS of the entire cohort was 13 months, with corresponding 12-month and 24-month OS rates of 54% and 39%, respectively. The mOS of pts with RUNX1::RUNX1T1 was similar to those with CBFB::MYH11 (12 months vs. 14 months, p=0.447). As compared to the corresponding wild-type pts, neither ELN 2024 intermediate risk variants nor KIT mutations adversely impacted mOS (FLT3-ITD/NRAS/KRAS, 11 months vs. NR, p=0.191; KIT, 14 months vs. 12 months, p=0.681). Conclusions Based on our analysis in a larger cohort of international pts, treatment with HMA/Ven seems to be a valuable treatment option for CBF AML pts ineligible for intensive chemotherapy. In contrast to previous studies in mostly Chinese cohorts, we did not observe a significant difference in response or survival between pts with AML with CBFB::MYH11 and those with RUNX1::RUNX1T1. With a mOS of around 12 months, outcomes of HMA/Ven treated CBF AML pts appear less favorable than those previously reported for pts with ELN 2024 favorable risk disease, and similar to those observed in ELN 2024 intermediate risk non-CBF AML pts, irrespective of FLT3-ITD, NRAS and/or KRAS mutations. Therefore, we propose that all pts with CBF AML treated with HMA/Ven should be provisionally classified as having ELN 2024 intermediate risk disease.

Supplementary Figure S1 shows enrichment of FBL in LSC, correlation of FBL with LSC signature and rRNA 2'-O-Me in healthy hematopoietic cells. Supplementary Figure S2 shows the association of rRNA 2’-O-Me with LSC and hematopoietic differentiation signatures. Supplementary Figure S3 shows strategy for LSC sorting and distribution of LSC methylation sites on ribosomes. Supplementary Figure S4 shows the effect of FBL knockdown on rRNA 2’-O-methylation and in vitro proliferation of leukemia cells. Supplementary Figure S5 show the effect of FBL overexpression on in vivo engraftment of primary AML cells. Supplementary Figure S6 shows the effect of FBL knockdown on nascent proteome and metabolism of AML cells. Supplementary Figure S7 shows the ribosome footprinting analysis after FBL knockdown. Supplementary Figure S8 shows the association of Gm1447 with LSC signature and the effect of Gm1447 supression on cellular amino acid levels. Supplementary Figure S9 shows regulatory effect of Gm1447 on in vivo leukemic self-renewal.

Introduction The randomized-controlled SAL DaunoDouble trial (NCT02140242) showed no significant differences in complete remission (CR) rates or survival between 60 mg/m² and 90 mg/m² of daunorubicin (Dauno), or between single and double induction in patients (pts) with newly diagnosed AML eligible for intensive chemotherapy (Röllig et al., JCO, 2024). To further study the impact of Dauno dose-escalation and the number of induction cycles in more detail, and to evaluate possible implications for post-remission therapy, we aimed to retrospectively analyze NPM1 measurable residual disease (MRD) kinetics in this study population. Methods Pts treated within DaunoDouble were retrospectively screened for data on NPM1 MRD. MRD was assessed within standard of care on bone marrow (BM) samples using a quantitative reverse transcription polymerase chain reaction (RT-qPCR) assay at the central reference laboratory. The assay achieves a validated sensitivity of 0.001%; pts were analyzed based on a threshold of 0.1% NPM1/ABL1. Time-to-event variables were estimated using the Kaplan-Meier method. Statistical analyses were performed using R v4.5.0. Results Data on post-induction MRD were available for 79 pts with NPM1-mutated AML who achieved a CR/CRi. The median age of this cohort was 51 years (range, 24-64), 49 pts (62%) were female. Most pts (97%) had de-novo AML. According to the European LeukemiaNet (ELN) 2017 risk classification, 78%, 22%, and 0% of pts had favorable, intermediate, and adverse risk disease, respectively. A FLT3-ITD co-mutation was observed in 31% of pts. Eighteen pts (23%) underwent allogeneic hematopoietic cell transplantation (alloHCT) in CR1. The median post-induction NPM1 load showed no significant differences after induction with Dauno 60 mg/m2 versus 90 mg/m2 (0.6% vs. 0.3%, p=0.601). Likewise, rates of MRD-negative remission did not significantly differ between pts receiving Dauno 60 mg/m2 and those receiving 90 mg/m2 (27% vs. 30%, p=0.804). In contrast, when comparing one versus two induction cycles, we detected a significantly lower post-induction MRD load in pts receiving double induction (0.1% vs. 1.5%, p<0.001). Pts receiving two cycles were significantly more likely to attain MRD-negative remissions (40% vs. 15%, p=0.015). As per March 22nd 2025, after a median follow-up of 43.7 months, the median overall survival (OS) of the entire cohort was not reached, with a corresponding 2-year OS rate of 86%. The 2-year relapse-free survival (RFS) rate of the entire cohort was 59%. When analyzing the entire cohort irrespective of post-remission treatment, the RFS in pts achieving MRD-negative CR was numerically longer when compared to pts being MRD-positive (median RFS (mRFS), NR vs. 42.5 months, 2-year RFS rate, 77% vs. 52%, p=0.084). When focusing on pts with cytarabine-based consolidation in CR1, the RFS benefit in MRD-negative pts was more pronounced and significantly longer than in MRD-positive pts (median RFS, NR vs. 15.3 months, 2-year RFS rate, 78% vs. 39%, p=0.014). With respect to the benefit of alloHCT in CR1, RFS was relevantly longer in MRD-positive pts undergoing alloHCT compared to chemotherapy consolidation (2-year RFS rate 83% vs. 41%, p=0.059). In contrast, alloHCT in CR1 did not improve RFS in pts considered MRD-negative post-induction (2-year RFS rate 69% vs. 67%, p=0.597). In terms of OS, we did not observe any relevant differences in pts with positive versus negative NPM1 MRD, regardless of post-remission therapy. Conclusions Dauno dose escalation from 60 to 90 mg/m2 in first induction did not improve depth of response in terms of NPM1 MRD levels. However, pts with double induction were more likely to achieve MRD-negative remissions and attained a higher reduction in their MRD load. Pts achieving MRD-negative remissions showed improved RFS which was statistically significant in pts not undergoing alloHCT in CR1. Regarding RFS, MRD-positive pts seemed to benefit from alloHCT in CR1, whereas MRD negative pts did not. Post-induction MRD status was not predictive of OS, most likely due to retained chemosensitivity and high efficacy of salvage alloHCT in pts with relapsed NPM1-mutated AML. Our results confirm that 60 mg/m² of Dauno is equipotent to 90 mg/m² on a molecular level, and sufficient for induction therapy. In addition, we demonstrate the predictive value of NPM1 MRD on relapse risk, as well as the value of alloHCT in pts with an insufficient response to induction.