Abstract Book for the 27^th Congress of the European Hematology Association * (BUTTON) Article notes * (BUTTON) Copyright and License information Collection date 2022 Jun. Copyright © 2022 The Author(s). Published by Wolters Kluwer Health, Inc. This is an open access Abstract Book distributed under the [19]Attribution-NonCommercial-NoDerivs (CC BY-NC-ND) which allows third parties to download the articles and share them with others as long as they credit the author and the Abstract Book, but they cannot change the content in any way or use them commercially. [20]PMC Copyright notice PMCID: PMC9429973 Presidential Symposium S100: QUIZARTINIB PROLONGED SURVIVAL VS PLACEBO PLUS INTENSIVE INDUCTION AND CONSOLIDATION THERAPY FOLLOWED BY SINGLE-AGENT CONTINUATION IN PATIENTS AGED 18-75 YEARS WITH NEWLY DIAGNOSED FLT3-ITD+ AML H. Erba^1,*, P. Montesinos^2, R. Vrhovac^3, E. Patkowska^4, H.-J. Kim^5, P. Zak^6, P.-N. Wang^7, T. Mitov^8, J. Hanyok^9, L. Liu^9, A. Benzohra^9, A. Lesegretain^9, J. Cortes^10, A. Perl^11, M. Sekeres^12, H. Dombret^13, S. Amadori^14, J. Wang^15, M. Levis^16, R. Schlenk^17 ^1Duke Cancer Institute, Durham, NC, United States of America; ^2La Fe University and Polytechnic Hospital, Valencia, Spain; ^3University Hospital Centre Zagreb, Zagreb, Croatia; ^4Institute of Hematology and Blood Transfusion, Warsaw, Poland; ^5The Catholic University of Korea, Seoul St. Mary’s Hospital, Seoul, South Korea; ^6University Hospital Hradec Kralove, Hradec Kralove, Czechia; ^7Chang Gung Medical Foundation, Linkou, Taiwan; ^8Daiichi Sankyo UK Ltd, Uxbridge, United Kingdom; ^9Daiichi Sankyo, Inc, Basking Ridge, NJ; ^10Augusta University Medical Center, Augusta, GA; ^11University of Pennsylvania, Philadelphia, PA; ^12University of Miami Health System, Miami, FL, United States of America; ^13Saint Louis Hospital, University of Paris, Paris, France; ^14Tor Vergata Polyclinic Hospital Rome, Rome, Italy; ^15Institute of Hematology and Blood Diseases Hospital, Tianjin, China; ^16Johns Hopkins University, Baltimore, MD, United States of America; ^17Heidelberg University Hospital and German Cancer Research Center, Heidelberg, Germany Background: Quizartinib (Quiz) is an oral, highly potent, and selective type II FLT3 inhibitor with single-agent activity in relapsed/refractory FLT3–internal tandem duplication positive (FLT3-ITD+) acute myeloid leukemia (AML). This is the first report of the global, randomized, double-blind, placebo (PBO)-controlled phase 3 QuANTUM-First trial ([21]NCT02668653). Aims: QuANTUM-First aimed to determine if the addition of Quiz to standard induction (IND) and post remission (including allogeneic hematopoietic cell transplant [allo-HCT]) in first complete remission [CR1]) consolidation followed by single-agent continuation therapy for up to 3 years improved survival compared with chemotherapy alone in patients (pts) with newly diagnosed FLT3-ITD+ AML. Methods: Pts aged 18-75 y with newly diagnosed AML were centrally screened for FLT3-ITD prior to initiation of IND with cytarabine 100 mg/m^2/day (200 mg/m^2/day if institutional standard) for 7 days and anthracycline (daunorubicin 60 mg/m^2/day or idarubicin 12 mg/m^2/day) for 3 days. Pts at 193 sites in 26 countries who were FLT3-ITD+ provided informed consent and were randomized to Quiz (40 mg/day days 8-21) or PBO and were stratified by region (North America, Europe, and Asia/Other regions), age (<60 y, ≥60 y), and white blood cell count (<40×10^9/L, ≥40×10^9/L) at diagnosis. A second IND was allowed if residual AML was noted at the post-IND marrow exam. Pts who achieved CR or CR with incomplete hematologic recovery (CRi) received up to 4 cycles of high-dose cytarabine plus Quiz (40 mg/day) or PBO and/or allo-HCT followed by up to 3 y of continuation therapy with Quiz (30-60 mg/day) or PBO. The primary endpoint was overall survival (OS). Results: Between September 2016 and August 2019, 3468 pts were screened, and 539 pts with FLT3-ITD+ AML were randomized to Quiz (n=268) or PBO (n=271). The median age was 56 y (range, 20-75 y). Baseline pt and disease characteristics, including FLT3-ITD variant allele frequency, were balanced between the 2 arms. At data cutoff (August 2021), the median follow-up was 39.2 months and 58 pts remained on continuation therapy. OS was significantly longer in the Quiz arm than the PBO arm (hazard ratio [HR], 0.776; 95% CI, 0.615-0.979; 2-sided P=.0324). Median OS was 31.9 mo with Quiz vs 15.1 mo with PBO (Figure). CR/CRi rates were 71.6% and 64.9%, respectively. Allo-HCT in CR1 was performed in 157 pts (Quiz, 31%; PBO, 27%). When censored for allo-HCT, OS trended longer with Quiz vs PBO (HR, 0.752; 95% CI, 0.562-1.008; 2-sided P=0.055). Relapse-free survival was longer with Quiz than PBO (HR, 0.733; 95% CI, 0.554-0.969). Although rates of grade ≥3 adverse events (AEs) were similar across arms, grade ≥3 neutropenia was more frequent in the Quiz arm (18.1% vs 8.6%). Discontinuations due to AEs occurred in 20.4% of Quiz and 8.6% of PBO pts. A total of 56 treatment-emergent AEs were associated with a fatal outcome (Quiz, 11.3%; PBO, 9.7%), mostly due to infections. Grade 3/4 electrocardiogram QT prolonged occurred in 3.0% of Quiz vs 1.1% of PBO pts. Image: graphic file with name hs9-6-1-g001.jpg [22]Open in a new tab Summary/Conclusion: These pivotal findings show that the addition of Quiz to standard chemotherapy and up to 3 years of continuation therapy yielded statistically significant and clinically meaningful improvements to OS in adults with newly diagnosed FLT3-ITD+ AML up to age 75 y. The manageable safety profile further supports use of Quiz in combination with standard therapy, including allo-HCT, in FLT3-ITD+ AML. S101: GENETIC AND EPIGENETIC FACTORS DRIVING PRIMARY MEDIASTINAL B-CELL LYMPHOMA PATHOGENESIS AND OUTCOME D. Noerenberg^1,*, F. Briest^1, C. Hennch^1, K. Yoshida^2,3, J. Nimo^1, R. Hablesreiter^1, Y. Takeuchi^2, D. Sasca^4, H. Ueno^2, L. Mansouri^5, Y. Inoue^2, L. Wiegand^1, A. M. Staiger^6,7, B. Casadei^8,9, M. Ziepert^10, F. Asmar^11, P. Korkolopoulou^12, M. Kirchner^13, P. Mertins^13, J. Weiner^14, E. Toth^15, T. Weber^16, A. Warth^17, T. Schneider^18, R.-M. Amini^19, W. Klapper^20, M. Hummel^21,22, V. Poeschel^23, G. Kanellis^24, A. Rosenwald^25, G. Held^23,26, E. Campo^27,28,29, K. Stamatopoulos^5,30, I. Anagnostopoulos^21,25, L. Bullinger^1,22, N. Goldschmidt^31, P. L. Zinzani^9,32, C. Bödor^33, R. Rosenquist^5,34, T. P. Vassilakopoulos^35, G. Ott^6, S. Ogawa^2,36,37, F. Damm^1,22 ^1Department of Hematology, Oncology and Cancer Immunology, Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Berlin, Germany; ^2Department of Pathology and Tumor Biology, Graduate School of Medicine, Kyoto University, Kyoto, Japan; ^3Wellcome Trust Sanger Institute, Hinxton, United Kingdom; ^4Department of Hematology, Oncology, and Pulmonary Medicine, University Medical Center, Johannes Gutenberg-University, Mainz, Germany; ^5Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden; ^6Department of Clinical Pathology, Robert-Bosch-Krankenhaus; ^7Dr Margarete Fischer-Bosch Institute of Clinical Pharmacology Stuttgart, and University of Tuebingen, Stuttgart, Germany; ^8Istituto di Ematologia “Seràgnoli”, IRCCS Azienda Ospedaliero-Universitaria di Bologna; ^9Dipartimento di Medicina Specialistica, Diagnostica e Sperimentale, Università di Bologna, Bologna, Italy; ^10Institute of Medical Informatics, Statistics and Epidemiology, University of Leipzig, Leipzig, Germany; ^11Department of Hematology, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark; ^12First Department of Pathology, National and Kapodistrian University of Athens, Laikon General Hospital, Athens, Greece; ^13Core Unit Proteomics, Berlin Institute of Health, Charité - Universitätsmedizin Berlin and Max-Delbrück-Center for Molecular Medicine; ^14Core Unit Bioinformatics, Berlin Institute of Health at Charité - Universitätsmedizin Berlin, Berlin, Germany; ^15National Institute of Oncology, Budapest, Hungary; ^16Department of Internal Medicine IV, Haematology and Oncology, University Hospital Halle (Saale), Martin-Luther-University Halle-Wittenberg, Halle; ^17Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany; ^18National Institute of Oncology, Budapest, Hungary; ^19Department of Immunology, Genetics and Pathology, Uppsala University and University Hospital, Uppsala, Sweden; ^20Department of Pathology, Hematopathology Section and Lymph Node Registry, Universitätsklinikum Schleswig-Holstein, Kiel; ^21Department of Pathology, Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Berlin; ^22German Cancer Consortium (DKTK) and German Cancer Research Center (DKFZ), Heidelberg; ^23Department of Internal Medicine 1 (Oncology, Hematology, Clinical Immunology, and Rheumatology), Saarland University Medical School, Homburg/Saar, Germany; ^24Department of Hematopathology, Evangelismos General Hospital, Athens, Greece; ^25Institute of Pathology, University of Würzburg and Comprehensive Cancer Center (CCC) Mainfranken, Würzburg; ^26Department Internal Medicine I, Westpfalzklinikum Kaiserslautern, Kaiserslautern, Germany; ^27Centro de Investigacion Biomedica en Red en Oncologia (CIBERONC), Madrid; ^28Hospital Clinic of Barcelona, University of Barcelona; ^29Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain; ^30Institute of Applied Biosciences, Centre for Research and Technology Hellas, Thessaloniki, Greece; ^31Hadassah-Hebrew University Medical Center, Jerusalem, Israel; ^32Istituto di Ematologia “Seràgnoli”, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy; ^33HCEMM-SE Momentum Molecular Oncohematology Research Group, 1st Department of Pathology and Experimental Cancer Research, Semmelweis University, Budapest, Hungary; ^34Clinical Genetics, Karolinska University Laboratory, Karolinska University Hospital, Stockholm, Sweden; ^35Department of Hematology and Bone Marrow Transplantation, National and Kapodistrian University of Athens, Laikon General Hospital, Athens, Greece; ^36Institute for the Advanced Study of Human Biology (WPI-ASHBi), Kyoto University, Kyoto, Japan; ^37Department of Medicine, Centre for Haematology and Regenerative Medicine, Karolinska Institutet, Stockholm, Sweden Background: Primary mediastinal large B-cell lymphoma (PMBCL) is an aggressive lymphoma affecting predominantly young female patients. Previous studies in this rare entity have focused on single genes or were limited in cohort size. Aims: To unravel the underlying genetic pathogenesis and its impact on outcome, we embarked on a comprehensive large-scale genetic investigation. Methods: Specimens of 486 previously untreated PMBCL patients were analyzed by paired tumor/normal whole-genome (WGS, n=14), whole-exome (WES, n=78) and targeted sequencing (TS, n=486). To understand the consequences of highly recurrent mutations in the chromatin-modifying gene ZNF217, we conducted functional and multi-omics analyses in CRISPR/Cas9 engineered cell lines. Results: WGS/WES revealed a complex genomic landscape in PMBCL with a median of 85 structural variants, a mutational burden of 5 mutations/Mb, 12 mutated coding candidate driver genes (CDG) and 4 focal somatic copy-number aberrations per sample (Fig.1a). Besides known targets, significant breakpoints were identified in genes previously not implicated in B-lymphomagenesis such as TOX and TP73 (36% and 21%). In addition, non-coding mutations clustered within the PAX5 enhancer region. With the identification of 50 recurrently mutated CDGs, we significantly expand the repertoire of known PMBCL drivers. The 10 most frequently mutated CDGs were SOCS1 (86%), B2M (67%), ITPKB (64%), ACTB (58%), STAT6 (58%), IGLL5 (56%), TNFAIP3 (53%), NFKBIE (49%), GNA13 (47%), and ZNF217 (36%), respectively. The operative mutational processes were attributed to aging, AID/APOBEC activity, defective MMR, and an unexpected infidelity of DNA-Polymerase ε. Next, we performed TS in 486 samples using a PMBCL-specific 106-gene panel. Recurrent lesions in 25 epigenetic modifiers were found in >90%, with ZNF217 being among the most frequently mutated genes (Fig.1b). After knockdown of ZNF217 in Karpas1106P and L428 cells, we demonstrated altered proliferation, migration, and apoptosis. Using mass spectrometry, we showed that ZNF217 is acting in a LSD1, CoREST and HDAC containing histone modifier complex. Accordingly, knockout of ZNF217 led to global changes in chromatin accessibility with an enrichment of differentially accessible motifs for crucial lymphoma-associated transcription factors, especially of the NF-κB, BATF/AP1, and IRF family, but also of CTCF, a major regulator of global 3D chromatin architecture. Resulting gene expression was characterized by changes in interferon-responsive genes and inflammation-associated transcription (Fig 1c). Clinical data were available for 329 cases, including 84 cases from clinical trials. Multivariate analysis using an IPI-corrected Cox regression model was performed. The estimated 5-year PFS and OS were 77% and 86%. Among the genetic lesions with the strongest association for poor outcome, we identified patients with mutatedCD58 having a significantly shorter survival (PFS: HR 2.96; p<.001; OS: HR 2.55; p=.006). In contrast, mutated DUSP2 indicated longer survival (PFS: HR 0.28; p=.002; OS: HR 0.15; p=.011) (Fig1d). Notably, DUSP2 mutated patients (25%) showed a similar outcome for CR rate, PFS and OS when comparing CHOP-like and intensified treatment regimens, suggesting no further benefit from treatment intensification in this very-low risk patient population. Image: graphic file with name hs9-6-1-g002.jpg [23]Open in a new tab Summary/Conclusion: Here, we present the genetic landscape of PMBCL highlighting a previously underappreciated role of chromatin modifying genes, identify novel treatment targets and provide a solid basis for guiding precision medicine approaches. S102: COMPREHENSIVE GENOME CHARACTERIZATION REVEALS NEW SUBTYPES AND MECHANISMS OF ONCOGENE DEREGULATION IN CHILDHOOD T-ALL P. Pölönen^1,*, A. Elsayed^1,2, L. Montefiori^1, S. Kimura^1, J. Myers^3, D. Hedges^3, J. Xu^4, Y. Hui^3, Z. Cheng^3, Y. Fan^3, Y. Chang^1, R. Shraim^5, M. Devidas^6, S. Winter^7, K. Dunsmore^8, J. J. Yang^9, T. L. Vincent^10, K. Tan^10,11,12,13,14, C. Chen^10, H. Newman^15, M. Loh^16, E. Raetz^17, S. P. Hunger^18, E. Rampersaud^3, T.-C. Chang^3, G. Wu^3, S. Pounds^2, C. G. Mullighan^1,19, D. T. Teachey^5,12,20 ^1Pathology; ^2Biostatistics; ^3Center for Applied Bioinformatics, St. Jude Children’s Research Hospital, Memphis; ^4Perelman School of Medicine at the University of Pennsylvania; ^5Department of Pediatrics and the Center for Childhood Cancer Research, Children’s Hospital of Philadelphia, Philadelphia; ^6Global Pediatric Medicine, St. Jude Children’s Research Hospital, Memphis; ^7Minnesota Research Institute and Cancer and Blood Disorders Program, Children’s Minnesota Research Institute, Minneapolis; ^8University of Virginia Children’s Hospital, Charlottesville; ^9Pharmaceutical Sciences, St. Jude Children’s Research Hospital, Memphis; ^10Division of Oncology and Center for Childhood Cancer Research; ^11Department of Biomedical and Health Informatics, Children’s Hospital of Philadelphia; ^12Perelman School of Medicine; ^13Institute for regenerative medicine; ^14Penn Epigenetics Institute, University of Pennsylvania; ^15Division of Oncology and Center for Childhood Cancer, Children’s Hospital of Philadelphia, Philadelphia; ^16Department of Pediatrics, Benioff Children’s Hospital and Helen Diller Family Comprehensive Cancer Center, University of California San Francisco, San Francisco; ^17Department of Pediatrics and Perlmutter Cancer Center, NYU Langone Health, New York; ^18Department of Pediatrics and the Center for Childhood Cancer Research, Children’s Hospital of Philadelphia and The Perelman School of Medicine at The University of Pennsylvania, Philadelphia; ^19Hematological Malignancies Program, St. Jude Children’s Research Hospital, Memphis; ^20Divisions of Hematology and Oncology, Children’s Hospital of Philadelphia, Philadelphia, United States of America Background: T-cell acute lymphoblastic leukemia (T-ALL) is an aggressive hematopoietic malignancy including leukemias of early T cell precursor acute lymphoblastic leukemia (ETP-ALL) and transformed thymocytes. Prior genomic studies of T-ALL had limited cohort size, excluded refractory disease and focused on alterations in coding parts of the genome. Aims: Investigate the genomic basis of T-ALL by identifying both coding and non-coding alterations and defining T-ALL subtypes. Methods: We performed whole-genome sequencing (WGS), whole-exome sequencing (WES), RNA-sequencing of 1,313 cases enrolled on the Children’s Oncology Group AALL0434 trial. Results: Uniform Manifold Approximation and Projection (UMAP) and gene expression clustering analyses of RNA-seq data identified 16 subtypes, of which 4 have not been reported previously. Furthermore, we could divide existing subtypes into smaller subgroups with common subtype-defining alterations, such as structural variation (SV) and copy number variation (CNV). TLX1 activation was linked with TLX1-TCR SVs and deletions in the TLX1 chromosomal domain, whereas TLX3 deregulation was associated with TLX3-BCL11B enhancer hijacking, but also through TLX3-TCR, TLX3-CDK6 rearrangements. We discovered two separate NKX2-1 deregulated groups, one characterized by TCR rearrangements and RPL10 mutations and the other by NKX2-1 CNVs, chromosome 14 chromothripsis, or MYB-TCR rearrangements. We also observed a distinct group of 9 cases aged 1-2 years, with recurrent STAG2-LMO2 rearrangements and a patient with inactivating STAG2 mutation and CELF1 enhancer hijacking by LMO2. Moreover, we found a group of 22 cases that were highly enriched for ETP-ALL with the following hallmark lesions: BCL11B enhancer amplification, BCL11B locus SVs involving enhancer hijacking of ARID1B, CCDC26, and novel CD34+ enhancer hijacking of lincRNA locus in chromosome 6. Gene expression-based clustering was unable to stratify patients based on TAL1, TAL2, LMO1, LMO2, LYL1 expression, and their respective activation mechanisms. However, WGS enabled further characterization of these patients, by identifying several types of activation mechanisms and co-occurring alterations for each oncogene. We identified canonical events, such as TAL1-STIL fusions, TCR rearrangements, and activation by TAL1/LMO1/LMO2 regulatory region indels, but also novel events, such as CD34 specific enhancer duplications downstream of TAL1 and BCL11B enhancer hijacking by LMO2. We also observed two smaller clusters with TAL1/LMO2 activation, where one with 39 cases was associated with TAL1/LMO2 activation and RPL10 mutations and the other with 8 cases had refractory disease (Day 29 MRD >5%). HOXA gene expression-associated subtypes were defined by fusions involving KMT2A or MLLT10 and PICALM/DDX3X or HOXA9-TCR SVs. Furthermore, our analysis revealed segregation of patients by HOXA13 or ZFP36L2 rearrangements and NUP98/NUP214 fusions. Interestingly, HOXA locus breakpoints involving HOXA13 and HOXA9 were in different chromatin compartments and were associated with mutually exclusive activation of either HOXA13 or other HOXA genes. Interestingly, HOXA13-deregulation was associated with ETP-ALL and frequent BCL11B enhancer hijacking, whereas HOXA9 breakpoints typically involved TCRg and were non-ETP. Image: graphic file with name hs9-6-1-g003.jpg [24]Open in a new tab Summary/Conclusion: Large-scale analysis of all children enrolled on the AALL0434 study has identified new subtype-defining lesions in T-ALL, including candidate novel enhancer hijacking events and enhancer duplications that are likely to result in oncogene deregulation in T-ALL. S103: EFFICACY AND SAFETY OF ARI0002H, AN ACADEMIC BCMA-DIRECTED CAR-T CELL THERAPY WITH FRACTIONATED INITIAL THERAPY AND BOOSTER DOSE IN PATIENTS WITH RELAPSED/REFRACTORY MULTIPLE MYELOMA C. Fernandez De Larrea^1,*, V. González-Calle^2, A. Oliver-Caldés^1, V. Cabañas^3, P. Rodríguez-Otero^4, M. Español-Rego^1, J. L. Reguera^5, L. López-Corral^2, B. Martin-Antonio^1, B. Paiva^4, S. Inogés^4, L. Rosiñol^1, A. López-Díaz de Cerio^4, N. Tovar^1, M. López-Parra^2, L. G. Rodríguez-Lobato^1, A. Sánchez-Salinas^3, S. Varea^1, V. Ortiz-Maldonado^1, J. A. Pérez Simón^5, F. Prósper^4, M. Juan^1, J. M. Moraleda^3, M. V. Mateos^2, M. Pascal^1, A. Urbano-Ispizua^1 ^1Hospital Clínic de Barcelona. IDIBAPS. University of Barcelona, Barcelona; ^2Hospital Universitario de Salamanca, Instituto de Investigacion Biomedica de Salamanca (IBSAL), Centro de Investigación del Cancer (IBMCC-USAL, CSIC), Salamanca; ^3Hospital Clínico Universitario Virgen de la Arrixaca. IMIB-Arrixaca. University of Murcia, Murcia; ^4Clínica Universidad de Navarra, Centro de Investigacion Medica Aplicada (CIMA), IDISNA, CIBERONC Pamplona, Pamplona; ^5Hospital Universitario Virgen del Rocío, Instituto de Biomedicina de Sevilla (IBIS/CSIC/CIBERONC), University of Sevilla, Sevilla, Spain Background: ARI0002h is a lentiviral autologous CAR T-cell product with a 4-1BB co-stimulatory domain and a humanized single chain variable fragment targeting BCMA. In pre-clinical studies, this academic CAR-T has demonstrated potent in vitro and in vivo activity. Aims: We report the safety and efficacy results of the CARTBCMA-HCB-01 multicenter clinical trial for patients with relapsed/refractory multiple myeloma (RRMM) ([25]NCT04309981) who received ARI0002h in 5 Spanish centers. Methods: Patients (pts) aged 18-75 years old with RRMM were eligible for this study if they had measurable disease, received ≥2 prior regimens, including a proteasome inhibitor, an immunomodulatory drug and an anti-CD38 antibody, and were refractory to the last line of treatment. Bridging therapy was allowed after apheresis. Cyclophosphamide (300 mg/m^2) and fludarabine (30 mg/m^2) were used as lymphodepletion regimen. The targeted dose was 3x10^6/kg CAR+ cells and was administered in a fractionated manner (10%/30%/60%), with at least 24h between infusions. A second dose of 3x10^6 CAR+ cells/kg was planned at least 4 months after the first dose in pts who achieved any grade of response any response and had not or serious complications after the first administration. Primary objectives were overall response rate (ORR; at least partial response -PR-) within 3 months of the first infusion and rate of cytokine release syndrome (CRS) and/or neurological toxicity in the first 30 days. Response was assessed as per IMWG criteria and bone marrow minimal residual disease (MRD) was analyzed by next-generation flow (NGF). Results: As of February 9^th 2022, 35 pts (median age 61 years) with RRMM were included in the trial. Four pts could not receive ARI-0002h due to MM progression and one died of infection. Therefore, 30 pts received ARI0002h cells (modified intention-to-treat population), of which 47% received bridging therapy. Median CAR-T cell production time was 11 days (range 9-14) with a 100% manufacture success. Median follow-up after ARI0002h administration for surviving pts was 16 months. The ORR of 30 evaluable pts was 100%, with a stringent complete remission (sCR) plus very good partial response (VGPR) rate of 90%. Median time to first response was one month. Of 26 MRD-evaluable pts at day +100, 92% were MRD-negative in bone marrow by NGF. 53% of patients were alive and without progression at 16 months. Median overall survival (OS) was not reached and the 16-month OS rate was 80% (Figure 1). AEs reported in >70% of pts were CRS (87%; grade [gr] 3/4 0%; gr 1 73%), neutropenia (97%; gr 3/4 100%), anemia (85%; gr 3/4 43%), and thrombocytopenia (79%; gr 3/4 70%). Median duration of CRS was 4 days (range 1-12). No CAR-T cell-related neurotoxicity cases were reported. Tocilizumab and corticosteroids were administered in 76% (mainly for persistent grade 1 CRS) and 12% of pts, respectively. ARI0002h cells demonstrated peak expansion on day 14 (range 7 days-6 months). 24 out of 28 eligible pts (86%) received the second dose (range 1.2-3x10^6 CAR+ cells/kg). Median time after first infusion was 4 months and 38% received a second lymphodepletion regimen. No relevant toxicities after second infusions were reported. 7 pts (29%) improved their response after reinfusion. Image: graphic file with name hs9-6-1-g004.jpg [26]Open in a new tab Summary/Conclusion: ARI0002h is the first European academic CART for RRMM that has demonstrated deep and durable responses and a favorable safety profile, including the absence of neurotoxicity and the feasibility of a second booster dose. S104: RBPS DYSREGULATION CAUSE HYPER-NUCLEOLI AND RIBOSOME GAIN-OF-FUNCTION DRIVING BONE MARROW FAILURE P. Aguilar-Garrido^1,*, M. Velasco^1, M. Hernández Sánchez^2, M. Á. Navarro Aguadero^1, P. Malaney^3, M. JL Aitken^3, X. Zhang^3, K. H Young^3, R. Duan^3, P. Hu^3, S. Kornblau^3, A. Fernández^1, A. Ortiz^1, Á. Otero-Sobrino^1, P. J. de Andrés^2, D. Megías^1, M. Pérez^1, J. Gómez^1, G. Mata^1, J. Martínez López^1, S. Post^3, M. Gallardo^1 ^1CNIO; ^2UCM, Madrid, Spain; ^3MD Anderson, Houston, United States of America Background: Nucleoli and ribosome cross-talk regulates cell translation capacity. Its dysregulation and impairment drive ribosome and nucleolus stress (NS), related to the biological mechanism of cancer. Ribosomopathies are a group of diseases characterized by ribosome defects leading to complex syndromes that include bone marrow failure. hnRNP K is an RNA binding protein (RBP) in charge of processing nascent RNAs (nucleoli) into mature mRNAs (ribosome). Our research found a novel ribosome gain-of-function ribosomopathy phenotype by hyper-nucleoli generation due RBP hnRNP K dysregulation. Aims: We aim to elucidate how hnRNP K dysregulation impact on haematopoietic stem cell biology. Methods: Hnrnpk overexpression was established in MEFs using CRISPR/SAM (Konermann S. et al, Nature). Global protein synthesis was tested using a Click-iT OPP and proteasome function was evaluated by Proteasome 20S activity NS hallmarks were analyzed by confocal microscopy evaluating Ncl, NS sensor marker. To evoke NS in our cells, we used Actinomycin D insult. Cell cycle FACS analysis (DAPI) and senescence assays (β-galactosidase staining) were performed. Molecular mechanism underlying was elucidated by qRT-PCR and WB (p21, p16, c-Myc, and mTor). To study the impact of hnRNP K overexpression in vivo, we developed an inducible tamoxifen mouse model activated 30-60 days after birth (Hnrnpk^Tg-Ubc-creERT2). Survival was evaluated by Kaplan-Meier, and phenotype described by symptoms, signs, CBC and bone marrow IHC panel (CD34, Gr1, B220, MPO). Results: Hnrnpk overexpressing cells led to an increment in protein and gene expression of Ncl, mTor and c-Myc (Figure 1A-B). Moreover, we found an increase in global protein synthesis (Figure 2C-D). Nevertheless, the elevation of hnRNP K inversely correlated with the proteasome function, which dropped significantly (Figure 1E). NS induction promoted higher hnRNP K expression. Additionally, hnRNP K overexpressing cells showed NS hallmarks associated with an increase of the number of nucleoli, and total area of the nucleoli and nucleus (Figure 1F-G). Cell cycle analysis confirmed an increment of arrested G2/M phase cells (Figure 1H-I), linked to an increment in p21 and p16 levels all leading towards a senescent cell phenotype (Figure 1J-L). HnrnpkTg-Ubc-creERT2 mice had widespread Hnrnpk overexpression (Figure 1M) and a reduction in lifespan (Figure 1N), mainly due to dysplastic and bone marrow failure phenotype, with dramatic reduction of CD34 and b-cells, leukopenia, anaemia and thrombocytopenia. (Figure 1O-Q). Image: graphic file with name hs9-6-1-g005.jpg [27]Open in a new tab Summary/Conclusion: The overexpression of hnRNP K drives to an increase in nucleoli activity, leading to ribosome biogenesis and higher global translation by the regulation of molecules involved in both systems: Ncl, c-Myc or mTor. Our work found that hnRNP K overexpression in vivo drives a bone marrow failure phenotype, promoting the exhaustion of haematopoietic stem cells by ribosome dysregulation that triggered cell senescence. Of note, this is the first time reported that a nucleoli/ribosome-gain-of-function induce bone marrow failure ribosomopathies phenotype. This work was financially supported by CRIS contra el Cancer Association (NGO) AES ISCIII (PI18/00295), ISCIII Miguel Servet (CP19/00140), Cancer Research UK [C355/A26819], FC AECC and AIRC under the Accelerator Award Program and National Cancer Institutes of Health Award (R01CA207204, SMP) Leukemia and Lymphoma Society (6577-19, SMP). Oral Sessions S105: IN VIVO PDX CRISPR/CAS9 SCREENS REVEAL MUTUAL THERAPEUTIC TARGETS TO OVERCOME HETEROGENEOUS ACQUIRED CHEMO-RESISTANCE A.-K. Wirth^1,*, L. Wange^2, S. Vosberg^3, A. K. Jayavelu^4, W Enard^2, T Herold^5, I Jeremias^1 ^1Apoptosis in Hematopoietic stem cells, Helmholtz Zentrum München, German Research Center for Environmental Health (HMGU), Munich; ^2Anthropology and Human Genomics, Faculty of Biology, Ludwig Maximilian University (LMU), Martinsried, Germany; ^3Clinical Division of Oncology, Department of Internal Medicine, Medical University of Graz, Graz, Austria; ^4Department of Proteomics and Signal Transduction, Max Planck Institute of Biochemistry, Martinsried; ^5Department of Medicine III, and Laboratory for Leukemia Diagnostics, Ludwig Maximilian University (LMU), Munich, Germany Background: Acquired resistance to conventional polychemotherapy regimens leads to relapse and poor prognosis, and remains a major unmet clinical need. Aims: To identify therapeutic options to overcome acquired chemo-resistance. Methods: We studied acute lymphoblastic leukemia (ALL) as model disease and combined long-term in vivo treatment in orthotopic patient-derived xenograft (PDX) models with multi-omics profiling and functional genomic CRISPR/Cas9 screens. Results: We adapted conventional chemotherapeutic protocols to allow treatment of mice for up to 18 consecutive weeks, using a combination of the widely used drugs cyclophosphamide and vincristine. Three luciferase-transgenic PDX models were monitored by repetitive in vivo imaging. Polychemotherapy strongly reduced PDX ALL cells within the first weeks, proving initial sensitivity of the sample. Under continuous treatment, tumor load persisted in mice at the level of minimal residual disease for several weeks, until tumors resumed growth despite treatment, indicating acquired resistance. In an exemplary PDX model, eight resistant derivatives were generated in replicate mice and characterized individually. Genomic profiling revealed profound genomic heterogeneity between distinct derivatives; individual resistant derivatives acquired different copy number alterations in regions associated with resistance and distinct point mutations, e.g. in TP53. In contrast to genomic heterogeneity, transcriptome and proteome profiling identified a group of genes differentially expressed between sensitive and resistant cells, but similar across all derivatives. To gain insights into underlying mechanisms and to identify therapeutic targets to overcome acquired resistance, a customized CRISPR/Cas9 in vivo dropout screen was performed in individual resistant PDX derivatives, to test the relevance of around 200 candidate genes under treatment. Among others, sgRNAs targeting BCL2, BRIP1 or COPS2 dropped out in the in vivo screen specifically during treatment; single knockout experiments confirmed that knockout of either BCL2, BRIP1 or COPS2 re-sensitized PDX ALL cells towards chemotherapy. Of direct translational relevance, treatment of mice with the BCL2 inhibitor ABT-199 sensitized resistant PDX cells towards polychemotherapy in vivo. Interestingly, BCL2 inhibition restored treatment response in resistant derivatives independently from the highly diverse underlying genetic alterations, e.g., in clones with and without mutation in TP53. Summary/Conclusion: Taken together, we established a highly clinically relevant PDX in vivo model of acquired resistance to conventional chemotherapy. Using this model, we demonstrate that heterogeneous genomic alterations evolved in parallel in replicate mice, which could be overcome by a single therapeutic approach to re-sensitize tumors towards conventional chemotherapy. S106: UBTF-ATXN7L3 GENE FUSION DUE TO 17Q21.31 DELETION DEFINES NOVEL HIGH-RISK ALL SUBTYPE AMENABLE TO MRD-BASED TREATMENT INTENSIFICATION L. Bastian^1,*, A. Hartmann^1, T. Beder^1, S. Hänzelmann^1, J. Kässens^1, M. Bultmann^1, M. P. Höppner^2, S. Franzenburg^2, M. Wittig^2, A. Franke^2, I. Nagel^3, M. Spielmann^3, N. Reimer^4, H. Busch^4, S. Schwartz^5, B. Steffen^6, A. Viardot^7, K. Döhner^7, M. Kondakci^8, G. Wulf^9, K. Wendelin^10, A. Renzelmann^11, A. Kiani^12, H. Trautmann^1, M. Neumann^1, N. Gökbuget^6, M. Brüggemann^1, C. Baldus^1 ^1Department of Medicine II, Hematology and Oncology, University Medical Center Schleswig-Holstein, Kiel; ^2Institute for Clinical Molecular Biology, Kiel University, Kiel; ^3Institute of Human Genetics, University Medical Center Schleswig-Holstein, Kiel and Lübeck, Kiel and Lübeck; ^4Medical Systems Biology Group and Institute for Cardiogenetics, University of Lübeck, Lübeck; ^5Department of Hematology, Oncology and Tumor Immunology (Campus Benjamin Franklin), Charité - Universitätsmedizin Berlin, Berlin; ^6Department of Medicine II, Hematology/Oncology, Goethe University Hospital, Frankfurt / Main; ^7Department of Internal Medicine III, University Hospital Ulm, Ulm; ^8Department of Hematology, Oncology and Clinical Immunology, University Hospital Düsseldorf, Düsseldorf; ^9Department of Hematology and Oncology, University Hospital Göttingen, Göttingen; ^10Medical Department V, Hospital Nürnberg, Paracelsus Medizinische Privatuniversität, Nürnberg; ^11Medical Department Oncology and Hematology, University Medical Center Oldenburg, Oldenburg; ^12Department of Medicine IV, Hematology/Oncology, Klinikum Bayreuth, Bayreuth, Germany Background: Response to induction chemotherapy assessed by quantification of minimal residual disease (MRD) is the strongest independent prognosticator in B precursor acute lymphoblastic leukemia (BCP-ALL). Molecular underpinnings of MRD poor response are insufficiently understood. Aims: We aimed to identify novel high-risk subtypes in adult BCP-ALL as cell-intrinsic determinants of MRD poor response. Methods: Adult BCP-ALL patients (n=565) were treated according to pediatric inspired protocols of the German Acute Lymphoblastic Leukemia Study group (GMALL) and profiled for integrative analyses by RNA-Seq (n=565), SNP-arrays (n=115), whole exome sequencing (WES; n=84) and whole genome sequencing (WGS; n=3). Results: Concordance between transcriptomic and genomic profiles was used to allocate samples to one of 15 established molecular driver subgroups (Figure A). Unsupervised clustering of gene expression from the remaining samples revealed a distinct cluster (n=12/565, 2.1%) with an in-frame gene fusion between upstream binding transcription factor (UBTF) and ataxin-7-like protein 3 (ATXN7L3) as exclusive event in these patients. Both fusion partners are in direct neighborship at 17q21.31. WGS revealed a 10.08 kb genomic deletion which truncated UBTF at exon 17/21 and comprised most of the intergenic region between both genes (Figure B). UBTF-ATXN7L3 rearranged cases frequently harbored 1q gains (n=5/7). Further genomic profiling showed a remarkable paucity of additional cooperating events compared to other molecular subtypes, supporting a prominent driver function of the newly identified fusion. UBTF and ATXN7L3 are global epigenetic regulators involved in transcriptional control. Both genes were highly expressed across the entire cohort. The gene fusion was associated with a marked increase of Caudal Homeobox 2 (CDX2) expression. Analysis of functional modules related CDX2 to upregulated HOXA9 and MEIS1, described essential co-regulators of KMT2A-driven leukemogenesis. NTRK3 expression was also strongly upregulated, suggesting a possible rationale for specific inhibitors. UBTF-AXTN7L3 rearranged patients were older, more frequently female and presented with normal leukocyte counts, low bone marrow infiltration and pro-B immunophenotypes or common ALL with reduced CD10 expression. Response to induction chemotherapy in evaluable patients (n=11) was poor with only 3 patients achieving MRD negativity after consolidation I compared to n=271/402 (67%; p=0.019) in the remaining cohort. Four patients suffered either cytologic (n=2) or molecular (n=2) relapse. Immunotherapeutic treatment intensification using blinatumomab (n=5) or inotuzumab ozogamizin (n=1) and / or allogenic stem cell transplantation (n=7) in MRD poor responders or relapsed cases resulted in an overall survival probability of 80% (+/- 12%) vs. 73% (+/- 2%; p=0.07) in the remaining cohort. Heterogeneous MRD responses were observed for other molecular subtypes (poor: ZNF384 - 48.2% MRD neg., p=0.056; Ph-like - 54.0% MRD neg., p=0.003; KMT2A - 55.8% MRD neg., p=0.127 / good: High Hyperdiploid - 90.9% MRD neg., p=0.01; TCF3-PBX1 - 94.1% MRD neg., p=0.016) indicating how molecular drivers affect chemo-sensitivity in adult BCP-ALL. Image: graphic file with name hs9-6-1-g006.jpg [28]Open in a new tab Summary/Conclusion: Molecular driver alterations determine sensitivity to induction chemotherapy in adult BCP-ALL. UBTF-ATXN7L3 ALL represents a novel subtype with poor induction chemotherapy response which could be successfully salvaged by MRD-based treatment intensification using immunotherapeutic strategies. S107: SOD2 PROMOTES ACUTE LEUKEMIA ADAPTATION TO AMINO ACID STARVATION THROUGH THE N-DEGRON PATHWAY N. K. Ibrahim^1,*, S. Schreek^1, B. Cinar^1, L. Loxha^1, B. Fehlhaber^1, J.-P. Bourquin^2, B. Bornhauser^2, C. Eckert^3, G. Cario^4, M. Forster^5, M. Stanulla^1, A. Gutierrez^6, L. Hinze^1 ^1Department of Pediatric Hematology and Oncology, Hannover Medical School, Hannover, Germany; ^2Department of Pediatric Hematology and Oncology, University Children’s Hospital Zurich, Zurich, Switzerland; ^3Department of Pediatric Hematology and Oncology, Charité Universitätsmedizin, Berlin; ^4Department of Pediatrics I, Christian-Albrecht University Kiel and University Medical Center Schleswig-Holstein; ^5Institute of Clinical Molecular Biology, Kiel, Germany; ^6Division of Pediatric Hematology and Oncology, Boston Children’s Hospital, Boston, United States of America Background: The ability of cells to tolerate amino acid starvation is fundamental for survival under cellular stress conditions. Some cancer cells are vulnerable to asparagine depletion, which is exploited therapeutically using asparaginase. However, the mechanisms of adaptation to amino acid starvation in leukemia cells remain incompletely understood. Aims: We recently performed a genome-wide CRISPR/Cas9 loss-of-function screen in the resistant T-ALL cell line CCRF-CEM to identify molecular pathways that promote asparaginase resistance. We found that Wnt-dependent stabilization of proteins (Wnt/STOP) induces a profound therapeutic vulnerability to asparaginase in acute leukemias and colorectal cancers (Hinze et. al., 2019; Hinze et al., 2020). Another unrelated gene on the top of the screen included SOD2, a mitochondrial superoxide dismutase. Intriguingly, to date, SOD2 activity has not been linked to a cellular amino acid starvation response, whose biologic basis we thus sought to further investigate. Methods: To evaluate the significance of SOD2 in mediating an asparaginase response, we employed genetic epistasis experiments as well as phenotypic assays including short hairpin RNA (shRNA) mediated knockdowns, quantitative PCRs, Western blots, amino acid starvation, and viability assays. Results: Knockdown of SOD2 (shSOD2) resulted in a profound sensitivity to asparaginase in several T-ALL and B-ALL cell lines (p<0.0001), and an increase in apoptosis, as assessed by caspase 3/7 activity (p<0.001). The sensitization was rescued by either overexpressing SOD2 cDNA (p<0.0001), or by adding the functional SOD2 mimetic MnTBAP (p<0.01). Of note, shSOD2 mediated sensitization was selective to asparaginase, as it could not be observed for other commonly used chemotherapeutic agents including vincristine, doxorubicin, dexamethasone, and 6-mercaptopurine (p=ns). Due to the selectivity to asparagine depletion, we then investigated whether SOD2 inhibition mediates a broader amino acid starvation response. Indeed, culturing SOD2-inhibited T-ALL cells in the absence of essential amino acids (EAA) or non-EAA, induced a significant decrease in cell viability (p<0.05). Sensitization appeared to be specific to the SOD2 isoform, and distinct from known SOD2-associated pathways including reactive oxygen species, cell cycle changes, alterations of mTOR signaling, or glutamine anaplerosis. To better understand the molecular underpinnings of SOD2 in regulating an amino acid starvation response, we leveraged the Bioplex Interactome database (Huttlin et al., 2020), and identified UBR2, an E3 ubiquitin ligase in the N-degron pathway, as a unique binding partner of SOD2. Ubiquitin E3 ligases target their substrates for ubiquitination, leading to proteasome-mediated degradation (Yang et al., 2010). Indeed, SOD2 and UBR2 were co-immunoprecipitated, suggesting the formation of a complex that can drive proteasome-dependent protein catabolism. In line, inhibition of SOD2 significantly decreased ubiquitin levels, suggesting that SOD2 positively regulates catabolic protein degradation through the N-degron pathway to promote cancer cell fitness in amino acid starved conditions. Summary/Conclusion: The interaction of SOD2 and the N-degron pathway represents a previously unknown molecular adaptation of cancer cells in response to amino acid starvation. These results serve as a strong proponent for an in-depth characterization of the N-degron pathway in mediating leukemia cell fitness upon amino acid starvation and thus provide a basis for therapeutic intervention in refractory leukemias. S108: PEDIATRIC T- ALL RELAPSE: CONSTITUTIONAL CANCER PREDISPOSITION AND HYPERMUTATATOR PHENOTYPES P. Richter-Pechanska^1 2,*, J. Kunz^1 2, T. Rausch^2 3, B. Erarslan-Uysal^1 2, B. Bornhauser^4, V. Frismantas^4, Y. Assenov^5, M. Zimmermann^6, M. Happich^1, C. von Knebel-Doeberitz^1, N. von Neuhoff^7, R. Koehler^8, M. Stanulla^6, M. Schrappe^9, G. Cario^9, G. Escherich^10, R. Kischner-Schwabe^11, C. Eckert^11, S. Avigad^12, S. Pfister^5 13, M. Muckenthaler^1 2, J.-P. Bourquin^4, J. Korbel^2 3, A. Kulozik^1 2 5 ^1University Hospital Heidelberg; ^2MMPU; ^3EMBL, Heidelberg, Germany; ^4University Children’s Hospital, Zürich, Switzerland; ^5DKFZ, Heidelberg; ^6Hannover Medical School, Hannover; ^7University Hospital, University of Duisburg-Essen, Essen; ^8University Heidelberg, Heidelberg; ^9University Hospital Schleswig-Holstein, Kiel; ^10University Medical Center Hamburg-Eppendorf, Hamburg; ^11Charite, Berlin, Germany; ^12Schneider Children’s Medical Center of Israel, Petah Tikva, Israel; ^13KiTZ, Heidelberg, Germany Background: Relapse is the main cause of death from pediatric acute precursor T-cell leukemia (T-ALL), but the underlying mechanisms of disease evolution from initial disease to relapse remain incompletely understood and show remarkable interpatient heterogeneity. Aims: As cross-sectional studies failed to identify unifying determinants of relapse, we adopted a longitudinal strategy and performed multi-omic analyses in 13 matched pairs of initial diagnosis and relapse samples and their matched PDXs. We extended this set by WES and methylome analyses in an additional cohort of 25 matched DNA samples from patient cells collected at initial diagnosis, remission, and relapse. Methods: Thirty-eight patients were recruited from the ALL‐BFM 2000/2009, CoALL97/03/09, and ALL‐REZ BFM 2002 trials or from Schneider Children’s Medical Center of Israel at time points of initial diagnosis, remission, and relapse. Material from 13 matched pairs of PDXs (RNA, cells) was used for multi-omic analyses, including DNA-Seq (WES), RNA-Seq, ATAC-Seq and methylation analysis with EPIC arrays. Results: Based on the profile of SNVs and InDels we distinguished 18 (47%) type-1 (derived from the major ancestral clone) and 20 (53%) type-2 relapses (derived from a minor ancestral clone). We observed stronger remodeling on the way to type 2 than to type 1 relapses reflected by more evident changes in methylation, chromatin accessibility and gene expression. At the time of relapse, 3/20 type 2 patients exhibited a hypermutator phenotype, probably caused by gains of mutations in TP53, BLM and BUB1B combined with PMS2. Moreover, type 2 T-ALLs were predominantly TAL1-driven (4/8) in contrary to type 1 (0/5). T-ALLs that later progressed to type-2 relapses exhibited a complex subclonal architecture, unexpectedly, already at the time of initial diagnosis. The fraction of subclonal mutations of those T-ALLs that later developed into a type-2 relapse was significantly higher already at the time of initial diagnosis than in those T-ALLs that later developed into a type-1 relapse (p=0.0387; Fisher’s exact), a difference that became even more pronounced at the time of relapse (p<0.0001). On the other hand, relapse type 1 T-ALLs exhibited overexpression of IL7R, its ligand HGF, and repressors of cytokine signaling (SOCS1, SOCS2, SOCS3) which regulates the IL7R pathway via negative feedback loop. Deconvolution analysis of ATAC-Seq profiles showed that T-ALLs later developing into type-1 relapses resembled a predominant immature thymic T-cell population, whereas T-ALLs developing into type-2 relapses resembled a mixture of normal T-cell precursors. Moreover, an analysis of remission samples revealed a significant enrichment of mutations in constitutional cancer predisposition genes (CPG) in type 2 patients, thus indicating fundamental differences between these two groups of patients. In both types of relapse, we observed known and novel drivers of drug resistance including MDR1 and MVP and NT5C2. Image: graphic file with name hs9-6-1-g007.jpg [29]Open in a new tab Summary/Conclusion: In sum, our comprehensive analyses revealed fundamentally different mechanisms driving either type-1 or type-2 T-ALL relapse and indicate that differential capacities of disease evolution are already inherent to the molecular setup of the initial leukemia. Leukemias of patients with type-1 relapses were often characterized by upregulation of the IL7R pathway, whereas type-2 relapses were characterized by (i) an enrichment of TAL-1 fusion, (ii) and of constitutional mutations in CPG, (iii) divergent genetic and epigenetic remodeling, and (iv) an enrichment of somatic hypermutator phenotypes. S109: ONCOGENIC DEUBIQUITINATION CONTROLS TYROSINE KINASE SIGNALING AND THERAPY RESPONSE IN ACUTE LYMPHOBLASTIC LEUKEMIA P. Ntziachristos^1,*, Q. Jin^2, B. Gutierrez^3 ^1Ghent University, Ghent, Belgium; ^2MD Anderson, Houston; ^3Northwestern University, Chicago, United States of America Background: Dysregulation of kinase signaling pathways via mutations favors tumor cell survival and resistance to therapy and it is common in cancer. Our data unveil how dysregulated deubiquitination controls signaling pathways, leading to cancer cell survival and drug non-response, and suggest novel therapeutic combinations towards targeting T-cell acute lymphoblastic leukemia (T-ALL). Here, we reveal a novel mechanism of post-translational regulation of kinase signaling and nuclear receptor activity via deubiquitination in acute leukemia. Aims: This study aims at 1) characterizing the function of an oncogenic complex composed by two deubiquitinating enzymes in in vitro and in vivo leukemia systems and 2) testing the association of deubiquitinase activity with resistance to therapy in acute lymphoblastic leukemia. Methods: We use genetic mouse and human:mouse xenograft models of T-cell leukemia, biochemical studies (quantitative global proteomics, phosphoproteomics and ubiquitination analysis) and high-throughput molecular biology (chromatin conformation capture (HiC), chromatin accessibility (ATAC-Seq) and gene expression (RNA-Seq)) analyses. Results: We observed that the ubiquitin specific protease 11 (USP11) is highly expressed in lymphoblastic leukemia and associates with poor prognosis in this disease. USP11 ablation inhibits leukemia growth in vitro and in vivo, sparing normal hematopoiesis and thymus development, suggesting that USP11 could be a therapeutic target in leukemia. USP11 forms a complex with USP7 to deubiquitinate the oncogenic lymphocyte cell-specific protein-tyrosine kinase (LCK). Deubiquitination of LCK controls its activity, thereby altering T cell receptor signaling. Impairment of LCK activity leads to increased expression of the glucocorticoid receptor transcript, culminating into transcriptional activation of pro-apoptotic target genes, and sensitizes cells to glucocorticoids in T cell leukemia patient samples. The transcriptional activation of pro-apoptotic target genes, such as BCL2L11, is orchestrated by the deubiquitinase activity and mediated via an increase in enhancer-promoter interaction intensity. Pharmacological inhibition of USP7 or genetic knockout of USP7 in combination treatment of glucocorticoid displayed improved anti-T-ALL efficacy in vivo. Image: graphic file with name hs9-6-1-g008.jpg [30]Open in a new tab Summary/Conclusion: Our data unveil how dysregulated deubiquitination controls signaling pathways, leading to cancer cell survival and drug non-response, and suggest novel therapeutic combinations towards targeting T-cell leukemia. S110: A NOVEL AND SUCCESSFUL CD7 GENE KNOCKOUT CAR-T CELL THERAPY FOR RELAPSED OR REFRACTORY T-CELL HEMATOLOGIC MALIGNANCIES J. Yang^1, J. Li^1, X. Zhang^1, L. Qiu^1, P. Lu^1,* ^1Hebei Yanda Lu Daopei Hospital, Langfang, China Background: T cell malignancies represent a group of hematologic cancers with high relapse and mortality rates. The shared expression of target antigens between chimeric antigen receptor (CAR) T cells and malignant T cells has limited the development of CAR-T due to unintended CAR-T fratricide. Here, we develop a fratricide-resistant anti-CD7 CAR-T modified by CD7 ablation through CRISPR/CAS9 gene editing (KO7CAR). Aims: In a phase I clinical trial, we explored the efficacy and safety of KO7CAR T-cells for relapsed or refractory (R/R) T-cell malignancies ([31]NCT04916860 & [32]NCT04938115). Methods: Peripheral blood mononuclear cells were collected from patients (n=13) or the transplant donor (n=2) by leukapheresis. CD7-ablated CAR T cells (KO7CAR) were derived by electroporation of bulk T cells with CD7-targeting Cas9-gRNA RNP 24 hours before 7CAR transduction. This KO7CAR is a second-generation CAR-T with the co-stimulatory domain of 4-1BB and CD3ζ targeting CD7. Intravenous fludarabine (30mg/m^2/d) and cyclophosphamide (300mg/m^2/d) were given to all patients on day -5 to day -3 prior to KO7CAR-T cells infusion. Results: From Oct. 2020 to Oct. 2021, 15 patients with T-cell acute lymphoblastic leukemia (n=10), T-cell lymphoblastic lymphoma (n=3), and mixed phenotype acute leukemia (n=2) were enrolled and received KO7 CAR-T cells (Table 1). The median age was 28 (8-46) years old. Four patients had prior hematopoietic stem cell transplantation (HSCT). At enrollment, 10 patients had bone marrow (BM) blasts >5% by morphology, and 10 patients had the extramedullary disease (EMD, diffuse involvement, n=8, and bulky mediastinal masses, n=2). Both patient- and donor-derived KO7CAR-T cells were successfully generated with a transduction efficiency of 59.0% (22.5%-97.4%). A single dose of KO7CAR-T cells was infused to patients at low dose (1.5~5x10^5 cells/kg, n=8), medium dose (1x10^6 cells/kg, n=6) or high dose (2x10^6 cells/kg, n=1). On day 28, 15/15 (100%) patients achieved minimal residual disease (MRD) negative complete remission (CR). Among the 10 patients with EMD, 7 achieved EMD CR on day 30, 2 achieved partial response (PR), and 1 who relapsed post 2^nd transplant had no response on day 35 then withdrew. Up to data cutoff Feb.10, 2022, the median follow-up time was of 309 days (35~407 days). About 2 months post KO7CAR, 12 patients bridged into allogeneic HSCT, and all remained progression-free after a median time of 253 (30~388) days after HSCT except for 1 who relapsed on day 147 then died from intracerebral hemorrhage on day 249. The other 2 patients without subsequent HSCT (all had a prior transplant) died from infection on day 78 and GVHD on day 103, respectively, post KO7CAR. Mild cytokine release syndrome (CRS, ≤grade II) occurred in 10/15 (66.7%) patients, and 5/15(33.3%) patients had grade III CRS. One patient had grade I neurotoxicity, and 2 had grade III/IV neurotoxicity. All was controlled after the administration of corticosteroids and/or tocilizumab. Following infusion, the median peak of circulating KO7CAR-T cells was 1.77×10^5 (0.279~14.3×10^5) copies/μg genomic DNA which occurred around day 20 (day10~ day25) and 63.47% (23.1%~94.18%) occurring on day15 (day 9~day25) by q-PCR and flow cytometry respectively. Image: [33]graphic file with name hs9-6-1-g009.jpg [34]Open in a new tab Summary/Conclusion: This study demonstrated KO7CAR-T therapy had a high efficacy for CD7+ T-cell malignancies even for those who relapsed post-transplant. Safety was manageable, however, more data on additional patients and longer observation time are needed to evaluate the efficacy of KO7 CAR-T products further. S111: REPEATED INFUSIONS OF ESCALATING DOSES OF EXPANDED AND ACTIVATED AUTOLOGOUS NATURAL KILLER CELLS IN MINIMAL RESIDUAL DISEASE-POSITIVE PH+ ACUTE LYMPHOBLASTIC LEUKEMIA PATIENTS. A GIMEMA PHASE 1 TRIAL G. F. Torelli^1,*, S. Chiaretti^1, N. Peragine^1, W. Barberi^1, L. Santodonato^2, G. D’Agostino^2, E. Abruzzese^3, M. I. Del Principe^4, A. Mancino^5, M. Matarazzo^1, M. S. Bafti^1, M Mancini^1, M. Messina^5, L. Castiello^2, A. Guarini^1, R. Foà^1 ^1Hematology, Department of Translational and Precision Medicine, Sapienza University; ^2FaBioCell Cell Factory, Istituto Superiore di Sanità; ^3Hematology, Sant’Eugenio Hospital, ASL Roma^2, Tor Vergata University; ^4Hematology, Department of Biomedicine and Prevention, Tor Vergata University; ^5Fondazione GIMEMA Onlus, Rome, Italy Background: Due to age and co-morbidities, many Philadelphia-positive acute lymphoblastic leukemia (Ph+ ALL) patients are ineligible to undergo high-dose chemotherapy or allogeneic transplant as consolidation treatment. Our group reported the promising results of the chemo-free scheme D-ALBA based on dasatinib/blinatumomab in induction/consolidation, underlying the potential role of immunotherapy in this setting (Foà et al, NEJM 2020;383:1616-23). Considerable interest has been raised by natural killer (NK) cells. We developed a GMP protocol for NK cell ex vivo expansion in the presence of IL-2 and IL-15, and report the results of a phase 1 protocol of adoptive immunotherapy with activated and expanded autologous NK cells for Ph+ ALL patients in complete hematologic remission (CHR) but with persistent/recurrent minimal residual disease (MRD) ≥60 years or ineligible for other post-CHR treatment modalities. Aims: The primary endpoint was to determine the maximum tolerated dose of NK cells and the recommended dose for subsequent studies. Secondary endpoints were the assessment of safety and tolerability of the treatment, the immunologic modifications induced by the procedure and the clinical response to treatment. Methods: The planned 6 patients were enrolled: 5 in 1^st CHR and 1 in 2^nd CHR. Patients underwent repeated infusions (maximum 5) of escalating doses of NK cells, ranging from 1x10^6 to 5x10^7/kg of body weight (BW). No conditioning therapies were administered before the infusion; patients were allowed to continue tyrosine kinase inhibitors. Patients underwent a comprehensive MRD monitoring by Q-RT-PCR with a one-year follow-up. Immunophenotypic analysis on the NK cell product was performed before and after the expansion. Intracellular cytokine production and PBMC cytotoxic activity against K562 cells, allogeneic and autologous blasts were evaluated after expansion and at time 0 and 7 days from each NK cell infusion. Results: NK cells presented a 12.3-fold ex vivo expansion. Expanded cells showed an increased expression of activating receptors and measurable cytotoxicity against primary allogeneic and autologous blasts. One patient received a maximum NK cell dose of 5x10^6 cells/kg, 2 patients 1x10^7 cells/kg and 3 5x10^7 cells/kg/BW. No patient experienced infusion-related toxicities. Two adverse events were recorded (grade 1 and 2), both judged not treatment-related, that resolved after TKI suspension. The higher cell dose infusion resulted in a significantly increased expression of natural cytotoxicity receptors, a greater cytokine production by NK, T and NKT cells, and in an increased capacity of PBMC to lyse K562 cells. These modifications appear persistent over time. At a 1-year follow-up from the last infusion, 5/6 patients are alive in CHR (Table 1). The MRD levels reduced over time and 4/6 patients reached a complete molecular response (CMR) or a positive-not-quantifiable (PNQ) status during the study period. At a median follow-up of 30.8 months from the last infusion, the 5 patients who received the NK treatment in 1^st CHR are still in CMR or PNQ, though 1 patient required additional treatment. The patient in 2^nd CHR at the time of the infusions showed a rise in MRD and died of disease progression. Image: graphic file with name hs9-6-1-g010.jpg [35]Open in a new tab Summary/Conclusion: This phase 1 study demonstrates that autologous NK cells can be efficiently expanded ex vivo from MRD-positive Ph+ ALL patients in CHR. The infusion of these expanded cells is safe and induces a marked in vivo host immune response, suggesting that this approach represents a tolerable and feasible model worthy of being investigated in larger clinical studies. S112: TISAGENLECLEUCEL IN PEDIATRIC AND YOUNG ADULT PATIENTS (PTS) WITH RELAPSED/REFRACTORY (R/R) B-CELL ACUTE LYMPHOBLASTIC LEUKEMIA (B-ALL): FINAL ANALYSES FROM THE ELIANA STUDY S. Rives^1,*, S. L. Maude^2, H. Hiramatsu^3, A. Baruchel^4, P. Bader^5, H. Bittencourt^6, J. Buechner^7, T. Laetsch^2, B. De Moerloose^8, M. Qayed^9, H. E. Stefanski^10, K. L. Davis^11, P. L. Martin^12, E. Nemecek^13, C. Peters^14, G. Yanik^15, A. Balduzzi^16, N. Boissel^17, S. L. Khaw^18, J. Krueger^19, J. Levine^20, S. Davies^21, G. D. Myers^22, A. Yeo^23, D. O’Donovan^24, R. Ramos^23, M. Pulsipher^25, S. Grupp^2 ^1Department of Pediatric Hematology – Oncology, Hospital Sant Joan de Déu Barcelona, and Institut de Recerca Sant Joan de Déu, Barcelona, Spain; ^2Division of Oncology, Center for Childhood Cancer Research and Cancer Immunotherapy Program, Children’s Hospital of Philadelphia and Department of Pediatrics, Abramson Cancer Center, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, United States of America; ^3Department of Pediatrics, Graduate School of Medicine, Kyoto University, Kyoto, Japan; ^4University Hospital Robert Debré (APHP) and Université de Paris, Paris, France; ^5Division of Stem Cell Transplantation and Immunology, Hospital for Children and Adolescents, University Hospital Frankfurt, Frankfurt, Germany; ^6Department of Pediatrics, Faculty of Medicine, University of Montreal, and the Hematology Oncology Division and Charles-Bruneau Cancer Center, Centre Hospitalier Universitaire Sainte-Justine Research Centre, Montreal, QC, Canada; ^7Department of Pediatric Hematology and Oncology, Oslo University Hospital, Oslo, Norway; ^8Department of Pediatric Hematology-Oncology and Stem Cell Transplantation, Ghent University Hospital, Ghent, Belgium; ^9Aflac Cancer and Blood Disorders Center, Emory University, Atlanta, GA; ^10National Bone Marrow Donor Program, Be the Match, Division of Pediatric Blood and Marrow Transplant, University of Minnesota, Minneapolis, MN; ^11Division of Hematology, Oncology and Stem Cell Transplant, Department of Pediatrics, Stanford University School of Medicine, Stanford, CA; ^12Pediatric Transplant and Cellular Therapy, Duke University Medical Center, Durham, NC; ^13Oregon Health and Science University, Portland, OR, United States of America; ^14Stem Cell Transplantation Unit, St. Anna Children’s Hospital, Vienna, Austria; ^15Department of Pediatrics, University of Michigan Medical Center, Ann Arbor, MI, United States of America; ^16Clinica Pediatrica Università degli Studi di Milano Bicocca, Fondazione MBBM, Ospedale San Gerardo, Monza, Italy; ^17Saint-Louis Hospital (APHP) and Université de Paris, Paris, France; ^18Children’s Cancer Centre, Royal Children’s Hospital and Murdoch Children’s Research Institute, Parkville, VIC, Australia; ^19Division of Haematology/Oncology/Bone Marrow Transplantation, Department of Paediatrics, The Hospital for Sick Children, Toronto, ON, Canada; ^20Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York, NY; ^21Cincinnati Children’s Hospital Medical Center, Cincinnati, OH; ^22Children’s Mercy Hospital and Clinics, Kansas City, MO; ^23Novartis Pharmaceuticals Corporation, East Hanover, NJ, United States of America; ^24Novartis Pharmaceuticals Corporation, Dublin, Ireland; ^25Division of Pediatric Hematology and Oncology, Intermountain Primary Children’s Hospital, Huntsman Cancer Institute at the University of Utah, Salt Lake City, UT, United States of America Background: Pediatric and young adult pts with R/R B-ALL experience a treatment journey characterized by diminishing likelihood of cure and increasing morbidity. Tisagenlecleucel is an autologous CD19-directed chimeric antigen receptor (CAR) T-cell therapy approved for use in pediatric and young adults with B-ALL and adults with B-cell lymphomas. Tisagenlecleucel provided high rates of remission (>80%) in children and young adults with R/R B-ALL in ELIANA, with 62% of responders remaining relapse-free at 24 mo (Grupp et al, Blood, 2018). Aims: Here, we report the final efficacy and safety analyses in pts followed up to 5.9 years post-tisagenlecleucel infusion. Methods: ELIANA ([36]NCT02435849) was a pivotal, Phase II, open-label, multicenter, global study of tisagenlecleucel in pediatric and young adult pts with R/R B-ALL. Pts received a single infusion of tisagenlecleucel at 0.2-5.0×10^6 CAR+ viable T cells/kg body weight for pts ≤50 kg and 0.1-2.5×10^8 CAR+ viable T cells for pts >50 kg. Endpoints included overall remission rate (ORR) within 3 mo, relapse-free survival (RFS), duration of remission (DOR), overall survival (OS), persistence of B-cell aplasia, and short- and long-term safety events. Results: Results: As of September 24, 2021, 97 pts were enrolled and 79 pts (81%) received tisagenlecleucel. Median time from infusion to data cutoff was 5.5 y; 64 pts had ≥5 y of follow-up. At study entry, the median age was 11 y (range, 3-24). Pts were heavily pretreated with a median of 3 prior lines of therapy (range, 1-8) and 61% had a history of prior stem cell transplant (SCT). ORR (complete remission [CR] or CR with incomplete hematologic recovery within 3 mo after infusion) was 82% (95% CI, 72-90). Among pts in remission (CR/CRi), the 5y RFS rate was 49% (95% CI, 34-62), and the median RFS was not reached (Figure, 46.8 mo when censoring for SCT; n=15). The median time to B-cell recovery was 38.6 mo (95% CI, 23-not reached) and the probability of B-cell aplasia at 6 mo and 12 mo was 83% (95% CI, 71-91) and 71% (95% CI, 57-82), respectively. Pts with B-cell recovery (<6 mo, n=10; 6-12 mo, n=4; >12 mo, n=7) experienced a 2y cumulative incidence of relapse of 25.2% (with SCT treated as a competing risk). Among all pts, the 5y EFS and OS rates were 42% (95%CI, 29-54) and 55% (95% CI, 43-66), respectively. There were no significant differences in any efficacy endpoint between pediatric (<18 y; n=65) and young adult (≥18 y; n=14) pts. No new or unexpected AEs were reported during long-term follow-up. Among pts in remission, the most commonly reported grade ≥3 AEs occurring >1 y post-infusion were infection (20%) and cytopenias (6%). Ten (14%) pts in remission experienced long-term cytopenias persisting for >1 y; however, none of these pts experienced cytopenias persisting for >5 y (median 2 y; range, 1.1-5y). Eighty-two percent of pts received IVIG any time post-infusion. Image: graphic file with name hs9-6-1-g011.jpg [37]Open in a new tab Summary/Conclusion: This >5 y follow-up study demonstrates continued durable efficacy of tisagenlecleucel without late adverse effects in heavily pretreated pediatric and young adult pts with R/R B-ALL. Tisagenlecleucel continues to be a potentially curative treatment option for pediatric and young adult patients with R/R B-ALL. S113: NATIONAL PEGASPARGASE-MODIFIED RISK-ORIENTED PROGRAM FOR PHILADELPHIA-NEGATIVE ADULT ACUTE LYMPHOBLASTIC LEUKEMIA/LYMPHOBLASTIC LYMPHOMA (PH− ALL/LL). GIMEMA LAL 1913 FINAL RESULTS. R. Bassan^1,*, S. Chiaretti^2, I. Della Starza^3, O. Spinelli^4, A. Santoro^5, L. Elia^3, M. S. De Propris^3, A. M. Scattolin^1, F. Paoloni^6, M. Messina^7, E. Audisio^8, L. Marbello^9, E. Borlenghi^10, P. Zappasodi^11, C. Vetro^12, G. Martinelli^13, D. Mattei^14, N. Fracchiolla^15, M. Bocchia^16, P. De Fabritiis^17, M. Bonifacio^18, A. Candoni^19, V. Cassibba^20, P. Di Bartolomeo^21, G. Latte^22, S. Trappolini^23, A. Guarini^24, A. Vitale^3, P. Fazi^6, M. Vignetti^6, A. Rambaldi^4, R. Foà^3 ^1Hematology, Ospedale dell’Angelo, Venice; ^2Translational and Precision Medicine, Sapienza Univesrity of Rome; ^3Translational and Precision Medicine, Sapienza University of Rome, Rome; ^4UOC Ematologia, ASST-Papa Giovanni XXIII, Bergamo; ^5Divisione di Ematologia con UTMO, Ospedali Riuniti Villa Sofia-Cervello, Palermo; ^6GIMEMA Data Center; ^7GIIMEMA Data Center, Fondazione GIMEMA – Franco Mandelli Onlus, Rome; ^8Ematologia, Città della Salute, Torino; ^9Hematology, Niguarda Ca’ Granda Hospital, Milan; ^10Hematology, Spedali Civili, Brescia, Italy, Spedali Civili, Brescia; ^11Hematology, Foundation IRCCS Policlinico San Matteo, Pavia; ^12General Surgery and Medical-Surgical Specialties, University of Catania, Catania; ^13Institute of Hematology, L. and A. Seràgnoli, Bologna; ^14Hematology, Ospedale S. Croce, Cuneo, Italy, Cuneo; ^15UOC Oncoematologia, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico di Milano, Milan; ^16Hematology Unit, Azienda Ospedaliera Universitaria Senese, Siena; ^17Hematology Division, S. Eugenio Hospital, Rome; ^18Ospedale Policlinico “G.B. Rossi”, University of Verona, Verona; ^19Clinica Ematologica, Azienda Sanitaria Universitaria Integrata di Udine, Udine; ^20Divisione di Ematologia, Ospedale Civile, Bolzano; ^21Oncology Hematology, Ospedale Civile, Pescara; ^22Hematology, S. Franceso Hospital, Nuoro; ^23Clinica di Ematologia, Azienda Ospedaliero - Universitaria Ospedali Riuniti Umberto I, Ancona; ^24Molecular Medicine, Sapienza University of Rome, Rome, Italy Background: Pediatric-inspired chemotherapy is standard of care for younger adults with Ph− ALL/LL. An essential component of these regimens is pegaspargase, here incorporated into a national treatment program for patients 18-65 years. Aims: To assess in the GIMEMA Phase 2 LAL 1913 study the feasibility and efficacy of a pegaspargase-containing induction and consolidation regimen sustaining a risk-oriented strategy for adult Ph− ALL/LL (ClinicalTrials.gov ID [38]NCT02067143). Methods: Our prior, reference 8-block chemotherapy protocol (Blood Cancer J 2020;10:119) was modified to include pegaspargase 2000 IU/m^2 at courses 1 (d10), 2 (d8), 5 (d3, with HD-MTX) and 6 (d8), with dose reductions in patients >55 years (pegaspargase 1000 IU/m^2). Serum drug activity was not assessed in this study. Responders were risk-stratified for allogeneic stem cell transplantation (SCT) or maintenance according to a mixed risk model based on WBC count, immunophenotype, genetics and post-remission molecular minimal residual disease (MRD): patients with high-risk (HR) features or MRD ≥ 10^-4 at weeks 10-16 or positive at week 22 were eligible to SCT; standard-risk (SR) patients were eligible to maintenance. Results: Two hundred and three patients entered the study (median age 39.8 years; 139 B- and 64 T-phenotype). The complete remission (CR) rate was 91% (100% in T-ALL/LL), with a 3-year cumulative relapse incidence and non-relapse mortality of 24.2% and 12.6%, respectively; 60 patients underwent a SCT. Overall (OS), event-free (EFS) and disease-free (DFS) survival were 66.7% (95% CI, 60.1-74.1%), 57.7% (95% CI, 51.0-65.3%) and 63.3% (95% CI, 56.3-71.1%) at 3 years. HR class (n=95) and LL diagnosis (n=20) did not affect prognosis. T-cell phenotype (CR 100%, P=0.001; EFS 67.1%, P=0.038), age 18-40 years (EFS 72.6%, P<0.0001) and MRD <10^-4 after courses 1 (55%: DFS 77.9%, P=0.023) and 3 (79%: DFS 75.2%, P=0.048) were prognostically favorable. One hundred and eighty-seven patients had pegaspargase at course 1 (92.1%, 11 delayed, 3 reduced), 154 at course 2 (84.6%; 11 delayed, 12 reduced), 110 at course 5 (83.9%; 2 delayed, 11 reduced) and 73 at course 6 (68.8%; 3 delayed, 7 reduced). Dose reductions and delays were related to high-risk profile (liver dysfunction/steatosis, obesity etc.) or treatment toxicity. Toxicity of grade 2 or more was mainly observed at course 1 (hepatic 12.8%, coagulation/thrombosis 3.2% [enoxaparin prophylaxis recommended with platelets >30-50], pancreatic 1.6%), contributing to an induction death in 3 patients (1.4%), but was rare afterwards. Image: graphic file with name hs9-6-1-g012.jpg [39]Open in a new tab Summary/Conclusion: This pegaspargase-based ALL regimen was safely applicable to the majority of study patients, resulting in 3-year OS, EFS and DFS rates >50% in a patient population aged 18-65. The results were more favorable in patients up to the age of 55, especially in those aged 18-40 years, and in those who achieved maximum MRD response regardless of age (Figure). Subsequently, a pegaspargase dosing algorithm based on patient age, body mass index, hepatosteatosis and selected toxicities at first or prior drug exposure was developed to minimize toxicity, and was used in a successor GIMEMA trial of sequential chemotherapy-blinatumomab for CD19+ adult B-ALL (EHA Congress 2021, abstract S114). S114: PONATINIB AND BLINATUMOMAB FOR PATIENTS WITH PHILADELPHIA CHROMOSOME-POSITIVE ACUTE LYMPHOBLASTIC LEUKEMIA: UPDATED RESULTS FROM A PHASE II STUDY N. Short^1,*, H. Kantarjian^1, M. Konopleva^1, N. Jain^1, F. Ravandi^1, X. Huang^2, W. Macaron^1, W. Wierda^1, G. Borthakur^1, T. Kadia^1, K. Sasaki^1, G. Issa^1, G. Montalban-Bravo^1, Y. Alvarado^1, G. Garcia-Manero^1, C. Dinardo^1, J. Thankachan^1, R. Delumpa^1, E. Mayor^1, W. Deen^1, A. Milton^1, J. Rivera^1, L. Waller^1, C. Loiselle^1, R. Garris^1, E. Jabbour^1 ^1Department of Leukemia; ^2Department of Biostatistics, The University of Texas MD Anderson Cancer Center, Houston, United States of America Background: Ponatinib and blinatumomab are both highly effective therapies for Philadelphia chromosome-positive (Ph+) acute lymphoblastic leukemia (ALL). The combination of these two agents may offer an effective chemotherapy-free strategy in these patients (pts). Aims: We evaluated the efficacy and safety of ponatinib and blinatumomab in pts with newly diagnosed (ND), relapsed/refractory (R/R) Ph+ ALL or CML in lymphoid blast phase (CML-LBP). For pts with ND Ph+ ALL, the primary endpoint was the complete molecular response (CMR) rate. For pts with R/R Ph+ ALL, the primary endpoint was the CR/CRi rate. Secondary endpoints included safety, event-free survival (EFS) and overall survival (OS). Methods: In this phase II study, adults with ND Ph+ ALL, R/R Ph+ ALL, or CML-LBP were eligible. Pts were required to have a performance status of ≤2, total bilirubin ≤2x the upper limit of normal (ULN), and ALT and AST ≤3x the ULN. Pts with uncontrolled cardiovascular disease or clinically significant central nervous system (CNS) comorbidities (except for CNS leukemia) were excluded. Pts received up to 5 cycles of blinatumomab as a continuous infusion at standard doses. Ponatinib 30mg daily was given during cycle 1 and was decreased to 15mg daily once CMR was achieved. After 5 cycles of blinatumomab, ponatinib was continued for at least 5 years. Twelve doses of prophylactic IT chemotherapy with alternating cytarabine and methotrexate were administered. Results: Between 2/2018 to 1/2022, 55 pts were treated (35 with ND Ph+ ALL, 14 with R/R Ph+ ALL and 6 with CML-LBP). Baseline characteristics are shown in Table 1. Among the 35 pts with ND Ph+ ALL, 12 were in CR at enrollment (including 2 pts in CMR). 22 of the 23 evaluable pts (96%) achieved CR/CRi. One pt died on day 18 from intracranial hemorrhage in the setting of chemotherapy administered prior to enrollment. After one cycle, 21/33 pts (64%) achieved CMR, and 28/33 pts (85%) achieved CMR at any time. 11 of 15 tested pts (73%) also became MRD-negative by an NGS assay with sensitivity of 1x10^-6. CR/CRi was achieved in 12/13 (92%) evaluable pts with R/R Ph+ ALL. CMR was achieved in 10 pts (71%) after cycle 1 and in 11 pts (79%) overall. 5 of 6 pts with CML-LBP achieved CR/CRi, and 1 pt achieved PR as best response. 2 pts (40%) achieved CMR. In the ND Ph+ ALL cohort, 1 of 34 pts who received at least 1 full cycle died in CR; the other 33 are in ongoing hematologic remission. Only one pt underwent stem cell transplant (SCT) in first remission for persistently detectable BCR/ABL1 transcripts. Among 13 responding pts in the R/R Ph+ ALL cohort, 6 proceeded to SCT, 4 did not undergo SCT and subsequently relapsed, 1 died in CR, and 2 are in ongoing remission without SCT. In the CML-LBP cohort, 3 of the 5 responding pts subsequently relapsed. The median follow-up is 11 months (range, 1-46+). For ND Ph+ ALL, the 2-year EFS and OS are both 93% (Figure 2). There were no relapses or leukemia-related deaths in this cohort. In the R/R Ph+ ALL cohort, the 2-year EFS rate was 42% and the 2-year OS rate was 61%. In the CML-LBP cohort, the 2-year EFS was 33% and the 2-year OS was 60%. The treatment was well-tolerated, and most toxicities were grade 1-2 and consistent with the known toxicities of the two agents. Two pts discontinued ponatinib due to toxicity (1 due to stroke and 1 due to DVT). One pt discontinued blinatumomab due to persistent grade 2 tremor. Image: graphic file with name hs9-6-1-g013.jpg [40]Open in a new tab Summary/Conclusion: The chemotherapy-free regimen of simultaneous ponatinib and blinatumomab is safe and effective in pts with Ph+ ALL. For pts with ND Ph+ ALL, SCT does not appear to be needed in first remission. S115: MUTANT NPM1 BINDS CHROMATIN AND COOPERATES WITH MLL1 TO REGULATE ONCOGENIC TRANSCRIPTION H. Uckelmann^1,*, S. Armstrong^1, E. Haarer^1, E. Wong^1, C. Hatton^1, F. Perner^1, C. Marinaccio^1, C.-W. Chen^2 ^1Pediatric Oncology, Dana-Farber Cancer Institute, Boston; ^2Department of Systems Biology, Beckman Research Institute, City of Hope, United States of America Background: The dysregulation of stem cell and self-renewal associated genes is a common phenomenon during leukemia development. In acute myeloid leukemia (AML) around 50 % of cases express high levels of HOXA cluster genes and MEIS1. Most of these AML cases harbor an NPM1 mutation (NPM1c), which encodes for an oncogene that is mislocalized from the nucleolus to the cytoplasm. Hence, most studies of NPM1c have focused on a potential cytoplasmic role. However, it remains unclear how NPM1c expression in hematopoietic cells leads to its characteristic gene expression pattern. Furthermore, NPM1c AMLs are highly sensitive to the disruption of the MLL1 histone methyltransferase complex. Small molecule inhibitors that block the interaction between MLL1 and its adaptor protein Menin have been shown to impair binding of MLL1 to a subset of its target genes and to inhibit leukemia cell proliferation and self renewal. Several MLL1-Menin inhibitors are currently in Phase I/II clinical trials and show promising activity in patients with NPM1c AML. The effectiveness of these molecules in NPM1c AML prompts the question whether NPM1c and the wildtype MLL complex cooperate directly on chromatin to drive leukemic self-renewal. Aims: In this study we investigated the potential role of NPM1c in regulating oncogenic transcription on chromatin and the interplay between NPM1c and the histone methyltransferase complex KMT2A (MLL1). Methods: We used an endogenously degrader tagged NPM1c leukemia cell line that allows rapid small molecule induced degradation to show that NPM1c occupies specific chromatin targets in AML. To characterize the effects of NPM1c degradation on the chromatin landscape and transcriptional output at genomic loci that are bound by NPM1c we used ChIPseq, PROseq and nascent RNAseq methods. Results: Our results show that endogenous NPM1c directly binds to chromatin at specific target genes, such as HOXA9 and MEIS1, which are highly expressed in NPM1c patient samples. The loss of NPM1c from its targets leads to specific alterations in active chromatin marks and RNA Polymerase II (Pol II) chromatin occupancy which are accompanied by rapid changes in gene expression as well as Pol II transcriptional activity. The recruitment of NPM1c to chromatin is dependent on the nuclear exporter CRM1 as well as one of the acidic domains of NPM1c. We further show that NPM1c is lost from specific loci after treatment with small molecules that disrupt the MLL1-Menin complex interaction thus functionally linking targeted epigenetic therapy and NPM1c function. Summary/Conclusion: Overall, we demonstrate that NPM1c directly regulates a network of leukemia self-renewal associated genes through direct chromatin interaction. We further found that NPM1c acts in collaboration with the MLL1 complex and define the mechanism by which MLL1-Menin small molecule inhibitors produce clinical responses in patients with NPM1-mutated AML. S116: CELLULAR AND MOLECULAR MECHANISMS OF EVI1-EXPRESSING MLL-REARRANGED ACUTE MYELOID LEUKEMIA Hugues-Etienne Châtel-Soulet^1, Sabine Juge^1, Ana Luisa Pereira^2, Frederik Otzen Bagger^3, Alexandar Tzankov^4, Mineo Kurokawa^5, Athimed El Taher^6, Jonathan Seguin^1, César Nombela Arrieta^2, Juerg Schwaller^1 ^1Biomedicine, University Children’s Hospital, Basel, Switzerland; ^2Medical Oncology & Hematology, University Hospital, Zürich, Switzerland; ^3Center for Genomic Medicine, University Hospital, Kopenhagen, Denmark; ^4Institute for Pathology, University Hospital, Basel, Switzerland; ^5Hematology & Oncology,The University of Tokyo, Tokyo, Japan; ^6Biomedicine, University of Basel, Basel, Switzerland Background: Expression of a doxycycline (DOX)-inducible acute myeloid leukaemia (AML)-associated iMLL-AF9 fusion transgene in long-term haematopoietic stem cells (LT-HSC) can lead to an invasive and chemo-resistant disease expressing the transcription factor EVI1. High EVI1 expression has been suggested as marker of poor outcome in AML patients even without rearrangements of the EVI1 locus at 3q26. Aims: We addressed the association between EVI1 expression, the cellular origin and poor disease outcome in AML driven by the iMLL-AF9 fusion gene. Methods: The role of EVI1 expression was studied in iMLL-AF9 transgenic mice carrying an Evi1-IRES GFP reporter in vitro using flow cytometry, colony formation and RT-qPCR assays, ex vivo with high-resolution bone marrow (BM) imaging, and in vivo, by transplantation of enriched naïve Evi1+ iMLL-AF9 hematopoietic stem and progenitor cells (HSPC) into irradiated recipients on DOX. Haematopoiesis of symptomatic mice was analysed by flow cytometry and histology. For mechanistic studies, single cell and bulk RNA sequencing was performed on enriched HSPC or BM samples from diseased mice. Results: Analysis of BM cells from Evi1-IRES-GFP reporter mice revealed that not only the mostly quiescent LT-HSC but also fractions of the more proliferating multipotent progenitors (MPP1-3) express abundant Evi1 (“Evi1high”). Induction of the iMLL-AF9 fusion did not result in significant changes in numbers of Evi1+ cells nor levels of Evi1 mRNA expression in the LT-HSC and MPP1 compartments. However, in colony assays, Evi1high iMLL-AF9 cells retained a more immature phenotype and produced more colonies with an invasive morphology than Evi1low cells (n=11, p<0.05). While Evi1 expression did not influence disease induction upon transplantation of LT-HSC, recipients of Evi1+ MPP1 cells developed AML earlier than Evi1- MPP1 (n=11, 79 vs. 269d, p<0.05). Disease induced by Evi1+ cells presented with more extensive leukemic organ infiltration than Evi1- AML. Evi1 expression also correlated with in vitro Ara-C resistance. We also examined whether some exogenous factors may increase AML susceptibility by expanding the Evi1+ HSPC. Although a single injection of recombinant mouse thrombopoietin (TPO) only increased the number of LT-HSC and not of MPP1, the Evi1high cell fraction was enlarged in both compartments (LT-HSC: 23 vs. 50%; MPP1: 22 vs. 47%; n=29, p<0.0001) supported by high-resolution imaging. Interestingly, increased TPO induced HSPC cycling was confined to the Evi1high cell population (n=3, p<0.05). Transplantation of TPO-treated iMLL-AF9 LT-HSC or MPP1 resulted in a significantly faster induction of Evi1+ AML than controls (n=19, MPP1: 35 vs. 79d, p<0.001; LT-HSC: 41 vs. 90d, p<0.001). To better understand mechanisms of aggravated AML after TPO mediated expansion of Evi1+ HSCP we performed multiplexed single cell RNA sequencing of highly enriched HSPC cells in iMLL-AF9 and control mice. While we observed no changes of cellular cluster organisation 2 days after TPO injection, we found some differentially expressed genes in TPO-stimulated cycling iMLL-AF9 HSPC, including potential stemness regulators. Summary/Conclusion: Our results suggest that expansion of Evi1-expressing HSPC by exogenous factors can result in a more aggressive MLL-AF9-driven AML. Ongoing data exploration and validation may characterize aberrantly expressed genes in TPO-stimulated Evi1+ iMLL-AF9-expressing HSPC as potential therapeutic targets to impair stemness of AML cells. S117: EVI1 DRIVES LEUKEMOGENESIS THROUGH ABERRANT ERG ACTIVATION J. Schmoellerl^1,*, I. Barbosa^1, M. Minnich^1, F. Andersch^1, L. Smeenk^2, M. Havermans^2, T. Eder^3, T. Neumann^1, J. Jude^1, M. Fellner^1, A. Ebert^1, M. Steininger^1, R. Delwel^2, F. Grebien^3, J. Zuber^1 ^1Research Institute of Molecular Pathology (IMP), Vienna, Austria; ^2Department of Hematology, Erasmus MC Cancer Institute, Rotterdam, Netherlands; ^3Institute for Medical Biochemistry, University of Veterinary Medicine Vienna, Vienna, Austria Background: Chromosomal rearrangements leading to overexpression of EVI1 (MECOM) on chromosome 3q26 define a distinct subtype of acute myeloid leukemia (AML) that is associated with chemotherapy resistance and a 2-year survival of <10%. While genetic events driving aberrant expression of EVI1 are increasingly understood, the molecular functions of EVI1 that drive leukemogenesis are unclear, which has so far precluded the development of targeted therapeutics. Aims: We aimed to elucidate transcriptional programs that are maintained by aberrant EVI1 expression and to systematically identify vulnerabilities of EVI1-driven AML. Methods: We developed a panel of mouse models that recapitulate phenotypic and transcriptional hallmarks of patients suffering from EVI1-driven AML, allow tetracycline-controllable EVI1 expression and the functional interrogation of genetic targets using CRISPR/Cas9. We mapped transcriptional programs upon acute EVI1 repression in vivo and in vitro, profiled global EVI1 chromatin occupancy in human AML cell lines and primary patient-derived AML cells and performed comparative genome-wide CRISPR/Cas9-based loss-of-function screens in murine and human EVI1-driven AML. Results: Integration of these datasets revealed a conserved core of genes that is transcriptionally regulated by EVI1 in murine and human AML, among which we identified the ETS transcription factor ERG as the only dependency that is highly selective for EVI1-driven AML. Suppression of ERG specifically triggered cellular differentiation and apoptosis of EVI1-driven leukemia cells while other AML cell lines were unaffected. Strikingly, ectopic expression of ERG was sufficient to functionally rescue loss of EVI1 in EVI1-driven AML cells, suggesting that the major oncogenic function of EVI1 in AML is the aberrant activation of ERG. Summary/Conclusion: Interfering with the EVI1/ERG regulatory axis may provide entry points for the development of rational targeted therapies that are urgently needed for this group of AML patients. S118: IDENTIFICATION OF DIRECT TRANSCRIPTIONAL TARGET GENES OF NUP98-KDM5A REVEALS REGULATORY NETWORKS IN ACUTE MYELOID LEUKEMIA S. Troester^1,*, J. Schmoellerl^2, T. Eder^1, G. Manhart^1, G. Winter^3, J. Zuber^2, F. Grebien^1 ^1Institute for Medical Biochemistry, University for Veterinary Medicine Vienna; ^2Research Institute of Molecular Pathology (IMP); ^3Research Center for Molecular Medicine of the Austrian Academy of Sciences (CeMM), Vienna, Austria Background: Oncogenic fusion proteins involving the Nucleoporin 98 (NUP98) gene are recurrently found in acute myeloid leukemia (AML) with a particular prevalence in pediatric patients. A chromosomal rearrangement resulting in the fusion of NUP98 to the gene encoding the lysine-specific demethylase 5A (KDM5A) is the most frequent NUP98-fusion in infant leukemia and is associated with particularly poor prognosis. The urgent need for the development of tailored treatments requires a better understanding of the effects of NUP98-KDM5A on the deregulation of gene expression programs. Although it has been shown that oncogenic NUP98-fusion proteins act as transcriptional regulators, it is unclear if and how NUP98-KDM5A directly regulates gene expression to drive leukemia. Aims: In this study, we aimed to identify immediate critical effectors of the NUP98-KDM5A fusion protein and to characterize the transcriptional programs through which they regulate the development and maintenance of NUP98-KDM5A-driven AML. Methods: We conducted a genome-scale CRISPR/Cas9 loss-of-function screen in a NUP98-KDM5A-driven murine AML cell line to unravel functional genetic dependencies that could be exploited to target leukemia cells. In parallel, we developed a new model for degradation tag (dTAG)-mediated ligand-induced degradation of the NUP98-KDM5A protein to gain a detailed understanding of direct transcriptional effects of NUP98-fusion-dependent gene regulation. We used this model to measure immediate changes in transcription upon acute NUP98-KDM5A degradation by nascent RNA-seq (SLAM-seq). An inducible shRNA system was used to assess the requirements of direct NUP98-KDM5A target genes for leukemia cell growth and to measure global gene expression changes by RNA-seq. Results: Analysis of the CRISPR/Cas9 screen identified 4105 genes that are required for the proliferation and survival of NUP98-KDM5A-driven AML cells. Complete loss of the dTAG-NUP98-KDM5A fusion protein was achieved within one hour after ligand addition, resulting in cell cycle arrest, terminal differentiation and apoptosis of leukemia cells. Global analysis of nascent mRNA expression by SLAM-seq revealed 45 immediate NUP98-KDM5A target genes, as their transcription was significantly downregulated upon fusion protein degradation. Among these genes, 12 were classified as essential factors for NUP98-KDM5A cell growth from the CRISPR/Cas9 screen. This list included known target genes of NUP98-fusion proteins, such as members of the Hoxa gene cluster, but also other transcription factors, enzymes and RNA binding proteins. shRNA-mediated knockdown of the 12 candidate genes confirmed their essentiality for NUP98-KDM5A leukemia cell proliferation. Through RNA-seq studies, we found that the knockdown of a small subset among the 12 candidate genes was able to recapitulate global patterns of gene deregulation that are induced by NUP98-KDM5A knockdown. Summary/Conclusion: Using a combination of CRISPR/Cas9 screening and ligand-induced degradation, we identified direct transcriptional target genes of NUP98-KDM5A that are functionally essential in AML. Multi-layered investigations of the interplay between the members of this small network of genes using a variety of models including primary patient samples will allow us to further dissect their role in the regulation of aberrant gene expression in NUP98-KDM5A-expressing leukemia cells and might identify novel therapeutic targets. S119: CEBPA AND TET2 MUTATIONS COOPERATE TO INDUCE AGGRESSIVE AML VIA GATA-2 DOWNREGULATION E. Heyes^1,*, A. S. Wilhelmson^2 3 4, A. Wenzel^2 3 4, M. B. Schuster^2 3 4, M. Ali^3 4, T. D’Altri^2 3 4, T. Eder^1, G. Manhart^1, E. Rzepa^1, L. Schmidt^1, M. Meggendorfer^5, T. Haferlach^5, G. Volpe^6, C. Nerlov^7, J. Frampton^6, K. Jae Won^3 4, F. Grebien^1, B. Porse^2 3 4 ^1Institute of Medical Biochemistry, University of Veterinary Medicine Vienna, Vienna, Austria; ^2The Finsen Laboratory, Rigshospitalet, Faculty of Health Sciences; ^3Biotech Research and Innovation Center (BRIC); ^4Novo Nordisk Foundation Center for Stem Cell Biology, DanStem, Faculty of Health Sciences, University of Copenhagen, Copenhagen, Denmark; ^5MLL Munich Leukemia Laboratory, München, Germany; ^6Institute of Cancer and Genomic Sciences, College of Medical and Dental Sciences, University of Birmingham, Birmingham; ^7MRC Molecular Haematology Unit, Weatherall Institute of Molecular Medicine, University of Oxford, Oxford, United Kingdom Background: The transcription factor CCAAT-enhancer-binding protein alpha (C/EBPα) is a master regulator of granulopoiesis and is mutated in 10-15 % of Acute Myeloid Leukemia (AML) patients. N-terminal frameshifts represent the predominant type of lesions and ablate the expression of the full-length protein p42, leading to overexpression of the shorter isoform p30. Mutations in the C-terminal basic-region leucine zipper (bZip) can disrupt the DNA-binding ability of C/EBPα. AML patients harbor mono- or biallelic CEBPA mutations (CEBPAmo or CEBPAbi). The most commonly co-occurring mutations are loss-of-function mutations in the methylcytosine dioxygenase TET2, resulting in adverse overall survival. We hypothesized that combinatorial effects of CEBPA mutations together with TET2 loss specifically rewire transcriptional and epigenetic circuitries in AML cells, thereby strongly influencing disease outcome. Aims: We aimed to elucidate the molecular mechanisms behind cooperative effects of CEBPA and TET2 mutations through state-of-the-art transcriptomic and epigenomic analyses of relevant in vitro and in vivo models as well as data from AML patients. Methods: We used the CRISPR/Cas9 technology to introduce Tet2 mutations in murine cell lines expressing only p30 (Cebpa^p30/p30) or mimicking biallelic CEBPA mutations (Cebpa^p30/C-mut.) to study the functional cooperation of these mutations in vitro. A Cebpa^-/p30 Tet2^-/- mouse model was used to study effects of Tet2 loss in CEBPA mutated AML. We performed RNA-, C/EBPα-ChIP-, ATAC-, Bisulfite-, CUT&RUN and CRISPR/Cas9-mediated enhancer screens to generate a comprehensive dataset for in-depth comparative analysis and correlation with relevant patient data from the beatAML collection. Results: Integration of transcriptomic and epigenomic data from in vitro and in vivo models of CEBPA-TET2 co-mutated AML, in combination with gene expression analyses in AML patients, identified the transcription factor GATA2 as a conserved target of the CEBPA-TET2 axis. p30 and TET2 were strongly bound to the -77 kb enhancer of the Gata2 gene, and CRISPR/Cas9-induced enhancer deletions diminished Gata2 expression. Furthermore, TET2 loss reduced chromatin accessibility and increased DNA methylation of the Gata2 promoter, resulting in decreased Gata2 mRNA levels. RNAi-mediated silencing revealed a dose-dependent effect of Gata2 expression on leukemia cell fitness in vivo. Reduction of Gata2 levels by 25-50 % provided a strong competitive advantage to Cebpa^p30/p30 cells while near-complete downregulation of Gata2 expression (>75 %) dramatically reduced cellular fitness. Finally, treatment with the demethylating agent 5-azacytidine restored Gata2 expression in an AML model with Cebpa and Tet2 mutations and caused a significant survival benefit. Summary/Conclusion: The datasets generated from these models enable deeper insights into the epigenetic and transcriptomic changes that depend on CEBPA and TET2 mutations in a physiologically relevant mutational context. Our results reveal that mutational disruption of CEBPA and TET2 results in down-regulated GATA2 expression, causing aggressive AML. We propose a mechanism in which C/EBPα p30 mediates recruitment of TET2 to regulatory regions in the GATA2 gene to maintain its expression. Conversely, loss of TET2 leads to reduced GATA2 levels, which is restored by 5-azacytidine treatment. Thus, interference with the C/EBPα-TET2 axis may provide entry points for the development of rational targeted therapies in AML patients with these mutations. S120: ACUTE MYELOID LEUKEMIA REPRESENTS A FERROPTOSIS-SENSITIVE CANCER ENTITY RAISING THE POSSIBILITY FOR NOVEL TARGETING STRATEGIES A. Narr^1 2 3,*, H. Alborzinia^1 2, E. Donato^1 2, F. Vogel^4, T. Boch^1, A.-M. Leppä^1 2, A. Waclawiczek^1 2, S. Renders^1 2, A. Schulze^4, A. Trumpp^1 2 ^1Division of Stem Cells & Cancer, German Cancer Research Center (DKFZ); ^2Heidelberg Institute for Stem Cell Technology and Experimental Medicine (HI-STEM gGmbH); ^3University Heidelberg; ^4Division of Tumor Metabolismus und Microenvironment, German Cancer Research Center (DKFZ), Heidelberg, Germany Background: Despite advances in the treatment of Acute Myeloid Leukemia (AML), the 5-year patient survival rate remains poor with 15-20%. The majority of patients develop recurrence partly due to resistance mechanisms to classical cell death programs. Exploiting new forms of cell death therefore may allow novel therapeutic options to limit survival or occurrence of resistant clones. Ferroptosis has been identified as non-apoptotic and iron-dependent form of cell death, characterized by an excessive ROS-induced peroxidation of cell membranes. Recently it has been implicated with a higher sensitivity in cancer stem cells and drug-resistant cancer cells, potentially making ferroptosis-inducing (FIN) therapies a promising option. Aims: We aim to elucidate whether AML represents a ferroptosis-sensitive cancer entity and is susceptive for potential FIN-therapies. Methods: IC[50] values of ferroptosis inducers were determined and used in combination with the TCGA-LAML and a pan-cancer CRISPR screen dataset (CERES) to describe ferroptosis sensitivity of AML. A drug screen with known ROS-inducers was performed in 9 AML cell lines to identify ferroptosis induction. To mechanistically characterize ferroptotic cell death by the identified drug on the genetic and metabolic level, we performed SLAM-RNA-, RNA-Sequencing, Mass Spectrometry and Metabolite Tracing (Cystine/Ser/Gln) in HL-60 cells in vitro and in vivo. Genetic (CRISPR-Cas9-KO) and chemical inhibition were used to characterize the role of identified hits and pathways, aiming to design new synergistic combination therapies. Results: We find that AML belongs to the most dependent cancer entities for various ferroptosis-associated genes and represents a ferroptosis-sensitive cancer entity. Furthermore, a 12-gene-ferroptosis-signature allows us to predict the Overall Survival of AML patients and shall help to identify a potential cohort for FIN-therapies. By screening known ROS-inducing drugs for ferroptosis induction, we identify one drug capable of inducing lipid peroxidation and subsequently ferroptosis in multiple AML and other tumor cell lines. Notably, in detail characterization of the drug’s mechanism on the genetic and metabolic level reveal the activity of the iron-metabolism gene HMOX-1 as essential for ferroptosis induction. Mechanistically, we further find that this drug directly inhibits GPX-4, one of the major ferroptosis-suppressive regulators, and facilitates its degradation. Furthermore, we demonstrate that drug treatment results in an upregulation of different metabolic pathways involved in glutathione (GSH) synthesis; namely (1) SLC7A11-mediated Cystine-uptake, (2) transsulfuration pathway as well as (3) glutaminolysis. Genetical and chemical inhibition of these pathways together with drug treatment in in vitro as well as in vivo studies shows strong synergistic effects marking them as new targetable vulnerabilities in AML. Using these insights of the drug’s mechanism of action, we design a synergistic combination therapy and demonstrate its tumor-eradicating capacity in AML cell lines (in vitro & in vivo) and in primary de novo as well as relapse AML patient samples (ex vivo). Summary/Conclusion: These findings classify AML as ferroptosis sensitive and reveal the capacity of this drug to induce ferroptosis. Detailed characterization of the drug’s mechanism on the genetic and metabolic level allows us to design an effective combination therapy exploiting newly identified vulnerabilities in AML. These results provide a rational basis for the additional pre-clinical and subsequent clinical evaluation of FIN-therapies in AML patients. S121: CD123-CD33 COMPOUND CAR-T CELLS WITH NOVEL ANTIGEN BINDING DOMAINS PROVIDE A NEW HOPE FOR THE TREATMENT OF ACUTE MYELOID LEUKEMIA Z. Wang^1,*, Y. LU^1, S. Qiu^1, M. Wang^1, D. Xiong^1, J. Wang^1 ^1State Key Laboratory of Experimental Hematology, National Clinical Research Center for Blood Diseases, Tianjin Key Laboratory of Cell Therapy for Blood Diseases, Haihe Laboratory of Cell Ecosystem, Institute of Hematology and Blood Diseases Hospital, TianJin, China Background: CAR-T therapy needs to be optimized to treat myeloid malignancies. Antigen-escape mediated relapse after CAR-T treatment is the vital mechanism responsible for the nondurable response. Targeting CD33 and CD123 simultaneously could cover almost all patients with AML. The off-target cytotoxicity of CD123 CAR-T and CD33 CAR-T, especially on HSPCs, has always been concerned. Several researchers have explored this issue and the conclusions are inconsistent in different studies. Aims: To improve the outcome of relapse and/or refractory acute myeloid leukemia and explore the influence of CD123-CD33 CAR-T on hematopoiesis. Methods: We develpoed a panel of anti-CD123 monoclonal antibodies (mAbs) using hybridoma technology, named 6E11, 8D7, 12H7 and 13C3 and incorporated each of the four scFvs into 4-1BB/CD3ζ signaling domain to generate 4 second generation CARs. T cells were infected with lentiviral supernatants to generate CAR-T cells. Comparative analysis were made from the expression property of CAR protein, in vitro and in vivo anti-leukemia efficacy to screen out the best candidate CD123 CAR. Next, we generated a CD123-CD33 dual targeting CAR in a compound way. We established an escape-simulation model to explore whether the compound CAR-T could reduce the risk of relapse associated with antigen loss. Flowcytometry analysis, CFU test and single cell RNA sequecing methods were jointly used to explore the influence of our CD123-CD33 CAR-T on hematopoiesis. Results: CD123 CAR constructs with four different novel scFvs differed in CAR protein expression ratio. A changing trend of CAR proteins from dropping to rising after encountering target cells was observed in 3 CAR-T cell group except in 8D7 CAR-T. 13C3 CAR was the most suitable construct from the perspective of basic and dynamic expression. 123CAR-T cells with different scFvs mediated variable anti-leukemia effect in vitro and in vitro. 123CAR and 33CAR proteins expressed efficiently on the compound CAR-T cell surface without codon optimization. The compound CAR-T showed superiority in the treatment of antigen-escape mouse model. FCM analysis showed no significant difference of residual CD34^+ cells proportion between NCT cell group and compound CAR-T group. scRNA-seq analysis showed that the most undifferentiated HSC population barely changed among each group. GSEA revealed that the HSC population in CAR-T groups had no difference in the enrichment of quiescence, cell cycle, G2M checkpoint related genes compared with that of NCT group. In addition, no downregulation/enrichment of TGF-beta signaling pathway/ PI3K-AKT-mTOR signaling pathway was observed in compound CAR-T group compared with NCT group.The CFU assay demonstrated that there were still quite numbers of hematopoietic colonies formed in compound CAR-T treatment groups. Image: graphic file with name hs9-6-1-g014.jpg [41]Open in a new tab Summary/Conclusion: Taken together, in this study we developed a panel of second generation CD123 CARs with novel scFvs on the basis of 4 novel anti-CD123 mAbs. We proved that scFv sequence has a significantly impact on CAR expression both originally and dynamically and further affects CAR-T function, addressing the importance of scFv development in CAR-T therapy refining. The compound CAR-T in our study minimize the risk of antigen-loss mediated escape. Notably, wildtype DNA sequence of cCAR without codon optimization brings no potential risk such as immunogenic complications. scRNA-seq and CFU assay convincingly demonstrated that the CD123-CD33 compound CAR-T had limited toxicity on hematopoiesis. Thus, the CD123-CD33 compound CAR-T with novel scFvs provides new hope for R/R AML patients. S122: ROCK INHIBITORS TARGET SRSF2 LEUKEMIA BY DISRUPTING CELL MITOSIS AND NUCLEAR MORPHOLOGY M. Su^1 2,*, T. Fleisher^3, I. Grosheva^4, M. Horev^4, M. Olszewska^5, B. Haim^6, A. Plotnikov^6, S. Carvalho^6, Y. Moskovitz^3, M Minden^7, N. Chapal-Ilani^3, E. Papapetrou^8, N. Dezorella^9, T. Cheng^10, N. Kaushansky^3, B Geiger^4, L. Shlush^11 12 13 ^1Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China; ^2Molecular and Cellular Biology, Weizmann Institute of Science, Rehovot; ^3Molecular and Cellular Biology, Weizmann Institute of Science; ^4Immunology, Weizmann Institute of Science, Rehovot, Israel; ^5Icahn School of Medicine at Mount Sinai, New York, United States of America; ^6Wohl Institute for Drug Discovery, Weizmann Institute of Science, Rehovot, Israel; ^7University Health Network (UHN), Toronto, Canada; ^8Icahn School of Medicine at Mount Sinai, New York, United States of America; ^9Electron Microscopy Unit, Weizmann Institute of Science, Rehovot, Israel; ^10Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Tianjin, China, Tianjin, China; ^11Molecular and Cellular Biology, Weizmann Institute of Science, Rehovot; ^12Division of Hematology, Rambam Healthcare Campus, Haifa; ^13Molecular hematology clinic, Maccabi Healthcare, Tel Aviv, Israel Background: Spliceosome machinery mutations are common early mutations in myeloid malignancies, however effective targeted therapies against them are still lacking in clinical settings. Aims: Exploring safe and efficient methods to target hematopoietic cells carrying SRSF2 mutations might be a powerful tool not just for leukemia prevention and treatment, but also to understand the mechanisms behind SRSF2 mutations. Methods: We generated five SRSF2 Mut (P95H) isogenic cell lines with CRISPR/CAS9 system and performed high throughput drug screen with 3988 compounds in a single dose. After we narrowed down our targets, dose response assay was performed on four different ROCK inhibitors (ROCKi). The leading ROCKi compound was validated in vivo with SRSF2 Mut AML xenograft models. Next, we aimed at understanding why ROCKi target SRSF2 Mut cells and used proteomics, gene expression and imaging of the cytoskeletal system to study the effect of ROCKi on Mut and WT cells. Results: In the current study, we used an in vitro high-throughput drug screen among four different isogenic cell lines and identified ROCK inhibitors (ROCKi) as selective inhibitors of SRSF2 Mut in MOLM14 and AML2 cells. To study the efficacy of RKI-1447 on human samples, we conducted six AML patient-derived xenograft (PDX) experiments, with SRSF2 Mut samples. In four out of the six primary AML samples, RKI-1447 significantly reduced engraftment compared to the untreated group and inhibited the engraftment of both leukemic blasts and pre leukemic-HSPCs (preL-HSPCS). RKI-1447 was not toxic to mice nor human cells. ROCKi induced mitotic catastrophe (G2M arrest and multipolar spindles) through their apparent effects on microtubules and nuclear organization, and leading to cell death. Confocal imaging and transmission electron microscopy (TEM) data revealed that SRSF2 mutations induce deep nuclear indentation and segmentation, which is reminiscent of the nuclear shape of Pelger–Huët anomaly (PHA). The nuclear volume and area were significantly higher in SRSF2 Mut compared with WT MOLM14 cells regardless of ROCKi. To investigate why SRSF2 Mut cause the structural changes, we looked into the cytoplasmic intrusions that segment the nuclei, and the structures that connect the lobes of the nuclei by TEM. This examination revealed enrichment of microtubule bundles inside the nuclear indentations. More importantly, these microtubules were located at the tip of the cytoplasmic intrusions suggesting that they take an active role in the nuclear segmentation process. After exposure to RKI-1447, the cytoskeletal and nuclear morphological abnormalities of SRSF2 Mut are augmented to a level which are incompatible with cell survival. Image: graphic file with name hs9-6-1-g015.jpg [42]Open in a new tab Summary/Conclusion: We believe it is the first report describing in high resolution the changes in the nucleus after SRSF2 mutations are introduced. The mechanisms we identified are most probably relevant to other SMMs and to the dysplastic phenotype observed in MDS and AML. Accordingly, our findings have wide implications as ROCKi might be even more useful than describe here. Altogether, we provide data from many directions that ROCKi should be tested in clinical trials. S123: DECODING TRANSCRIPTOMIC AND EPIGENETIC CONSEQUENCES OF STRUCTURAL VARIANTS IN CK-AML AT SINGLE-CELL RESOLUTION A.-M. Leppä^1,*, K. Grimes^2, H. Jeong^2, T. Boch^1, D Karpova^1, A. Jauch^3, F. Grünschläger^1, A. Dolnik^4, L. Bullinger^4, A. Krämer^5, A. D. Sanders^2, J. O. Korbel^2, A. Trumpp^1 ^1Division of Stem Cells and Cancer, German Cancer Research Center; ^2Genome Biology Unit, European Molecular Biology Laboratory; ^3Institute of Human Genetics, University of Heidelberg, Heidelberg; ^4Hämatologie, Onkologie und Tumorimmunologie, Charité Universitätsmedizin Berlin, Berlin; ^5Molecular Hematology/Oncology, German Cancer Research Center, Heidelberg, Germany Background: Acute myeloid leukemia (AML) with complex karyotype (CK) is a heterogenous AML subgroup with dismal response to standard treatment. CK-AML has remained poorly characterized at the molecular level, with mainstream techniques struggling to unravel the complexity of the structural variants (SVs) and link them to phenotypic characteristics. To improve our understanding of molecular dynamics in CK-AML, we established an integrated single cell multi-omics approach that combines the analysis of SVs and nucleosome occupancy (NO) profiling (scNOVA) with concurrent immunophenotypic and transcriptomic profiling (CITE-seq). Aims: To characterize the genetic and non-genetic intra-patient heterogeneity at single cell resolution in CK-AML. Methods: For scNOVA analysis, blasts from four CK-AML patients were cultured in the presence of BrdU for the duration of one cell division, followed by single-cell template strand-sequencing (Strand-seq). For CITE-seq analysis, cells were stained with 38 antibody-derived tags for combined analysis of single cell transcriptome and immunophenotype. For functional studies, xenografts of the same CK-AML patients were generated by injection of AML cells into the femoral BM cavity of sublethally irradiated NOD.Prkdc^scid.Il2rg^null (NSG) mice. Results: To characterize cellular hierarchies together with epigenetic, transcriptomic, and surface protein features of CK-AML, we applied an integrated single-cell analytical framework, termed scNOVA-CITE, on over 15,000 leukemic cells from four diagnostic CK-AML patient samples. We detected complex clonal evolution patterns driven by simple and complex SVs in all patients. Three out of four patients harbored SVs leading to loss of one TP53 wild-type allele, often accompanied by loss-of-function mutations in the remaining allele. One patient showed parallel evolution of TP53-deleted and TP53 wild-type subclones in the same sample, with the latter clone harboring a lower abundance of SVs compared to the TP53-deleted cells. Moreover, based on NO we inferred lowest activity of TP53 pathway in the cells with TP53 aberrations, emphasizing the role of TP53 in intra-patient cytogenetic evolution as well as inter-patient heterogeneity. Thereby, we unraveled novel single-cell SV landscapes in CK-AML and explored their epigenetic impact. To assess in depth the functional outcomes of the genetic complexity, we also characterized the transcriptome and cell surface proteome using scNOVA-CITE. Two patients showed high genetic intra-patient heterogeneity and harbored multiple subclones at diagnosis with unique transcriptomic features. When transplanted into NSG mice, we detected predominantly monoclonal engraftment and strongly reduced genomic instability in the respective patient-derived xenografts. The engrafting clones could be traced back to a small subclone in the original patient samples. Molecular analysis of the subclones in the diagnostic samples by scNOVA-CITE revealed converging characteristics leading to stem-like phenotypes. Taken together, these data indicate that cytogenetic evolution resulting in stemness phenotype in CK-AML is advantageous for leukemic stem cell expansion in mice, unfolding novel ways to study regulation of stemness and to identify potential therapeutic targets. Summary/Conclusion: By developing scNOVA-CITE, we established an integrative single-cell multi-omics framework allowing to characterize genetic, epigenetic, transcriptomic and cell surface proteomic properties of CK-AML patient cells. We provide unique insights into the genetic complexity of CK-AML, and evidence of the resulting functional outcomes. S124: BONE MORPHOGENETIC PROTEIN 2 TRIGGERS OFF AN IMMUNOSUPPRESSIVE EFFECT OF ΓΔ T CELLS IN ACUTE MYELOID LEUKEMIA S. Liang^1 2,*, T. Dong^1, K. Yue^1, H. Gao^1, N. Wu^1, R. Liu^1, J. Liu^1, X.-J. Huang^1 2 ^1Peking University People’s Hospital, Peking University Institute of Hematology, National Clinical Research Center for Hematologic Disease, Beijing Key laboratory of Hematopoietic Stem Cell Transplantation; ^2Peking-Tsinghua Center for Life Sciences, Academy for Advanced Interdisciplinary Studies, Peking university, Beijing, China Background: Description of immune landscapes that indulge or restrain survival and proliferation of malignant cells is critical to the improvement of therapeutic strategies. Acute myeloid leukemia (AML) remains a severe life-threatening malignancy and often confronts treatment dilemma in clinic. Although γδ T cells exhibit independent and potent cytotoxicity against leukemic cells in vitro and in the mouse models, efficacy of γδ T cell-based immunotherapy on AML patients has seemed unsatisfying so far. How the anti-AML capacity of γδ T cells is suppressed in vivo is unknown. Aims: In this study, we aim to dissect the abnormal changes and functions of intrinsic γδ T cells in the context of AML and dig out the correlated factors in the AML tumor environment. Methods: 1. Immunophenotyping analyses of γδ T cells in bone marrows from newly diagnosed AML patients (n = 62) were detected using flow cytometry, compared with healthy donors (n=51). The levels of TGF-β family members in the supernatants of bone marrows were detected by ELISA kits. 2. Peripheral blood mononuclear cells (PBMCs) isolated from healthy donors were stimulated with AML blasts and morphogenetic protein 2 (BMP2) to validate clinical phenotypes. Functional experiments were performed to depict the role of a BMP2-induced subset of γδ T cells in the anti-AML activity of effector γδ T cells. 3. To confirm the findings described above, we established human CD34^+ cells-implanted mice models followed by injection with luciferase and GFP co-expressing AML cells. Results: The proportions of total γδ T cells and Vδ1^+ fraction were not significantly different between AML patients and donors. In contrast, the percentage of Vδ2^+ T cells was markedly decreased in AML patients. Meanwhile, a key activating receptor of γδ T cells, NKG2D, was dramatically decreased and regulatory receptors CD25 and PD-1 were increased on Vδ2^+ T cells in AML patients. The expansion capacity of Vδ2^+ T cells was profoundly impaired in AML patients. The functional impairments of γδ T cells were accompanied with an abnormally increased levels of BMP2, rather than TGF-β, BMP4 and ActivinA, in AML patients. Furthermore, we demonstrated that BMP2 directly induced an aberrant γδ T cells subset expressing Vδ2^+CD25^+CD127^low (abbreviated as BMP2-Vδ2) in PBMCs from healthy donors. Consistently, BMP2-Vδ2 cells were significantly increased in the bone marrows of AML patients compared with donors. The frequencies of BMP2-Vδ2 cells correlated to soluble BMP2 levels and the disease status. Distinct from effector Vδ2 cells, BMP2-Vδ2 not only lost the ability of killing AML cells, but also was able to reduce the anti-AML activity of effector Vδ2 cells after co-cultures. In AML-loaded humanized mice, treatment with BMP2 induced the occurrence of BMP2-Vδ2 cells and reduced the survival rate of mice. Injection of BMP2-Vδ2 cells significantly attenuated the activity of effector Vδ2 cells on the elimination of AML cells in mice. Together, these results demonstrated the immunoregulatory effect of BMP2 on the phenotype and function of γδ T cells in AML microenvironment. Summary/Conclusion: We first reported that dysregulated BMP2 in AML triggered an immunosuppressive subset of γδ T cells, leading to the impairment of anti-AML function of effector γδ T cells. Our findings provide a novel insight into the mechanisms of immunosuppression in the context of leukemia and suggest potential targets for the treatment of AML and other hematopoietic malignancies. S125: 10-DAY DECITABINE VS. CONVENTIONAL CHEMOTHERAPY (“3 + 7”) FOLLOWED BY ALLOGRAFTING (HSCT) IN AML PATIENTS ≥60 YEARS: A RANDOMIZED PHASE III STUDY OF THE EORTC LEUKEMIA GROUP, GIMEMA, CELG, AND GMDS-SG M. Lübbert^1,*, P. Wijermans^2, M. Kicinski^3, S. Chantepie^4, W. van der Velden^5, R. Noppeney^6, L. Griskevicius^7, A. Neubauer^8, M. Crysandt^9, R. Vrhovac^10, M. Luppi^11, S. Fuhrmann^12, E. Audisio^13, A. Candoni^14, O. Legrand^15, R. Foà^16, G. Gaidano^17, D. van Lammeren-Venema^2, E. F. Posthuma^18, M. Hoogendoorn^19, A. Giraut^3, M. Stevens-Kroef^20, J. H. Jansen^21, E. Ammatuna^22, J.-P. Vilque^4, R. Wäsch^1, H. Becker^1, N. Blijlevens^5, U. Dührsen^6, F. Baron^23, S. Suciu^3, S. Amadori^24, A. Venditti^24, G. Huls^22 ^1Department of Hematology, Oncology and Stem Cell Transplantation, Faculty of Medicine, University Medical Center Freiburg, University of Freiburg, Freiburg, Germany; ^2Department of Hematology, Haga Teaching Hospital, The Hague, Netherlands; ^3EORTC Headquarters, Brussels, Belgium; ^4Institut d’Hématologie de Basse Normandie, Centre Hospitalo-Universitaire de Caen, Caen, France; ^5Department of Hematology, Radboud University Medical Centre, Nijmegen, Netherlands; ^6Klinik für Hämatologie, Universitätsklinikum Essen, Essen, Germany; ^7Department of Hematology, Oncology and Transfusion Medicine Center, Vilnius University Hospital Santaros Klinikos, Vilnius University, Vilnius, Lithuania; ^8Department of Internal Medicine, Hematology, Oncology and Immunology, Philipps University Marburg and University Hospital Gießen and Marburg, Campus Marburg, Marburg; ^9Department of Hematology, Oncology, Hemostaseology and Stem Cell Transplantation, Medical Faculty, Center for Integrated Oncology Aachen Bonn Cologne Duesseldorf (CIO ABCD), Aachen, Germany; ^10Department of Haematology, University Hospital Centre Zagreb, Zagreb, Croatia; ^11Dipartimento di Scienze Mediche e Chirurgiche Materno-Infantili e dell’Adulto, University of Modena and Reggio Emilia, Azienda Ospedaliera Universitaria, Modena, Italy; ^12Department of Hematology and Oncology, HELIOS Hospital Berlin-Buch, Berlin, Germany; ^13SC Ematologia Città della Salute e della Scienza Torino, Torino; ^14Clinica Ematologica Azienda Sanitaria Universitaria Integrata di Udine, Udine, Italy; ^15Service d’Hématologie Clinique et de Thérapie cellulaire, Hôpital Saint Antoine, APHP, Paris, France; ^16Ematologia, Dipartimento di Medicina Traslazionale e di Precisione, “Sapienza” Università di Roma, Rome; ^17Division of Hematology, Department of Translational Medicine, Università del Piemonte Orientale and Azienda Ospedaliero-Universitaria Maggiore della Carità, Novara, Italy; ^18Department of Internal Medicine, Reinier de Graaf Hospital, Delft; ^19Department of Hematology, Medical Center Leeuwarden, Leeuwarden; ^20Radboud University Medical Center, Nijmegen, Netherlands; ^21Dept laboratory Medicine, Lab Hematology, Radboud University Medical Center, Nijmegen; ^22University Medical Center Groningen, Groningen, Netherlands; ^23University of Liège, Liège, Belgium; ^24Department of Biomedicine and Prevention, University of Rome Tor Vergata, Rome, Italy Background: Older, fit AML patients (pts) treated with induction chemotherapy (IC) have poor long-term survival unless HSCT is performed. DNA-hypomethylating agents have become the backbone of AML therapy in pts unfit for IC. Promising outcomes have been reported for the 10-day decitabine (DEC) schedule, suggesting it may be a better treatment prior to HSCT as compared to IC. Aims: To compare efficacy and safety of 10-day DEC followed by allografting to IC followed by allografting in older fit AML pts. Methods: This was an international open-label randomized phase III trial ([43]NCT02172872). Key inclusion criteria were newly diagnosed AML, age ≥60 years, eligible for IC, WHO performance status 0-2. DEC was administered 10 days consecutively in cycle 1 (20 mg/m^2), 10 or 5 days in subsequent cycles (depending on bone marrow blast clearance at day 28). IC treatment was daunorubicin 60 mg/m^2 x 3 days, cytarabine 200 mg/m^2 x 7 days, followed by 1-3 additional chemotherapy cycles. Pts who had an HLA-matched donor and at least stable disease were encouraged to undergo HSCT after ≥1 treatment cycle. Pts from the DEC arm not receiving HSCT could continue DEC treatment. The primary endpoint was overall survival (OS). Pts were randomized 1:1, stratified by de novo AML vs. secondary AML, age (60-64 vs 65-70 vs ≥70 yrs), and institution. The statistical design aimed to detect a hazard ratio (HR) for OS of 0.75 (HR<1 indicates longer survival for DEC), requiring 441 deaths (one-sided alpha 0.025, 85% power). Due to the slow accumulation of deaths, the final analysis was performed with a clinical cut-off (CCO) date June 30, 2021, following the Data Monitoring Committee recommendation. Results: Between 12/2014 and 8/2019, 606 pts were randomized, 303 in each arm. Median follow-up was 4.0 yrs. Median age was 68 yrs (range 60-81), 34% of pts were ≥70 yrs old and 57% were male, 21% and 32% had good and adverse ELN 2017 risk profile, respectively. A median of 3 DEC cycles (Q1-3: 2-5) and 2 IC cycles (Q1-3: 1-2) were administered. The CR/CRi rate was 48% with DEC and 61% with IC. HSCT as part of the protocol was performed in 122 pts (40%, 30 of them not in CR/CRi) from the DEC and 118 (39%, 11 of them not in CR/CRi) from the IC arm, and in 52% in both arms at any time. By the CCO, 423 deaths occurred. The OS was not significantly different between DEC and IC groups (HR=1.04, 95% confidence interval [CI]: 0.86-1.26; 2-sided p=0.68). The median OS was 15 months (95% CI: 13-18) in the DEC and 18 months (95% CI: 14-22) in the IC group. The OS rates (%) after 1, 2, 3 and 4 years for the DEC and IC groups were 58 vs 59, 38 vs 40, 30 vs 33, and 26 vs 30, respectively. In age subgroups, the estimated HR for OS for DEC vs IC was 1.34 (99% CI: 0.79-2.28) for pts aged 60-64, 1.14 (99% CI: 0.77-1.69) for pts aged 65-69, and 0.84 (99% CI: 0.55-1.26) for pts aged ≥70 yrs (p-value for trend: 0.058). Notable differences in the incidence of grade 3-5 adverse events (%) reported (before HSCT) were: febrile neutropenia (37% for DEC vs 57% for IC), decrease in platelets (24% for DEC vs 32 % for IC), oral mucositis (2% for DEC vs 10% for IC), diarrhea (1% for DEC vs 8% for IC), decrease in neutrophils (19% for DEC vs 13% for IC). The 30-day mortality rate was 3.6% for DEC and 6.4% for IC. The incidence of grade 5 treatment-related adverse events after HSCT was comparable in both treatment arms (25% for DEC and 22% for IC). Summary/Conclusion: Treatment with DEC resulted in a similar OS and HSCT rate but a better safety profile compared to IC in older AML pts ≥60 yrs, eligible for IC. S126: GEMTUZUMAB-BASED INDUCTION CHEMOTHERAPY COMBINED WITH MIDOSTAURIN FOR FLT3 MUTATED AML. RESULTS FROM THE NCRI AML19 “MIDOTARG” PILOT TRIAL N. Russell^1,*, C. Wilhelm-Benartzi^2, J. Othman^3, R. Dillon^3, N. Potter^3, J. Jovanovic^3, A. Gilkes^4, L. M. Batten^2, J. Canham^2, E. L. Hinson^2, P. Kottaridis^5, J. Cavenagh^6, C. Arnold^7, M. Dennis^8, S. Knapper^4 ^1Haematology, Guy’s and St Thomas’ NHS Foundation Trust, London; ^2Centre for Trials Research, Cardiff University, Cardiff; ^3Department of Medical and Molecular Genetics, Kings College London, London; ^4School of Medicine, Cardiff University, Cardiff; ^5Haematology, University College London Hospitals NHS Foundation Trust; ^6Department of Haematology, St Bartholomew’s Hospital, London; ^7Haematology, Belfast City Hospital, Belfast; ^8Haematology, The Christie NHS Foundation Trust, Manchester, United Kingdom Background: Following the RATIFY study Midostaurin in combination with “7 + 3” like chemotherapy has become the standard of care for patients with newly diagnosed FLT3 mutated AML. The ALFA 0701 trial also suggested a benefit for Gemtuzumab Ozogamicin (GO) in FLT3 mutated AML however the combination of GO-based induction with Midostaurin has not been formally assessed. Aims: To assess the safety of midostaurin in combination with Gemtuzumab and intensive induction therapy and to assess impact of the combination on MRD kinetics Methods: The NCRI AML19 trial randomised patients to receive DA 3 + 10 (Daunorubicin 60mg/m2 on days 1,3,5 plus AraC 100mg/m^2 bd on days 1-10) plus a single dose of GO 3mg/m^2 on day 1 (DAGO1) or 2 doses (3mg/m^2, maximum 5mg) on days 1 and 4 (DAGO2) plus 50mg bd of midostaurin (m) for 14 days following completion of chemotherapy. Eligibility included age 18-60 years with a FLT3-ITD or TKD mutation. Enhanced pharmacovigilance (PV) was performed for four weeks following the first induction course. Midostaurin was also given following second induction (DA 3 + 8 without GO) and 2 courses of HDAC consolidation and as maintenance for 12 cycles in non-transplanted patients. From November 2020 to November 2021, 77 patients were enrolled into the Midotarg pilot receiving DAGO1m (n=39) or DAGO2m (n=38). 59 had a FLT3 ITD and 22 a FLT3-TKD (and 4 had both). 59 patients have completed course 1 and are evaluable. RT-qPCR MRD monitoring for patients with an NPM1 mutation (n=48) was performed following each cycle of chemotherapy. A descriptive comparison is presented here of toxicity and MRD kinetics with FLT3 mutated patients receiving DAGO1 or DAGO2 without Midostaurin in the same trial. Results: Treatment compliance in course 1 was 88% for DAGO1m and 100% for DAGO2m. One patient did not receive any Midostaurin because of colitis during induction. Dose interruption for QTc prolongation occurred in 1 patient. A total of 17 SAEs (CTC grade 3 or greater) were reported (GO1, n=11, GO2 n= 6). Day 60 mortality was 0%. No cases of VOD were reported. Blood count recovery was not delayed. Time to neutrophil recovery to 1 x 10^9/L was 31 and 32 days with DAGO1m and DAGO2m compared to 31 and 32 days in DAGO1 and DAGO2 .. Time to platelet recovery to 100 x 10^9/L was 28 and 29 days with DAGO1m and DAGO2m respectively compared to 30 days in DAGO1 and DAGO2. Complete remission (CR plus CRi) was achieved in 51/59 (88%) evaluable patients who have completed induction with DAGOm. In 44 evaluable patients in remission with NPM1 mutation, 34 (74%) were MRD negative in the peripheral blood after course 2. This compares with 63% in 54 evaluable patients with DAGO only. Bone marrow NPM1 transcript levels after courses 1 to 4 were compared with FLT3 mutated patients receiving DAGO1 and DAGO2 without Midostaurin (Figure 1) with a higher proportion of patients becoming MRD negative from course 2 onwards and 81% being MRD negative after course 4. Image: graphic file with name hs9-6-1-g016.jpg [44]Open in a new tab Summary/Conclusion: The addition of midostaurin to DAGO using a single or fractionated dose of GO was well tolerated with no evidence of increased toxicity, high response rate and with encouraging clearance of NPM1 mutant transcripts. A randomised study of DAm versus DAGOm in FLT3 mutated AML is planned S127: QUIZARTINIB WITH DECITABINE AND VENETOCLAX (TRIPLET) IS ACTIVE IN PATIENTS WITH FLT3-ITD MUTATED ACUTE MYELOID LEUKEMIA - A PHASE I/II STUDY M. Yilmaz^1,*, M. Muftuoglu^1, H. Kantarjian^1, C. DiNardo^1, T. Kadia^1, G. Borthakur^1, N. Pemmaraju^1, N. Short^1, Y. Alvarado^1, A. Maiti^1, L. Masarova^1, G. Montalban Bravo^1, S. Loghavi^2, K. Patel^2, S. Kornblau^1, E. Jabbour^1, G. Garcia-Manero^1, M. Andreeff^1, N. Daver^1 ^1Leukemia; ^2Hematopathology, MD Anderson Cancer Center, Houston, United States of America Background: Quizartinib (QUIZ), a potent 2nd generation FLT3 inhibitor (FLT3i) demonstrated synergy with venetoclax (VEN) in AML cell lines and PDX models (Mali Haematologica 2020). Aims: To evaluate the safety and efficacy of Decitabine (DAC) + VEN + QUIZ triplet in patients (pts) with newly diagnosed (ineligible for intensive induction chemotherapy) or relapsed/refractory (R/R; up to 5 prior chemotherapies) FLT3 ITD mutated AML. Methods: All pts received 10 days of DAC (20 mg/m2) in Cycle 1. Pts underwent day 14 bone marrow (BM) biopsy, and VEN (400 mg/day starting from day1) was put on hold in pts with BM blasts ≤ 5% or aplasia. Those with day14 BM blast >5% continued VEN for 21 days during cycle 1. In subsequent cycles, DAC was reduced to 5 days. QUIZ (30 or 40 mg/day) was administered daily continuously. Results: Overall 28 pts were enrolled and evaluable at the time of this report. Of 23 pts with R/R AML (median 3 prior Rx, 78% with ≥1 prior FLT3i including prior gilteritinib in 70%, and 39% had a prior alloSCT), 78% achieved CRc (3 CR, 15 CRi) with 6/16 and 5/18 responders achieving FLT3-PCR and multicolor flow cytometry negativity, respectively. The CRc rates were 75% and 72% in pts who received prior gilteritinib and prior HMA+VEN, respectively. 60-day mortality rate was 5%. Of 5 pts with newly diagnosed AML (median age 69), all achieved CRc (2 CR, 3 CRi) with 4/5 and 2/4 responders achieving FLT3-PCR and MFC negativity, respectively. 60-day mortality in frontline was 0. RAS/MAPK mutations appear to drive both primary and secondary resistance. Pts with underlying RAS/MAPK mutations had the lowest response rates, at 40% compared to 94% in those without. None of the six pts who achieved a durable remission (> 6months) had RAS/MAPK mutations at baseline (Figure A). 4 of 16 (25%) of pts who relapsed (< 6 months of remission) or were refractory to this triplet regimen had baseline RAS/MAPK mutations. We had pre- and post-treatment NGS data from 8 pts who had a response but then relapsed (Figure B). Emergent RAS/MAPK mutations were noted in 37% of relapses (3/8), while emergent FLT3 F691L gatekeeper mutations was noted in 25% of relapses (2/8). Interestingly, there were no emergent FLT3 TKD mutations. No pts developed a dose limiting toxicity (DLT) with 30 mg/day quizartinib, however with the 40mg/day quizartinib 2 pts developed hematologic DLT (grade 4 neutropenia with a <5% cellular bone marrow lasting ≥42 days). Hence, quizartinib 30 mg/day dose was determined to be the recommended phase 2 dose for the triplet. Most common Grade 3/4 non-hematologic toxicities included lung infections (42%) and neutropenic fever (30%). No QTcF prolongations >480 msec were noted. With a median follow-up (f/u) of 13 months, the median OS was 7.6 months in R/R cohort. Median OS in prior Gilt exposed pts was 6.3 months and ≥1 prior FLT3i exposed pts was 6.3 months. 8/18 R/R pts (including 5/8 prior Gilt exposed pts) underwent ASCT with a median OS of 19 vs 8 months in pts who underwent ASCT versus not (p=0.26). Of the 5 frontline responding pts median OS was 14.5, 2 were alive in CR, 1 died in CR1 post-ASCT, 2 died due to progressive disease at the last f/u. Image: graphic file with name hs9-6-1-g017.jpg [45]Open in a new tab Summary/Conclusion: DAC + VEN + QUIZ is active in R/R FLT3-ITD mutated AML pts, with CRc rates of 78% and the median OS of 7.6 months. The high response rate was maintained in prior Gilteritinib exposed pts. Interestingly, RAS/MAPK mutations but not emergent TKD mutations were associated with primary and secondary resistance to the triplet. Accrual continues and updated clinical, NGS and mass cytometry (CyTOF) data will be presented. S128: A RANDOMISED COMPARISON OF CPX-351 AND FLAG-IDA IN HIGH RISK ACUTE MYELOID LEUKAEMIA. RESULTS FROM THE NCRI AML19 TRIAL N. Russell^1,*, C. Wilhelm-Benartzi^2, S. Knapper^3, L. Batten^2, J Canham^2, E. Hinson^2, U. Malthe Overgaard^4, J. Othman^5, R. Dillon^5, P. Mehta^6, P. Kottaridis^7, J. Cavenagh^8, C. Hemmaway^9, C. Arnold^10, M. Dennis^11 ^1Guy’s and St Thomas’ NHS Foundation Trust, Department of Haematology, London; ^2Centre for Trials Research; ^3School of Medicine, Cardiff University, Cardiff, United Kingdom; ^4Copenhagen University Hospital, Copenhagen, Denmark; ^5Department of Medical and Molecular Genetics, Kings College London, London; ^6University Hospitals of Bristol and Weston NHS Trust, Bristol; ^7University College London Hospitals NHS Foundation Trust; ^8Department of Haematology, St Bartholomew’s Hospital, London, United Kingdom; ^9Aukland Hospital, Aukland, New Zealand; ^10Belfast City Hospital, Belfast; ^11The Christie NHS Foundation Trust. On behalf of the NCRI AML Working Group, Manchester, United Kingdom Background: Liposomal daunorubicin and cytarabine (CPX-351) has shown a survival advantage for older patients (>60years) with secondary AML compared to 3 + 7 chemotherapy however in younger patients there is a lack of randomised evidence for benefit. We have previously reported improved survival with FLAG-Ida treatment as treatment intensification for younger patients identified with high risk (HR) AML following induction therapy and for patients with secondary AML (Burnett AK et al. Leukemia. 2018 Dec;32(12):2693-2697) and considered this regimen an appropriate comparator for trials in younger patients. Aims: We compared CPX-351 with FLAG-Ida in a randomised fashion in patients who were either HR at trial entry based on cytogenetics or identified as HR following induction or at relapse. Methods: The AML19 trial (ISRCTN78449203) randomised CPX-351 vs FLAG-Ida in 635 patients mainly <60 years with HR AML or MDS (>10% blasts) (median age 53.6 yrs). Three groups of HR patients were randomised 2:1 in favour of CPX with the aim of proceeding to allogeneic transplant. Group 1 (n=195) had known adverse risk cytogenetics (Grimwade et al, Blood 2010,116, 354) and were randomised at diagnosis between 4 courses of CPX-351 and 2 courses of FLAG-Ida followed by MACE/MidAC consolidation. Group 2 (n=263) were HR by validated risk score, had FLT3-ITD without an NPM1 mutation, had refractory disease and were randomised after induction course 1. Group 3 (n=177) were randomised after course 2 if they had persisting MRD at the time of relapse. Here we present results for Group 1. Group 1 was not powered to claim statistical significance; therefore, these results are intended to be exploratory and hypothesis generating. Results: Group 1 included 49.2% with de novo AML 20.3% patients with secondary AML and 30.5% with HR MDS. The Overall response rate(CR/CRi) was 64.8% for CPX-351 and 74.4% for FLAG-Ida (univariate OR:0.57, 95%CI 0.30-1.10, p=0.09). Overall survival (OS) at 3 years was 32% and 24%, median OS was 13.3 months vs 10.2 months (univariate HR:0.83, 95%CI 0.58-1.18, p=0.3) for CPX -351 and FLAG-Ida respectively (Figure 1a). Event free survival (EFS) was not significantly different (HR:0.91 95%CI: 0.50-1.64, p=0.76). Relapse free survival (RFS) at 3 years was 43% and 28%, median RFS was 22.1 months vs 14 months (univariate HR:0.66, 95%CI 0.41-1.06, p=0.09) for CPX -351 and FLAG-Ida respectively (Figure 1b). RFS was significant when adjusting for NPM1 mutation status or FLT3 mutation status using multivariable cox regression model with RFS being better with CPX-351 compared to FLAG-Ida (HR:0.58, 95% CI0.36-0.95, p=0.03). Median duration of remission favoured CPX and was 319.5 days vs 167 days (p=0.046) for CPX vs FLAG-Ida respectively. Haematological toxicity was greater in course 1 with CPX-351 with platelet recovery to 100x10^9/L at 34.5 days versus 29 days (p<0.001) and neutrophil recovery to 1.0x10^9/L at 32 days vs 30 days.(p=0.12). Day 30 and day 60 mortality were not significantly different between arms with 4.8% vs 7.3% (p=0.46) and 12.4% vs 11.0% (p=0.77) for CPX-351 and FLAG-Ida respectively.Compliance was better with CPX-351 with 90.7% vs 83.0% receiving the scheduled course 1 dose. More patients receiving CPX-351 were transplanted (50.5% vs 41.5%) with the median number of courses given prior to transplant 2 in both arms. Image: graphic file with name hs9-6-1-g018.jpg [46]Open in a new tab Summary/Conclusion: In patients with adverse cytogenetics CPX-351 did not improve response, OS or EFS compared to FLAG-Ida but was associated with better duration of remission and RFS. Further follow-up is needed to determine the clinical significance of those differences. S129: TAKEAIM LEUKEMIA- A PHASE 1/2A STUDY OF THE IRAK4 INHIBITOR EMAVUSERTIB (CA-4948) AS MONOTHERAPY OR IN COMBINATION WITH AZACITIDINE OR VENETOCLAX IN RELAPSED/REFRACTORY AML OR MDS G. Garcia-Manero^1,*, E. S. Winer^2, D. J. DeAngelo^2, S. Tarantolo^3, D. A. Sallman^4, J. Dugan^5, S. Groepper^6, A. Giagounidis^6, K. Götze^7, K. H. Metzeler^8, C.-C. Li^9, L. Zhou^9, E. Martinez^9, M. Lane^9, R. von Roemeling^9, M. Bohme^9, A. S. Kubasch^10, A. Verma^11, U. Platzbecker^10 ^1Luekemia, MD Anderson Cancer Center, Houston; ^2Dana-Faber Cancer Institute, Boston; ^3Nebraska Cancer Specialist, Omaha; ^4Moffitt Cancer Center, Tampa; ^5Novant Health Cancer Institute, Forsyth Medical Center, Winston-Salem, United States of America; ^6Marien Hospital/ Univ. of Dusseldorf Germany, Dusseldorf; ^7Faculty of Medicine Technical University of Munic, Munich; ^8Hematology, Cellular Therapy and Hemostaseology, University Hospital Leipzig, Leipzig, Germany; ^9Curis, Lexington, United States of America; ^10University Hospital Leipzig, Leipzig, Germany; ^11Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, United States of America Background: Emavusertib (CA-4948) is a novel oral inhibitor of interleukin-1 receptor-associated kinase 4 (IRAK4) and FLT3. IRAK4 is critical in triggering inflammation, oncogenesis, and survival of cancer cells. Genetic mutations in the splicing factors SF3B1 and U2AF1 drive overexpression of a highly active long isoform of IRAK4 and have been associated with disease progression and poor prognosis of high-risk myelodysplastic syndrome (HR-MDS) and acute myeloid leukemia (AML). Aims: Assessment of safety, clinical activity, and Recommended Phase 2 Dose (RP2D) of emavusertib as monotherapy or in combination with azacitidine (AZA) or venetoclax (VEN). Methods: This is an open-label, phase 1/2a dose escalation and cohort expansion trial ([47]NCT04278768). In phase 1 Dose Escalation, patients with R/R AML or HR-MDS are treated with emavusertib monotherapy. Phase 1b includes 2 arms of combination therapy: emavusertib + AZA and emavusertib + VEN. The primary objectives of this study are to assess the safety, clinical activity, and identify the RP2D of emavusertib as monotherapy or in combination with AZA or VEN in R/R AML or HR-MDS. The Phase 2a Dose Expansion includes patients for emavusertib monotherapy: R/R AML with FLT3 mutation, or AML and HR-MDS R/R to HMA with U2AF1 or SF3B1 mutations. Results: As of December 16^th, 2021, 49 patients have been treated in the phase 1 portion, of whom 43 started by September 30^th, allowing 2 on-study disease assessments. The median number of prior therapies was 2 (range 1-5). Four monotherapy dose levels of emavusertib were tested (200 to 500 mg orally BID). No dose-limiting toxicities were observed at 200 mg and 300 mg BID. No Grade 4 or 5 treatment-related AEs (TRAEs) were reported, and all the TRAEs were manageable. Reversible, manageable Grade 3 rhabdomyolysis occurred in 1/26 (4%) patients at 300 mg BID, 2/17 (12%) at 400 mg BID, and 1/3 (33%) at 500 mg BID. RP2D was determined as 300 mg BID. Of 43 patients starting before Sept 30^th, 2021, 14 had SF3B1, U2AF1 or FLT3 mutations and demonstrated more promising efficacy. In the 5 evaluable AML patients with spliceosome mutations, 40% reached CR/CRh (1 CR, 1 CRh), both with study duration >6 months. In the 7 spliceosome-mutated HR-MDS patients, 57% reached marrow CR, including 1 with RBC transfusion independence and 1 proceeding to HSCT. One of the three FLT3-mutated AML reached CR, and 2 became FLT3-negative. Among the 29 patients without SF3B1/U2AF1/FLT3 mutations, 1 reached CR and 2 PR. Phase 1b and Phase 2a are ongoing. RNA-seq on selected samples showed decrease in relative expression of IRAK4-long isoforms with response to emavusertib. Summary/Conclusion: Emavusertib is well tolerated and effective in heavily pretreated AML and HR-MDS patients, especially in those with U2AF1/SF3B1/FLT3 mutations. No dose-limiting myelosuppression was reported, suggesting emavusertib may be a candidate for combination therapy. Accrual of Phases 1b and 2a is ongoing. S130: NPM1 MUTATED AML: IMPACT OF CO-MUTATIONAL PATTERNS - RESULTS OF THE EUROPEAN HARMONY ALLIANCE A. Hernández-Sánchez^1 2,*, Á. Villaverde-Ramiro^2, J. Martínez Elicegui^2, T. González^2 3, A. Benner^4, E. Sträng^5, G. Castellani^6, C. A. Heckman^7 8, J. Versluis^9, M. Abáigar^2 3, M. Sobas^10, R. Azibeiro^1 2, L. Tur^11, P. J. Valk^9, K. H. Metzeler^12, R. Ayala^13, D. Dall’Olio^6, J. Tettero^14, J. Martínez-López^13, H. Dombret^15, M. Pratcorona^16, F. Damm^5, K. I. Mills^17, J. Mayer^18, C. Thiede^19, M. T. Voso^20, G. F. Sanz^21 22, F. Calado^23, K. Döhner^24, V. I. Gaidzik^25, M. Heuser^26, T. Haferlach^27, A. T. Turki^28 29, D. Reinhardt^28, R. Villoria Medina^11, M. van Speybroeck^30, R. Schulze-Rath^31, M. Barbus^32, J. E. Butler^33, J. M. Hernández Rivas^1 2 3, B. J. Huntly^34, G. J. Ossenkoppele^14 35 36, H. Döhner^24, L. Bullinger^5 ^1Hospital Universitario de Salamanca; ^2Instituto de Investigación Biomédica de Salamanca (IBSAL); ^3Centro de Investigación del Cáncer (CIC), Salamanca, Spain; ^4German Cancer Research Center (DKFZ), Heidelberg; ^5Charité Universitätsmedizin Berlin, Berlin, Germany; ^6University of Bologna, Bologna, Italy; ^7University of Helsinki; ^8Institute for Molecular Medicine Finland, Helsinki, Finland; ^9Erasmus University Medical Center Cancer Institute, Rotterdam, Netherlands; ^10Wroclaw Medical University, Wroclaw, Poland; ^11GMV Innovating Solutions, Valencia, Spain; ^12University of Munich, Munich, Germany; ^13Hospital Universitario; ^12de Octubre, Madrid, Spain; ^14Amsterdam University Medical Center, Amsterdam, Netherlands; ^15EA3518 Leukemia Translational Laboratory, Paris, France; ^16Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; ^17Queens University Belfast, Belfast, United Kingdom; ^18University Hospital Brnoand Masaryk University, Brno, Czechia; ^19University of Technics Dresden Medical Dept., Dresden, Germany; ^20University of Rome “Cattolica S. Cuore”, Rome, Italy; ^21Hospital Universitario y Politécnico La Fe; ^22Instituto de Salud Carlos III (CIBERONC), Valencia, Spain; ^23Novartis, Oncology Region Europe, Basel, Switzerland; ^24University Hospital of Ulm; ^25Ulm University Hospital, Ulm; ^26Hannover Medical School, Hannover; ^27MLL Munich Leukemia Laboratory, Munich; ^28Essen University Hospital; ^29West-German Cancer Center, Essen, Germany; ^30Janssen Pharmaceutica N.V., Beerse, Belgium; ^31Bayer Pharma AG, Berlin; ^32AbbVie Germany GmbH & Co. KG, Wiesbaden; ^33Bayer AG, Berlin, Germany; ^34Wellcome - MRC Cambridge Stem Cell Institute, Cambridge, United Kingdom; ^35VU University Medical Center; ^36Cancer Center Amsterdam, Amsterdam, Netherlands Background: Acute myeloid leukemia (AML) is a heterogeneous disease in terms of clinical features, outcomes and genetics. While mutations of NPM1 are usually considered as a favorable prognostic marker, the vast majority of the patients carry several co-mutations that might influence the prognosis. Therefore, a better understanding of the NPM1^mut AML mutational landscape is warranted. The large cohort of AML patients collected within the European HARMONY Alliance provides an excellent basis for this purpose. Aims: To identify clinically significant co-mutational patterns in NPM1^mut AML in order to establish a revised risk stratification model. Methods: From the HARMONY Alliance AML database, a total of 1001 NPM1^mut intensively treated patients were selected. Clinically significant co-mutations were evaluated using graphical patterns created with the Gephi tool and confirmed by detailed survival analysis using Kaplan-Meier and Cox regression models. Finally, a novel multi-state risk stratification model for NPM1^mut AML was established. Results: The study population of 1001 NPM1^mut AML patients included 57% females and median age was 53 years. Regarding ELN2017 classification, 68% of patients were classified into the favorable, 29% intermediate and 3% adverse risk groups. The most frequent co-mutations were DNMT3A (54%), followed by FLT3-ITD (38%). In total, 24% of patients presented with a high allelic mutant-to-wildtype ratio ≥0.5 (FLT3-ITD^high) while 14% had low allelic ratio <0.5 (FLT3-ITD^low). Other frequent co-mutations were NRAS (21%), TET2 (20%) and PTPN11 (15%). The triple mutation pattern of NPM1^mut + FLT3-ITD^high + DNMT3A^mut identified a subgroup with adverse prognosis (2-year OS of 25%), similar to NPM1^mut + TP53^mut. The combination of FLT3-ITD^low + DNMT3A^mut or FLT3-ITD^high + DNMT3A^wt was associated with intermediate prognosis (2-year OS of 45% and 53% respectively). Notably, mutations of NRAS, KRAS, PTPN11 or RAD21 were identified to be associated with better OS. However, in the context of NPM1^mut + DNMT3A^mut these mutations did not affect the prognosis when a FLT3-ITD was present. This information is summarized in a 3-category risk classification model (Figure 1). The revised NPM1^mut favorable group presented with a 2-year OS of 73%, while for intermediate and adverse groups the OS was 54% and 27% respectively (p<0.001). Regarding relapse free survival (RFS), the median was not reached in the favorable group, while it was 23 months for intermediate and 6 months for adverse group (p<0.001). It should be noted that 171 patients in the NPM1^mut intermediate group would be considered as favorable according to the ELN2017 criteria, as well as 162 patients in the NPM1^mut adverse group were previously classified as intermediate risk. Therefore, our model was able to reclassify 33% of NPM1^mut AML patients in comparison to ELN2017 criteria. Multivariate analysis of OS in NPM1^mut AML identified the following independent prognostic factors: NPM1^mut model (taking favorable group as reference, HR 1.6 for intermediate and HR 2.7 for adverse group, p<0.001); secondary or therapy-related AML (HR 1.8, p<0.001), WBC at diagnosis >100x10^3/μL (HR 1.5, p<0.001) and age >60 years (HR 1.4, p<0.001). Image: graphic file with name hs9-6-1-g019.jpg [48]Open in a new tab Summary/Conclusion: Analysis of large NPM1^mut AML cohorts allows the discovery of co-mutation patterns associated with prognostic outcome. In accordance, we propose a new genetic stratification model for NPM1^mut AML that identifies 3 groups with different OS and RFS. This model improves ELN2017 criteria as it is able to correctly reclassify 33% of NPM1^mut AML patients. S131: CLINICAL IMPLICATIONS OF SECONDARY-AML TYPE MUTATIONS IN PATIENTS WITH DE NOVO ACUTE MYELOID LEUKEMIA K. Sun^1,*, C.-H. Tsai^1, M.-Y. Lo^1, M.-H. Tseng^1, Y.-Y. Kuo^2, M.-C. Liu^3, C.-C. Lin^1, C.-L. Cheng^1, S.-J. Wu^1, C.-Y. Chen^1, B.-S. Ko^1 4, M. Yao^1, W.-C. Chou^1, H.-A. Hou^1, H.-F. Tien^1 1 ^1Hematology; ^2Oncology; ^3Pathology, National Taiwan University Hospital; ^4Hematology, National Taiwan University Cancer center, Taipei, Taiwan Background: More and more gene mutations have been identified in patients with de novo acute myeloid leukemia (AML). Among them, a set of gene mutations, including SRSF2, ZRSR2, SF3B1, U2AF1, ASXL1, EZH2, STAG2, and BCOR mutation, are categorized as secondary AML (sAML)-type mutations, for their distinct distribution in secondary AML, compared to primary AML (Lindsley et al, Blood 2015), but the reports regarding the prognostic impact have been scanty, especially in younger patients. Aims: In this study, we aimed to explore the clinical significance and prognostic implication of sAML-type mutations in non-M3 AML patients. Methods: We consecutively enrolled 921 de novo non-M3 AML patients; 368 were 60 years or older (older patients) and 553 were younger. Patients with an antecedent history of hematologic diseases, or therapy-related AML were excluded. sAML-type mutations were identified by targeted next-generation sequencing of 54 myeloid malignancies related gene mutations. Results: A total of 243 (26.4%) patients harbored sAML-type mutations (ST group), 40.2% in older patients and 17.2% in younger ones. Patients in the ST group were significantly older, had a lower WBC count, peripheral blast count and lactate dehydrogenase level at diagnosis. sAML-type mutations were negatively correlated with inv(16), monosomy 17, and complex karyotype, but positively associated with 2017 European LeukemiaNet (ELN)-defined unfavorable-risk genetic category. Among the patients receiving standard chemotherapy (n=686, 74.5%), the ST group had a significantly lower CR rate (62.9% vs. 81.5%, P< 0.01), especially among younger patients (70.8% vs. 86.7%, P<0.01), but only a trend in older patients (49.0% vs. 59.6, P=0.17). With a median follow-up of 4.7 years, patients in the ST group had a shorter overall survival (OS, median, 2.1 years vs. not reached, P< 0.01) and disease-free survival (DFS, median, 0.4 years vs. 0.9 years, P< 0.01) than the non-ST group. Subgroup analyses showed that sAML-type mutations conferred a significantly poorer DFS (median, 0 years vs. 0.5, P=0.03) and a trend of shorter OS (0.8 years vs. 1.1 years, P=0.08) in older patients, and a significantly worse OS (not reached vs. not reached, P=0.03) and a trend of shorter DFS (0.7 years vs. 1.0 years, P=0.16) in younger patients. The numbers of sAML-type mutations had prognostic impacts on both OS (median, 5.8 years vs. 1.5 vs. 1.3 for patients with 0, 1, and ≥2 mutations, respectively, P<0.01) and DFS (median, 0.9 years vs. 0.6 vs. 0 for those with 0, 1, and ≥2 mutations, respectively, P< 0.01). Intriguingly, these findings were valid among both the younger and older patients. Furthermore, the ELN-defined intermediate-risk patients with sAML-type mutations had similar poor OS and DFS to the ELN unfavorable-risk patients in total cohort (Figure 1), as well as in the older and younger patients. Among the patients with sAML-type mutations, allogeneic hematopoietic stem cell transplantation did improve their outcome (median OS 3.0 vs. 0.8 years, P< 0.01). Image: graphic file with name hs9-6-1-g020.jpg [49]Open in a new tab Summary/Conclusion: AML patients with sAML-type mutations had distinct clinical features and poorer outcomes. Incorporating sAML-type mutations can further refine the 2017 ELN risk stratification. It is suggested that AML patients with sAML-type mutations receive more intensive treatment, such as allo-HSCT, and/or novel therapies. S132: TOLERABILITY AND EFFICACY OF THE FIRST-IN-CLASS ANTI-CD47 ANTIBODY MAGROLIMAB COMBINED WITH AZACITIDINE IN FRONTLINE PATIENTS WITH TP53-MUTATED ACUTE MYELOID LEUKEMIA: PHASE 1B RESULTS N. G. Daver^1,*, P. Vyas^2, S. Kambhampati^3, M. M. Al Malki^4, R. Larson^5, A. Asch^6, G. Mannis^7, W. Chai-Ho^8, T. Tanaka^9, T. Bradley^10, D. Jeyakumar^11, E. Wang^12, G. Xing^13, M. Chao^13, G. Ramsingh^13, C. Renard^13, I. Lal^13, J. Zeidner^14, D. Sallman^15 ^1The University of Texas MD Anderson Cancer Center, Houston, United States of America; ^2University of Oxford, Oxford, United Kingdom; ^3Healthcare Midwest, Kansas City; ^4City of Hope National Medical Center, Duarte; ^5University of Chicago, Chicago; ^6University of Oklahoma, Oklahoma City; ^7Stanford University, Stanford; ^8University of California Los Angeles, Los Angeles; ^9University of California San Diego, San Diego; ^10University of Miami, Miami; ^11University of California Irvin, Irvine; ^12Roswell Park Comprehensive Cancer Center, Buffalo; ^13Gilead Sciences, Inc., Foster City; ^14Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill; ^15Moffitt Cancer Center, Tampa, United States of America Background: Patients (pts) with TP53-mutated acute myeloid leukemia (AML) have a poor prognosis, with limited responses to currently available therapies and low survival outcomes, representing a significant unmet medical need. Magrolimab is a monoclonal antibody blocking CD47, a “don’t eat me” signal overexpressed on cells in cancers such as AML. This blockade induces phagocytosis of tumor cells and is synergistic with azacitidine (AZA) via upregulation of “eat me” signals. Aims: To report tolerability and efficacy data from a phase 1b trial of magrolimab + AZA in frontline pts with TP53-mutated AML unsuitable for intensive chemotherapy ([50]NCT03248479). Methods: Frontline pts with AML not suitable for intensive chemotherapy received magrolimab IV starting with a priming dose (1 mg/kg) followed by ramp-up to 30 mg/kg QW or Q2W as the maintenance dose. AZA 75 mg/m^2 was given IV or SC on days 1-7 of each 28-day cycle. Primary endpoints were safety/tolerability and complete remission (CR) rate by European LeukemiaNet (ELN) 2017 criteria. Results: 72 pts with TP53-mutated AML were treated (Table). Common all-grade treatment-emergent adverse events (TEAEs) were constipation (52.8%), diarrhea (47.2%), febrile neutropenia (45.8%), nausea (43.1%), fatigue (37.5%), decreased appetite (37.5%), thrombocytopenia (31.9%), peripheral edema (30.6%), and cough (30.6%). Most common grade ≥3 TEAEs were febrile neutropenia (37.5%), anemia (29.2%; grade 3, 26.4%; grade 4, 2.8%), thrombocytopenia (29.2%), pneumonia (26.4%), and neutropenia (20.8%). Grade 3 hemolysis was reported in 1 pt (1.4%); no grade 4 hemolysis was reported. Objective response rate by intent to treat was 48.6% (CR, 33.3%; CR with incomplete hematologic recovery [CRi]/CR with partial hematologic recovery [CRh], 8.3%; morphologic leukemia-free state [MLFS], 1.4%; partial remission, 5.6%). Stable disease and progressive disease (PD) were reported in 16.7% and 5.6% of pts, respectively; 30- and 60-day mortality rates were 8.3% and 18.1%, respectively. Response assessments were unavailable in an additional 4.2% of pts who discontinued due to AEs and 6.9% due to other reasons, prior to the cycle 3 day 1 assessment. Median time to CR/CRi was 2.2 months (range, 1.7-7.2 months) and to CR was 3.0 months (range, 1.8-9.6 months); 14 of 31 (45.2%) evaluable pts with CR/CRi/CRh/MLFS achieved negative MRD by flow cytometry (investigator reported). Of 24 pts with CR, 8 had a longitudinal TP53 variant allele frequency (VAF) assessment and 5 of 8 (63%) had VAF decreased to ≤5%. Treatment was stopped due to stem cell transplant (9 [12.5%]), PD (26 [36.1%]), death (8 [11.1%]), AE (13 [18.1%]), and other (14 [19.4%]). Median durations of CR and CR/CRi were 7.7 months (95% CI, 4.7-10.9 months) and 8.7 months (95% CI, 5.3-10.9 months), respectively. Median overall survival (OS) in 72 pts was 10.8 months (95% CI, 6.8-12.8 months) (figure), with median follow-up of 8.3 months. Table. Baseline characteristics N=72 Age (range), years 73 (31-89) ECOG, n (%) 0-1 61 (84.7) 2 11 (15.3) ELN cytogenetic risk, n (%) Favorable 1 (1.4) Intermediate 2 (2.8) Adverse 57 (79.2) Unknown 12 (16.7) AML with MDS-related changes, n (%) 34 (47.2) Therapy-related AML, n (%) 15 (20.8) [51]Open in a new tab ECOG, Eastern Cooperative Oncology Group; MDS, myelodysplastic syndrome. Image: graphic file with name hs9-6-1-g021.jpg [52]Open in a new tab Summary/Conclusion: In high-risk frontline pts with TP53-mutated AML unsuitable for intensive chemotherapy, magrolimab + AZA showed durable responses and encouraging OS in a single-arm study. A Phase 3 trial of this combination vs standard of care in TP53-mutated AML (ENHANCE-2; [53]NCT04778397) is ongoing. S133: OFF-THE-SHELF CD33 CAR-NK CELL THERAPY FOR RELAPSE/REFRACTORY AML: FIRST-IN-HUMAN, PHASE I TRIAL R. Huang^1,*, Q. Wen^1, X. Wang^1, H. Yan^1, Y. Ma^1, M. Wang^1, X. Han^1, L. Gao^1, L. Gao^1, C. Zhang^1, X. Zhang^1 ^1Medical Center of Hematology, Xinqiao Hospital of Army Medical University, Chongqing, China Background: The primary results of CAR-T therapy for patients with R/R AML has shown limited efficacy and severe side-effect. One of the main challenges is that current targets for myeloid malignancies are either widely expressed on healthy hemopoietic stem cells such as CD33 or specific for a group of tumor cells presented as Lewis Y antigen, which could cause lasting bone marrow depression induced by “on target off tumor” side-effect or target negative relapse. Therefore, to receive a balance, we designed a CD33 CAR to recognize AML cells and using NK cells to replace T cells as the carrier to eliminate tumor cells. The CD33 CAR NK cells have combined the wide-expression advantage of CD33 target and the safety of NK cells. Aims: To evaluate the safety and primary efficacy of CD33 CAR NK cells Methods: 5 qualified subjects with R/R AML aged between 18 and 65 years-old were enrolled and received round(s) of infusion of anti-CD33 CAR NK cells (6×10^8, 1.2×10^9 or 1.8×10^9 cells per round after the precondition with Fludarabine (30mg/m^2) and Cytoxan 300-500mg/m^2 for 3 days to 5 days, determined by tumor burden at baseline. We investigated the response rate at D28 and treatment related side-effect after the CAR NK cell infusion and the long-term efficacy. Results: As of data cut (February 26, 2021), 5 pts have finished CAR NK cells infusion. The median age was 43 (18-65) years-old and the median tumor burden before infusion is 31% (21%-77.5%). 4 of 5 patients have received MRD negetive CR at day 28 assessment. In dose group one, three patients have received 3 rounds of CAR NK cells (6×10^8, 1.2×10^9 and 1.8×10^9 cells) with the interval of 7 days after last round, and only patient 1 developed grade 1 CRS represented as fever after the infusion of 1.8×10^9 CAR NK cells and alleviated within 24h after symptomatic treatment. Patient 1 and patient 2 received MRD negetive CR at day 14 and 21, but patient 2 relapsed at day 43 after first round infusion. In dosage group 2, patient 4 and patient 5 both received one dose of 1.8 ×10^9 cells CD33 CAR NK cells after precondition, patient 5 developed grade 2 CRS presenting as lasting fever for 6 days after infusion and alleviated after 5mg Dexamethasone I.V., both patient 4 and patient 5 recieved MRD negetive CR at day 28 after infusion. At the time of this abstract being uploaded, three patients remain MRD negetive CR. Image: graphic file with name hs9-6-1-g022.jpg [54]Open in a new tab Summary/Conclusion: Our primary data of the phase I trial have proved the primary efficacy and safety of CD33 CAR NK cells for patients with R/R AML. The efficacy needs expanded samples and longer follow up. S134: INTRA-PATIENT FUNCTIONAL HETEROGENEITY OF AML DETERMINES FIRST-LINE TREATMENT RESPONSE Y. Severin^1,*, Y. Festl^1, M. Roiss^2, T. Benoit^3, T. Heinemann^1, R. Wegmann^1, A.-K. Kienzler^3, M. Bissig^3, M. Scharl^3, M. Manz^3, A. M. Müller^3, B. Snijder^1 ^1Institute of Molecular Systems Biology, ETH Zurich; ^2Medical clinic of Oncology and Haematology, University Hospital Zurich; ^3Medical clinic of Oncology and Haematology, University Hospital Zurich, Zürich, Switzerland Background: Acute myeloid leukemia (AML) is a heterogeneous disease with limited treatment options and poor long-term survival. AML is characterized by a fast clonal expansion of premature myeloid cells with highly variable combinations of chromosomal aberrations and somatic mutations. Patient-to-patient variability is further complicated by the fact that AML partially recapitulates myeloid maturation with a rare leukemic stem cell population at the top of the hierarchy. Although eradication of bulk tumor cells is necessary, matched targeted therapies against the AML stem cell compartment are likely essential for a long lasting cure of patients. Despite this large tumor heterogeneity in de-novo AML patients, most patients receive the same standard-of-care treatment resulting in large differences of treatment success and overall survival. Functional ex vivo drug-response profiling offers a possible route to tailor individual treatments for AML patients. However, the intra-patient functional heterogeneity of divergent AML cell subpopulations, and their contribution to clinical outcome, have not yet been systematically studied. Aims: This study has two aims: 1) To analyze the intra-patient functional heterogeneity of AML blasts, and their contribution to the response to first-line treatment. 2) To improve the clinical response predictions from functional ex vivo drug response profiling by explicit analysis of functional tumor heterogeneity. Methods: In this prospective, non-interventional study, 180 AML patient biopsies were collected with written informed consent (from 44 patients with newly diagnosed AML undergoing intensive induction chemotherapy). Patient matched bone marrow and blood samples were obtained for each donor at three different time points across the course of the treatment. For each sample, we perform ex vivo drug-response profiling analyzed by multiplexed immunofluorescence, automated microscopy, and deep learning-based morphological profiling to determine drug sensitivities on a single-cell level. The drug library includes 100 distinct drugs and drug combinations, including standard of care induction chemotherapy, and single-cell drug sensitivity data is analyzed as a function of cellular clone and blast maturation state. The drug response data is integrated with patient-matched serum cytokine profiling, RNA-sequencing, and clinical data, allowing to identify the cellular subpopulations whose functional response is predictive of treatment response. Results: Our workflow excelled at predicting patient responses to first line treatment, with an overall predictive power of 85% accuracy and a diagnostic odds ratio of 47. Molecular and morphology-based phenotyping of cancer subpopulations strengthened clinical associations and revealed blast maturity as well as immune activation to be strong predictors of treatment success and the risk of chemotherapy-related infectious complications. Furthermore, the screening of 100 unique drugs and combinations enabled individualized ex vivo treatment predictions, suggesting more potent treatment options than those currently deployed. Image: graphic file with name hs9-6-1-g023.jpg [55]Open in a new tab Summary/Conclusion: Maturation-associated intra-tumor drug response heterogeneity is a strong predictor of clinical response to first-line intensive induction chemotherapy in AML. Our multiplexed ex vivo drug screening framework is highly flexible, allowing us to discover, molecularly describe, and target clinically relevant AML subpopulations on a personal basis, leading to improved personalized treatment recommendations. S135: CRISPR/CAS9 EDITING REVEALS DEPENDENCE OF HUMAN RICHTER SYNDROME AND MURINE CLL CELLS ON SIGNALS FROM B CELL RECEPTOR, CXCR4 RECEPTOR AND MACROPHAGES BUT NOT FROM TOLL-LIKE RECEPTORS IN VIVO C. Martines^1,*, S. Chakraborty^1, V. Guastafierro^1, M. Vujovikj^2, T. Vaisitti^3, S. Deaglio^3, L. Laurenti^4, A. Dimovski^2, D. Efremov^1 ^1Molecular Hematology, International Centre for Genetic Engineering and Biotechnology, Trieste, Italy; ^2Research Center for Genetic Engineering and Biotechnology, Macedonian Academy of Sciences and Arts, Skopje, North Macedonia, Republic of; ^3Functional Genomics Unit, Department of Medical Sciences, University of Turin, Torino; ^4Department of Hematology, Catholic University Hospital “A. Gemelli”, Rome, Italy Background: Numerous signals from the microenvironment have been identified that can increase the survival or induce the proliferation of chronic lymphocytic leukemia (CLL) cells in vitro. These signals typically represent various secreted or cell surface ligands that are expressed by different cell types present in the lymph node tumor microenvironment, such as T cells, macrophages and stromal cells, or molecules that would be expected to be released by apoptotic cells, such as apoptosis associated autoantigens or CpG-unmethylated mitochondrial DNA. However, the extent to which these different signals contribute to the growth and survival of the leukemic cells in vivo has still not been fully established. Aims: To determine the relevance of signals from Toll-like receptors (TLRs), the B cell receptor (BCR) and the chemokine receptor CXCR4 for the growth of murine Eμ-TCL1 CLL cells and human Richter Syndrome (RS) patient-derived xenograft (PDX) cells in vivo. Methods: To understand the impact of pharmacological inhibition of the TLR pathway, immunocompetent C57BL/6 or immunodeficient NSG mice were transplanted with murine Eμ-TCL1 CLL or human RS-PDX cells, respectively, and were treated with an inhibitor of the kinase IRAK4 or vehicle control. To determine the effects of genetic disruption of the TLR, BCR and CXCR4 pathways, the IRAK4, IgM heavy chain (IGHM) and CXCR4 genes were disrupted by CRISPR/Cas9 editing in the murine Eμ-TCL1 CLL cells or the human RS-PDX lines RS9737, RS1316 and IP867/17 prior to transplantation. The effects of genetic disruption of these pathways on the growth of the malignant cells in vivo were investigated by analyzing the proportion of mutant and wild type alleles in different anatomical compartments of the transplanted mice. Results: Treatment with the IRAK4 inhibitor significantly prolonged the survival of C57BL/6 mice transplanted with murine Eμ-TCL1 CLL cells and significantly delayed the growth of the human RS-PDX lines RS9737 and RS1316 xenografted in immunodeficient NSG mice. However, genetic disruption of IRAK4 in the human RS-PDX lines RS9737, RS1316 and IP867/17 or of the TLR adaptor MyD88 in the murine Eμ-TCL1 CLL cells did not result in negative selection of these cells in vivo, suggesting that these tumors do not receive or do not rely on TLR signals for their growth and that the therapeutic activity of the IRAK4 inhibitor is not caused by disruption of TLR signaling in the malignant cells themselves. In contrast, genetic disruption of the IGHM or CXCR4 gene was associated with significantly reduced growth of these cells compared to their wild type counterparts and resulted in almost complete disappearance of the mutated cells at later timepoints following transplantation. Analysis of the effects of the IRAK4 inhibitor on other cell types from the tumor microenvironment revealed a significant reduction in the number of macrophages, which in co-culture experiments were shown to strongly protect human Richter syndrome and murine Eμ-TCL1 leukemia cells from spontaneous apoptosis. Summary/Conclusion: These data provide evidence that signals from the BCR, CXCR4 and macrophages support the growth and survival of CLL and RS cells in vivo and argue against a role for TLR signals in the pathogenesis of CLL. In addition, they suggest that targeting the TLR pathway may potentially provide a therapeutic benefit in CLL, but that this benefit would be derived from inhibition of TLR signaling in monocytes and macrophages rather than inhibition of TLR signaling in the malignant cells themselves. S136: CONSTITUTIVE VLA-4 ACTIVATION IN CHRONIC LYMPHOCYTIC LEUKEMIA. THE OTHER SIDE OF BCR AUTONOMOUS SIGNALING. E. Tissino^1,*, R. Bomben^1, P. C. Maity^2, F. Pozzo^1, G. Forestieri^3, A. Nicolò^2, A. Härzschel^4, T. Bittolo^1, F. M. Rossi^1, M. Datta^2, F. Zaja^5, G. Capasso^1, G. D’Arena^6, A. Chiarenza^7, F. Di Raimondo^7, H. Jumaa^2, D. Rossi^3, G. Del Poeta^8, T. N. Hartmann^4, A. Zucchetto^1, V. Gattei^1 ^1Clinical and Experimental Onco-Hematology, Centro di Riferimento Oncologico, Aviano, Italy; ^2Institut für Immunologie, Universitätsklinikum Ulm, Ulm, Germany; ^3Division of Hematology, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland; ^4Department of Internal Medicine, Faculty of Medicine, University Medical Center Freiburg, University of Freiburg, Freiburg, Germany; ^5Dipartimento di Scienze Mediche e Chirurgiche e della Salute, University of Trieste, Trieste; ^6IRCCS Centro Di Riferimento Oncologico Della Basilicata, IRCCS Centro Di Riferimento Oncologico Della Basilicata, Rionero In Vulture; ^7Divisione di Ematologia, Ospedale Ferrarotto, A.O.U. Policlinico-OVE, Università di Catania, Catania; ^8Division Hematology, S.Eugenio Hospital and University of Tor Vergata, Rome, Italy Background: In chronic lymphocytic leukemia (CLL), CD49d, the alpha chain of the heterodimer CD49d/CD29 (VLA-4), is a strong negative prognosticator and key player of CLL microenvironmental interactions. The adhesive properties of VLA-4 can be rapidly inside-out activated by signals through the B-cell receptor (BCR), thus favoring the capability of the integrin to interact with its ligands. In CLL, beside the canonical antigen (Ag)-dependent mechanism, BCR signaling has been recently demonstrated to occur via an autonomous Ag-independent manner. Aims: To investigate the role of autonomous BCR signaling on constitutive VLA-4 activation state in CLL. Methods: VLA-4 activation/affinity was determined by flow cytometry (FC) using the conformation-sensitive anti-CD29 mAb HUTS21 and/or by “real-time” (FC) measuring the binding of the VLA-4 ligand LDV-FITC as reported (Tissino et al, J Exp Med, 2018) in: i) 1,984 consecutive CLL all with IGHV gene mutations/BCR features available; ii) sequential samples (0, 14, 30, 60 90 days) from CLL patients (n=26) treated in vivo with ibrutinib (IB) in real-world and from a clinical trial ([56]NCT02827617). HUTS21 staining was also performed in the presence of: plasma depletion/replacement, soluble (s) VCAM-1, fibronectin (FN) and blocking anti-CD49d (HP1/2) mAbs. ELISA assays were used to quantify sVCAM-1 in plasma samples (n=122). BCR signaling were investigated by Ca^++ influx assay in a 4-OH tamoxifen (4-OHT)-inducible murine TKO cell model (Dühren-von Minden M et al, Nature, 2012). Results: Out of 1,984 CLL, 1,070 (54%) expressed CD49d (cutoff 30%) and, among them, 250/1,070 (23%) were HUTS21+ (cutoff 20%), indicating an activated VLA-4 conformation. HUTS21 staining was: i) impaired by depletion of plasma from whole blood samples, and reconstituted by specific plasma components (sVCAM-1, FN); ii) impaired by pre-incubation with anti-CD49d HP1/2 blocking mAbs before addition of plasma, sVCAM-1 and FN. sVCAM-1 was higher in CD49d+ vs CD49d- CLL (p<0.0001); among CD49d+ cases, sVCAM-1 was lower in HUTS21+ (i.e. VLA-4 activated) cases (p=0.0096), suggesting ligand sequestration by activated surface VLA-4. CLL with mutated IGHV expressed higher levels of activated VLA-4 compared to unmutated IGHV CLL (p=0.001). Higher levels of activated VLA-4 were found in CLL using the IGHV3 and IGHV4 families, compared to cases using the IGHV1 family (p=0.043 and p=0.004). Finally, analysis of BCR stereotypy highlighted higher VLA-4 activation levels in CLL from subset#2 compared to CLL from subset#1 (p=0.02). To validate these data, murine TKO cells, expressing high VLA-4 levels, were transfected with different BCRs derived from 4 CLL with high level of constitutively activated VLA-4 (TKO-high) and 4 CLL with low level of constitutively activated VLA-4 (TKO-low). Compared to TKO-low cells, TKO-high cells showed a higher autonomous Ca^++ influx (p=0.03), and consistently higher VLA-4 affinity (p=0.01). IB treatment impaired both BCR autonomous signaling and VLA-4 affinity. Notably, anti-IgM stimulation induced high Ca^++ influx and high VLA-4 affinity state in both TKO-high and TKO-low, irrespective of IB treatment. According to TKO data, a decreased constitutive VLA-4 activation was observed in CLL cells collected at pre-treatment and at day 14, 30, 60 and 90 from patients on IB, confirming an IB-dependent impairment of VLA-4 activation via BCR signal. Summary/Conclusion: The presence of a constitutively activated form of VLA-4 is observed in a fraction of CD49d+ CLL, due to a continuous VLA-4 inside-out stimulation derived from autonomous BCR signaling. S137: SMALL EXTRACELLULAR VESICLES IN THE LEUKEMIA MICROENVIRONMENT SUSTAIN CLL PROGRESSION BY HAMPERING T CELL-MEDIATED ANTI-TUMOR IMMUNITY E. Gargiulo^1,*, E. Viry^1, P. E. Morande^1 2, A. Largeot^1, S. Gonder^1 3, F. Xian^1 4, N. Ioannou^5, M. Benzarti^1 3, F. Kleine Borgmann^1 3 6, M. Mittelbronn^1 3 7, G. Dittmar^1 3, P. V. Nazarov^1, J. Meiser^1, B. Stamatopoulos^8, A. G. Ramsay^5, E. Moussay^1, J. Paggetti^1 ^1Luxembourg Institute of Health, Luxembourg, Luxembourg; ^2Instituto de Medicina Experimental, Buenos Aires, Argentina; ^3University of Luxembourg, Luxembourg, Luxembourg; ^4University of Vienna, Vienna, Austria; ^5King’s College London, London, United Kingdom; ^6Centre Hospitalier de Luxembourg, Luxembourg; ^7Laboratoire national de santé, Dudelange, Luxembourg; ^8Université Libre de Bruxelles, Brussels, Belgium Background: Small extracellular vesicles (sEV) are nano-sized particles released by every cell and found in all biofluids. Given their composition and abundance, sEV are commonly involved in cell-to-cell communication through the transfer of genetic material and proteins. Furthermore, sEV possess direct functions carried out by sEV-ligands capable to affect the biological functions of targeted cells. In cancer, tumor-derived sEV are involved in the re-education of microenvironment (ME) cells promoting tumor proliferation, immune escape and metastasis. We previously demonstrated that leukemia-derived sEV are involved in the re-education of surrounding cells and increased immune escape. Indeed, chronic lymphocytic leukemia (CLL)-derived sEV induce stromal cell conversion into cancer-associated fibroblasts (Paggetti et al., Blood, 2015), and modulate PD-L1 expression in monocytes (Haderk et al. Science Immunology, 2017). Aims: The goal of the present work was to characterize leukemia ME-derived sEV (LME-sEV) and to evaluate their role in the disease development and progression in vivo. Methods: To obtain a biological representation of sEV in CLL microenvironment, we isolated LME-sEV directly from spleens of leukemic mice, obtaining a complex mix of sEV released by both CLL and ME cells alike. Small EV characterization was performed using a wide range of techniques, including qPCR, mass spectrometry and single sEV flow cytometry (FC). The effect on target cells was evaluated both ex vivo and in vivo using high-throughput techniques, FC, qPCR and cytotoxic assay. Small EV impact on CLL development in vivo was evaluated by generating a novel preclinical mouse model in which sEV release is genetically impaired due to Rab27a/b knock-out. Finally, we analyzed the expression of sEV-related genes in a cohort of 144 CLL patients using qPCR followed by regression analysis. Results: LME-sEV showed a distinct proteome (A) and RNA contents compared to healthy counterparts (HCME-sEV), including miRNA enriched in the plasma of CLL patients. Furthermore, FC-based immune checkpoint (ICP) screening showed the presence of multiple ICP ligands anchored on CLL-derived sEV (CD20^+ subset of LME-sEV) (B), while high expression of the corresponding ICP receptors was found on T cells from matching LME. We also found that LME-sEV are internalized by different T cell subsets, thus we performed in vivo and ex vivo functional studies to assess sEV impact on T cells. High-throughput analysis of cells isolated from spleens of control mice treated with LME-sEV revealed considerable physiological changes mainly in CD8^+ T cells. Indeed, CD8^+ T cells showed alterations in their transcriptome, proteome and metabolome leading to cell exhaustion, decreased functions and survival. In line with this, absence of sEV dramatically delayed CLL progression in vivo. This effect was due to CLL inability to escape immune surveillance in absence of sEV and this was rescued by LME-sEV treatment (C). Finally, we identified a consistent sEV gene signatures in CLL patients correlating with treatment-free survival, overall survival, and with unfavorable clinical parameters routinely used in CLL diagnosis and prognosis (D). Image: graphic file with name hs9-6-1-g024.jpg [57]Open in a new tab Summary/Conclusion: By using different preclinical murine models and strategies, our results demonstrated for the first time that sEV in CLL ME play a key pro-tumoral role in leukemia development by negatively affecting the anti-tumor immune response. Furthermore, high expression of sEV-related genes correlated with poor survival and clinical parameters in CLL patients, suggesting sEV profiling as prognostic tool in CLL. S138: MICROENVIRONMENT-REGULATED TRANSCRIPTIONS FACTORS AHR AND HIF-1Α EXPRESSION IN TREGS PROMOTE CLL PROGRESSION BY IMPAIRING CD8+ T-CELL MEDIATED ANTI-TUMOR IMMUNITY G. Pagano^1,*, M. Wierz^1, I. Fernandez Botana^1, S. Gonder^1, E. Gargiulo^1, E. Moussay^1, J. Paggetti^1 ^1Department of Cancer Research, Luxembourg Institute of Health, Luxembourg, Luxembourg Background: Chronic Lymphocytic Leukemia (CLL) progression is highly dependent on complex interactions between tumor cells and the tumor microenvironment (TME). Indeed, CLL cells can modify stromal cells and immune cells to promote the survival of the leukemic clone and to escape from the immune system surveillance. Within the TME, regulatory T cells (Tregs) represent a subtype of CD4+ T cells with immunosuppressive abilities, causing the evasion of cancer cells from the immune system. We previously characterized extensively the immune microenvironment of pre-clinical CLL mouse models using mass cytometry, and we described a significant increase in the Tregs subsets with an enhanced immunosuppressive and activated phenotype compared to non-leukemic animals (Wierz et al., Blood, 2018). Interestingly, TIGIT+ Tregs are more immunosuppressive than their TIGIT- Treg counterparts and express higher levels of several transcription factors, including Ahr and Hif1α (Joller et al., Immunity, 2014), both involved in the cellular response to microenvironment-mediated stimuli. Aims: The aim of the present study is to investigate the role of AHR and HIF-1α in the suppressive ability of Tregs during CLL development. Methods: We generated conditional knock out mice (cKO) lacking Ahr or Hif1a genes exclusively in Tregs (Foxp3^YFP-Cre Ahr^fx/fx and Foxp3^YFP-Cre Hif1a^fx/fx mice). We then performed adoptive transfer (AT) of CLL cells obtained from diseased Eµ-TCL1 mice into cKO and control mice. In order to decipher the mechanism by which AHR and HIF-1α pathways in Tregs affect CLL progression, we analyzed the splenic TME of recipient mice and evaluated immune checkpoint expression and cytokine production. Finally, we evaluated the suppression ability of the regulatory T cells with ex vivo suppression assays. Results: Generated cKO mice showed no sign of abnormalities or autoimmune phenotypes, and immune cells phenotyping revealed no major differences. However, we showed that CLL growth in cKO mice was drastically delayed compared to the control mice (A). Interestingly, this decrease was mitigated when CD8+ T cells were depleted. The analysis of the splenic TME in recipient cKO mice revealed an increase in IL-17 and TNF-α production, two major T-cell cytokines, in CD4+ T cells as compared to their WT counterparts. We also measured the expression of immune checkpoints and activation markers in the different T cell subpopulations and observed that Tregs lacking AHR or HIF-1α show decreased levels of the immune checkpoints CTLA-4 and TIGIT, two key proteins for the suppressive functions of Tregs. Ex vivo suppression assays demonstrated an increased proliferation of CD8+ T cells in the presence cKO Tregs (B-C). This result confirmed the decreased suppressive ability of Tregs in absence of the two transcription factors, explaining the observed delay in CLL progression in cKO mice. Image: graphic file with name hs9-6-1-g025.jpg [58]Open in a new tab Summary/Conclusion: Altogether, these results indicate that the TME-regulated transcription factors AHR and HIF-1α in Tregs are crucial for CLL development by promoting escape to anti-tumor immune response, and therefore represent potential therapeutic targets during CLL progression. S139: BCOR DELETION SUSTAINS NOTCH1 SIGNALLING ACTIVATION TO ACCELERATE CHRONIC LYMPHOCYTIC LEUKEMIA (CLL) PROGRESSION TOWARD RICHTER TRANSFORMATION IN MICE C. Rompietti^1,*, D. Sorcini^1, F. De Falco^1, E. Dorillo^1, F. M. Adamo^1, E. C. Silva Barcelos^1, A. Stella^1, A. Scialdone^1, A. Esposito^1, R. Arcaleni^1, B. Bigerna^2, G. Martino^3, L. Moretti^2, M. G. Mameli^2, C. Geraci^1, L. Sandoletti^1, A. Cipiciani^1, E. Rosati^4, B. Falini^1, P. Sportoletti^1 ^1Medicine and Surgery, University of Perugia (Center for Hemato-Oncology Research); ^2Hospital of Perugia, Perugia; ^3Pathology Unit, Azienda Ospedaliera Santa Maria di Terni, University of Perugia, Terni; ^4Medicine and Surgery, University of Perugia, Perugia, Italy Background: BCL6 co-repressor (BCOR) is a transcription factor involved in various biological processes including lymphoid development. BCOR disruptive mutations were found in up to 2% of CLL and frequently associated with aberrations of NOTCH1, one of the most common genetic alterations with poor prognosis in CLL and Richter transformation (RT). Recent evidence also indicates that BCOR is involved in NOTCH signalling suppression during embryogenesis. These data indicate the need for further investigation on the role of BCOR mutation and its interplay with NOTCH1 in CLL pathogenesis. Aims: We aim to elucidate the impact of Bcor deficiency in CLL and RT, focusing on the role of active NOTCH1 signalling. Methods: We used a conditional knockout mouse of Bcor (Sportoletti et al, Leukemia 2021) crossed with CD19-Cre mice, to specifically delete Bcor in B-cells, and with the Eμ-TCL1 mouse model of CLL. Mice were characterized for disease phenotype (using flow-cytometry and histological analyses), overall survival and drug response. NOTCH1 activity was assessed by western blot (WB) for the NOTCH1-intracellular domain (NOTCH1-IC) and HES1 expression. Results: B-cell restricted loss of Bcor in Eµ-TCL1 mice significantly expanded leukemic CD5+CD19+ cells in the peripheral blood (PB; p<0.05), spleen (p<0.01), bone marrow (BM; p<0.001), compared to Eµ-TCL1 control mice. No leukemic cells were found in BcorCD19Cre+ mice. Cellular changes resulted in a poorer survival of double mutant mice compared to single-mutant and wild-type (WT) controls (median survival of 334 days vs unreached in BcorCD19Cre+;Eμ-TCL1 vs the other groups, respectively). Adoptive transfer of spleen cells (CD5+CD19+ leukemic burden >50%) resulted in a more rapidly lethal disease in recipient of BcorCD19Cre+;Eμ-TCL1 compared to Eμ-TCL1 cells (median survival of 37.5 vs 74 days, respectively, p<0.05; Figure1A). At necropsy, double mutant mice presented massive splenomegaly, whose histopathological examination revealed a diffuse infiltration by high-mitotic blastoid cells, significantly increased in size, defining a high-grade lymphoid malignancy distinct from the leukemia of Eμ-TCL1 counterparts (Fig.1B). In order to gain mechanistic insight relating to the phenotypic observations, we measured NOTCH1 activity. BcorCD19Cre+;Eμ-TCL1 splenic CD5+CD19+ cells showed significantly increased levels of the active NOTCH1-IC and the up-regulation of its direct target HES1, compared to leukemic cells from Eμ-TCL1 mice and CD19+ cells from non-leukemic BcorCD19Cre+ and WT mice used as control (Fig.1C). In vivo treatment with the NOTCH1 inhibitor Bepridil resulted in a significant growth delay of CD5+CD19+ cells in the PB of mice transplanted with BcorCD19Cre+;Eμ-TCL1 leukemic cells compared to vehicle. At sacrifice, Bepridil caused a significant reduction of spleen dimensions and CD5+CD19+ cellularity, associated with reduced levels of active NOTCH1-IC, c-MYC and HES1 compared to vehicle (28%, 54% and 38% reduction, respectively). NOTCH1 inhibition significantly improved the survival of diseased mice (median survival of 39 vs 33 days in Bepridil vs vehicle, respectively; N=4, p<0.05). Image: graphic file with name hs9-6-1-g026.jpg [59]Open in a new tab Summary/Conclusion: We showed for the first time the tumour suppressor activity of Bcor in a CLL mouse model, ultimately leading to transformation towards a high-grade lymphoma, mimicking human RT. Mechanistically, we implied NOTCH1 signalling activation in Bcor loss mediated tumorigenesis with potential for targeted treatment for high-risk CLL and RT patients. S140: SINGLE-CELL MULTIOMICS ANALYSES REVEAL COMPLEX INTRA-PATIENT HETEROGENEITY IN RELAPSED CLL FOLLOWING VENETOCLAX THERAPY R. Thijssen^1 2,*, L. Tian^1 2, C. Flensburg^1 2, M. A. Anderson^1 2 3 4, A. Jarratt^1 2, H. Peng^1 2, I. Majewski^1 2, C. Tam^3 4, J. Seymour^3 4, P. Blombery^3 4, M. Ritchie^1 2, D. Huang^1 2, A. Roberts^1 2 3 4 ^1The Walter and Eliza Hall Institute of Medical Research; ^2Medical Biology, University of Melbourne; ^3Clinical Haematology, Royal Melbourne Hospital and Peter MacCallum Cancer Centre; ^4Medicine, University of Melbourne, Melbourne, Australia Background: Venetoclax (VEN) is the first approved-in-class BH3-mimetic therapy that inhibits pro-survival BCL2 to induce apoptosis. It is now a standard of care for patients with chronic lymphocytic leukaemia (CLL) and acute myeloid leukaemia. Despite high remission rates, secondary resistance remains problematic with disease progressing on continuous therapy in most patients. While previous studies by ourselves and others demonstrated a BCL2 mutation (Blombery et al., 2019 Cancer Disc) or MCL1 amplification (Guièze et al., 2019 Cancer Cell) confer resistance in patient samples, it is clear that these only account for a fraction of tumour cells at disease progression and are not present in all leukaemias. Given that these changes only provide partial explanations, other mechanisms must operate to subvert the action of VEN. Aims: To uncover why VEN fails in patients by assessing heterogeneity, distinguishing cells with known changes (e.g., BCL2 mutation, MCL1 amplification) and to elucidate what is happening in the cells not harbouring these alterations. Methods: We applied a single-cell (sc) sequencing method that simultaneously measures gene expression and full-length transcripts to determine genotype on samples from patients with progressive CLL with long term follow-up on trials of continuous VEN therapy (Figure 1). Results: By applying scRNA-seq to 161,499 tumour cells, we discovered a high degree of inter- and intra-patient heterogeneity. The majority of CLL cells displayed an altered transcriptional profile at relapse compared to before VEN treatment. 8/13 VEN-relapse samples demonstrated a complex sub-clonal architecture (Figure 1). By applying scLong-read sequencing, we detected BCL2 mutations in 4/13 samples at relapse. We also identified subclones with loss of NOXA or BAX, but overall, alterations in genes for the pro-apoptotic proteins were uncommon. In contrast, altered expression of the pro-survival genes was common and high MCL1 expression was seen in most of the CLL cells in 11/13 relapses. This could only be fully explained by MCL1 amplification in 1 patient and partially in 2 (Figure 1). At VEN resistance, we identified near universal activation of NF-ΚB in circulating tumour cells and this was highly correlated with MCL1 expression. We next demonstrated that the MCL1 locus was transcriptionally targeted by NF-ΚB accounting for the MCL1 increase. NF-ΚB activation and associated high MCL1 was also observed in cells harbouring the BCL2 mutations. Strikingly, NF-ΚB activation (and consequent increased MCL1) dissipated after VEN cessation and commencement of a BTK inhibitor as the leukaemic cells substantially recovered their in vitro sensitivity to VEN. Consistent with the hypothesis that VEN was driving these changes, NF-ΚB activation (and high MCL1) was not detected in a patient (CLL26) with disease relapse 3 years after ceasing VEN in CR. The disease in this patient responded promptly to VEN reinduction. Image: graphic file with name hs9-6-1-g027.jpg [60]Open in a new tab Summary/Conclusion: Taken together, our findings provide a much clearer understanding of resistance to VEN and provide impetus for improved VEN-based clinical management of CLL patients. Given the multiple ways a CLL population in a patient can become resistant to ongoing VEN, it seems unlikely that simply adding other drugs at relapse will be durably effective in most patients. The data pinpoint NF-ΚB activation as a biomarker for in vivo VEN-resistance and provide a specific biological rationale for ceasing VEN in deep response, as is currently being used in time-limited and explored in response-directed regimens. S141: ELICITING ANTI-TUMOR T CELL ACTIVITY IN CHRONIC LYMPHOCYTIC LEUKEMIA WITH BISPECIFIC ANTIBODY-BASED COMBINATION THERAPY D. Papazoglou^1,*, L. Ysebaert^2, N. Ioannou^1, B. Apollonio^1, P. Patten^3, S. Herter^4, M. Bacac^4, A. Deutsch^5, C. Klein^4, A. Vardi^6, A. Quillet-Mary^2, A. Ramsay^1 ^1Hemato-oncology, King’s College London, London, United Kingdom; ^2Centre de Recherches en Cancérologie de Toulouse, Toulouse, France; ^3King’s Health Partners, King’s College Hospital, London, United Kingdom; ^4Roche Innovation Center Zurich, Roche Pharma Research and Early Development, Zurich, Switzerland; ^5Medical University of Graz, Graz, Austria; ^6Hematology Department and HCT Unit, G. Papanikolaou Hospital, Thessaloniki, Greece Background: Identifying effective combination immunotherapy could offer hope to chronic lymphocytic leukemia (CLL) patients who relapse on previous lines of therapy. Although prior studies have described an exhausted and pro-tumor T cell state, there may be potential to stimulate anti-CLL cytolytic T cell activity with novel therapy. Aims: Investigate the ability of the CD20-targeted T-cell-engaging bispecific antibody glofitamab (CD20-TCB) to overcome T cell exhaustion and tumor microenvironment (TME)-mediated immunosuppression in CLL. Methods: We pre-clinically investigate CD20-TCB efficacy in CLL utilizing immune and TME model assays that include in vitro functional T cell assays, patient-derived xenografts and lymph node organotypic models. Results: Our previous studies of the CLL lymph node (LN) TMEs suggested that low numbers of infiltrated CD8^+ T cells likely contribute to the generation of a “cold” TME that could represent a barrier to effective therapy. Here, we show that CD20-TCB treatment induced the activation, proliferation and migration of previously exhausted patient CD4^+ and CD8^+T cells and triggered the release of immunostimulatory cytokines related to immune cell recruitment and cytotoxic function. Additionally, we detected enhanced cytolytic (perforin^+) immune synapse activity in CD20-TCB-treated cultures which correlated with high levels of T cell-mediated CLL cell death. Importantly, bispecific antibody treatment also induced high levels of anti-CLL T cell activity in both ibrutinib responding and refractory patient samples. Given the emerging evidence that endothelial cells (ECs) that reside within the TME can possess immunomodulatory activity, we next evaluated CD20-TCB-mediated T cell responses in the presence of this stromal cell compartment. We established triple culture assays that allowed patient T and CLL cells to be cultured with CLL-activated ECs (CLL-ECs). These assays revealed that TCB-activated T cells in the presence of CLL-ECs showed a reduced ability to kill target CLL cells compared to healthy ECs. Flow cytometric and multicolor confocal microscopy analysis revealed that CLL-ECs in vitro and in situ expressed increased levels of ICAM-1 and PD-1 ligands that we demonstrated contribute to T cell suppression. In order to overcome this TME-mediated inhibition and enhance anti-CLL CD20-TCB activity, we next investigated combination immunotherapy. Combination of CD20-TCB with a CD19-4-1BBL fusion protein (tumor-targeted co-stimulatory drug) resulted in enhanced CLL cell killing in both T:CLL and triple T:CLL-EC culture assays. Patient-derived xenograft models further confirmed the ability of bispecific antibody combination therapy to induce higher levels of anti-CLL T cell activity compared to monotherapy. Additionally, the pairing of CD20-TCB with CD19-4-1BBL promoted the proliferation and activation of patient T cells within splenic TMEs. Finally, we established 3D organotypic cultures of excess diagnostic LN biopsies that preserve the intact TME cellular composition and spatial organization. LNs of treatment naïve, as well as ibrutinib and venetoclax relapsed CLL patient samples, were treated ex vivo with CD20-TCB combination for subsequent 3D image analysis. Importantly, combination immunotherapy triggered improved anti-tumor T cell killing function and promoted an inflammatory cytokine-rich “hot” TME. Summary/Conclusion: Collectively, this preclinical study supports the concept of triggering cytolytic T cell activity in CLL with bispecific antibody combination immunotherapy that could help overcome T cell exhaustion and the immunosuppressive TME. S142: DECIPHERING THE COMPLEXITY OF T CELLS IN BLOOD AND LYMPH NODES OF PATIENTS WITH CLL BY INTEGRATIVE SINGLE-CELL RNA-SEQ AND MASS CYTOMETRY ANALYSES L. Llaó Cid^1, J. Wong^1, M. Wierz^2, Y. Paul^1, S. Gonder^2, I. Fernandez Botana^2, T. Roider^3, S. Dietrich^3, D. Colomer^4, P. Lichter^1, M. Zapatka^1, E. Moussay^2, J. Paggetti^2, M. Seiffert^1,* ^1Molecular Genetics, B060, German Cancer Research Center, Heidelberg, Germany; ^2Department of Oncology, Luxembourg Institute of Health, Luxembourg, Luxembourg; ^3Department of Medicine V, University Clinic Heidelberg, Heidelberg, Germany; ^4Hematopathology Unit, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain Background: Abnormal distribution and impaired function of T cells are key features of chronic lymphocytic leukemia (CLL), a malignancy of mature B lymphocytes that develops in secondary lymphoid tissue and blood. These defects have been linked to failure of immune control and resistance to immunotherapy including immune checkpoint or CAR T-cell therapy. A better characterization of T cells and their loss of function in CLL will help to improve such treatment approaches. Aims: The goal of this study was an in-depth characterization of the T-cell compartment in blood and tissue samples of patients and mouse models with CLL to gain insights into the spectrum of phenotypes and transcriptional programs of T cells and the underlying mechanisms of their development. Methods: We performed mass cytometry (CyTOF) with a panel of 35 antibodies to characterize T cells in blood (n=8), bone marrow (n=3), and lymph nodes (n=21) of CLL patients, as well as reactive lymph nodes of non-tumor patients (n=13). We further generated single-cell transcriptome and T-cell receptor sequencing data of T cells from lymph nodes of CLL patients (n=5) and spleen samples of the Eµ-TCL1 mouse model (n=3), and performed integrative analyses of all data sets. Results: CyTOF analysis allowed for the identification and quantification of 15 clusters of CD4^+ and 14 clusters of CD8^+ T cells in blood, lymph node and bone marrow samples. T cells in blood and bone marrow were clearly distinct from lymph node derived cells, and several cell subsets showed a positively correlated abundance. A comparative analysis revealed an accumulation of several regulatory T-cell subsets, as well as T cells harboring an exhausted and dysfunctional phenotype in CLL lymph nodes. Single-cell transcriptome and T-cell receptor sequencing demonstrated the presence of clonally expanded and gradually exhausted T-cell clusters in human and murine CLL, and provided insights into the transcriptional programs and regulatory networks of these cell subtypes. Using CellPhoneDB (Efremova et al., Nature Protocols, 2020) and CellChat (Jin et al., Nature Communications, 2021) repositories, we identified novel ligand-receptor-interactions between CLL and T-cell clusters with impact on anti-tumor immune control which are currently tested in preclinical models for intervention therapy. Summary/Conclusion: Altogether, our study provides a detailed characterization of the T-cell compartment in CLL that helps us to understand T-cell exhaustion and suggests novel targets to improve immunotherapy for patients with CLL and likely also other malignancies. MZ, EM, JP and MS share senior authorship. S143: TRANSCRIPTOMIC CHARACTERIZATION OF MRD RESPONSE AND NON-RESPONSE IN PATIENTS TREATED WITH FIXED-DURATION VENETOCLAX-OBINUTUZUMAB O. Al-Sawaf^1 2 3,*, H. Y. Jin^4, C. Zhang^1, Y. Choi^4, S. Balasubramanian^4, S. Robrecht^1, A. Kotak^5, N. Chang^5, A.-M. Fink^1, E. Tausch^6, C. Schneider^6, M. Ritgen^7, K.-A. Kreuzer^1, B. Chyla^8, J. Paulson^4, B. Eichhorst^1, S. Stilgenbauer^6, Y. Jiang^4, M. Hallek^1, K. Fischer^1 ^1Department I of Internal Medicine and Center of Integrated Oncology Aachen Cologne Bonn Duesseldorf, University Hospital of Cologne, Cologne, Germany; ^2Cancer Institute, University College London; ^3Francis Crick Institute, London, United Kingdom; ^4Genentech Inc., South San Francisco, CA, United States of America; ^5Roche Products Ltd, Welwyn Garden City, United Kingdom; ^6Department III of Internal Medicine, Ulm University, Ulm; ^7Department II of Internal Medicine, University of Schleswig Holstein, Kiel, Germany; ^8AbbVie Inc., North Chicago, IL, United States of America Background: Minimal residual disease (MRD) is a key surrogate for the depth of remission of CLL. The highest rates of undetectable MRD have been observed with regimens using the BH3 mimetic venetoclax. In CLL14, most patients (pts) receiving venetoclax-obinutuzumab (Ven-Obi) had undetectable MRD (uMRD <10^-4) at the end of treatment (EoT). However, a subgroup of pts remained MRD positive, regardless of TP53 aberrations or unmutated IGHV status. The biological drivers of response or non-response have so far not been elucidated. Aims: This study explores whether differential transcriptomic profiles can discriminate between pts with vs without deep MRD remissions, particularly after Ven exposure. This might eventually allow for biology-informed treatment development for pts in this high-risk disease setting. Methods: NGS-based MRD measurements (clonoSEQ) from peripheral blood (PB) at follow-up month 3 were grouped into uMRD <10^-6 (MRD6), 10^-6 ≤ and <10^-4 (MRD4/5), and detectable MRD ≥10^-4 (MRD+). Pre-treatment CD19-enriched PB samples, and samples collected at relapse from a subset of pts, were subjected to bulk RNAseq on the Illumina NovaSeq platform. Transcriptomic clustering, linear modeling, differential expression (DE), and gene set enrichment analyses were run using the R packages Seurat, DESeq2, fGSEA, and GSVA. Top DE genes were selected based on P-values <0.05 and log2-fold-change >0.5, comparing MRD+ and MRD6 populations. Results: Within the intention-to-treat (ITT) population after a median observation time of 65.4 months, pts with MRD6 had a significantly longer PFS than pts who had MRD4/MRD5 (A). Pre-treatment RNAseq data were available for 405 of the 432 pts in the ITT population (202 Clb-Obi, 203 Ven-Obi), and at relapse for 41 pts (14 Clb-Obi and 27 Ven-Obi). The pre-treatment transcriptomic profile correlated with the IGHV status, trisomy 12, and del13q (B). No clustering according to MRD status was observed, suggesting that individual pathways/genes rather than global differences are associated with MRD response. Individual gene-level analysis of pre-treatment cohorts showed that MRD+ status was associated with resistance markers such as multi-drug response (MDR) gene or ABCB1, but MRD6 was associated with pro-apoptotic BCL2L11 (BIM) (C). Hallmark apoptotic pathways (P53 and Apoptosis), as well as canonical oncogenic pathways (MYC, mTORC1, TNFɑ/NFκB), were enriched in MRD6 pts in both Ven-Obi and Clb-Obi-treated pts (D). In contrast, inflammatory pathway gene sets (Inflammatory response, IFNγ response, IL2/STAT5) were enriched in MRD+ pts in the Ven-Obi arm, suggesting their role in Ven resistance. Gene sets enriched at progression compared to baseline were mostly related to cell proliferation and oncogenic signaling (E). Moreover, in a GSVA analysis, inflammatory pathway gene sets were also markedly upregulated at progression compared to baseline, providing an orthogonal validation that inflammatory signaling might be associated with resistance to venetoclax therapy (Fig F). Common leading edge genes included IL2RB, LCP2, BST2, IRF7, CASP3, CD86, IRF8, and NCOA3. Image: graphic file with name hs9-6-1-g028.jpg [61]Open in a new tab Summary/Conclusion: This analysis of a large, prospective CLL RNAseq dataset confirms that global transcriptomic profiles cluster according to IGHV, trisomy 12 and del13q. Response and non-response to therapy, as assessed by NGS-based MRD status, was associated with a distinct transcriptomic profile that was characterized by upregulated oncogenic pathways and inflammatory signaling, respectively. These data suggest possible biological vulnerabilities that could be leveraged to overcome resistance to venetoclax. S144: IMMUNE RESTORATION AND SYNERGISTIC ACTIVITY WITH FIRST-LINE (1L) IBRUTINIB (IBR) PLUS VENETOCLAX (VEN): TRANSLATIONAL ANALYSES OF CAPTIVATE PATIENTS WITH CLL I. Solman^1,*, R. Singh Mali^2, L. Scarfo^3, M. Choi^4, C. Moreno^5 6, A. Grigg^7, J. P. Dean^1, E. Szafer-Glusman^1 ^1Pharmacyclics LLC, an AbbVie Company, South San Francisco, CA; ^2AbbVie, North Chicago, IL, United States of America; ^3Ospedale San Raffaele, Milan, Italy; ^4University of California San Diego, San Diego, CA, United States of America; ^5Hospital de la Santa Creu I Sant Pau, Autonomous University of Barcelona; ^6Josep Carreras Leukaemia Research Institute, Barcelona, Spain; ^7Austin Hospital, Heidelberg, VIC, Australia Background: Ibr and Ven work synergistically through their distinct and complementary modes of action: BTK inhibition by Ibr decreases levels of anti-apoptotic MCL-1 and BCL-XL proteins, but not BCL-2, and therefore increases sensitization to BCL-2 inhibition by Ven in vivo. Single agent Ibr has been shown to normalize CLL-associated immune cell alterations in number and function but the impact of the Ibr plus Ven combination (I+V) on immune cells has not been evaluated. Aims: To 1) confirm BCL-2 sensitization by single-agent Ibr and 2) monitor changes in the cellular immune profile of patients (pts) treated with I+V in CAPTIVATE ([62]NCT02910583), a multicenter phase 2 study evaluating 1L I+V treatment (tx) in CLL and SLL. Methods: Ibr (420mg po daily) effects on anti-apoptotic proteins were evaluated by flow cytometry with samples from 4 previously untreated pts with CLL treated for 1 cycle (28 days). Immune restoration was evaluated in 79 previously untreated pts with CLL enrolled in the CAPTIVATE Minimal Residual Disease (MRD) cohort. Pts received 3 cycles of Ibr, followed by 12 cycles of I+V (Ibr 420 mg/d orally; Ven ramp-up to 400 mg/d orally). Pts who met criteria for confirmed undetectable MRD (Confirmed uMRD) were then randomized 1:1 to placebo fixed-duration (FD) tx or Ibr at cycle 16; pts who did not meet Confirmed uMRD criteria (uMRD Not Confirmed) were randomized 1:1 to Ibr or I+V. Cryopreserved peripheral blood mononuclear cells at baseline (pre-dose cycle 1) and day 1 of cycles 4, 7, 16, 20, 23, and 29 in each of the 4 arms were analyzed by high-dimensional flow cytometry (34-color panel). Immune cell subset counts of B cells, T cells, monocytes, dendritic cells (DCs), myeloid-derived suppressor cells, natural killer cells, and innate lymphoid cells were calculated and compared to those of 20 age-matched healthy donors. Median changes from baseline are reported. Results: In pts treated with single-agent Ibr for 1 cycle, MCL-1, BCL-XL and BCL-2 expression decreased by 74%, 95% and 10%, respectively, in lymph node emigrant (CD5^+ CXCR4^dim) CLL cells, confirming increased BCL-2-dependence and corresponding inhibitor sensitization. In CAPTIVATE pts treated with I+V, a rapid and significant decrease in circulating CLL cells was observed within the first 3 cycles of Ven tx initiation (Fig 1A, cycle 7). From cycle 16 and onwards, Confirmed uMRD pts had CLL cell counts similar to healthy donors (≤0.8 CLL cell/mL), while in uMRD Not Confirmed pts counts were 1.5 to 22.9 CLL cell/mL. Subsequently, normal B-cell count restoration occurred in Confirmed uMRD pts randomized to placebo (FD tx) with a recovery to levels similar to healthy donors (+332% at cycle 29; Fig 1B). Across all arms, abnormal counts of T cell subsets, classical monocytes and conventional DCs characteristic of CLL were normalized to healthy donor levels (‒49%, +101% and +91% respectively) within the first 6 months of tx and were maintained thereafter regardless of randomized tx. Plasmacytoid DCs counts recovered by cycle 20 (+598%) in all 4 arms. Image: graphic file with name hs9-6-1-g029.jpg [63]Open in a new tab Summary/Conclusion: Sensitization to BCL-2 inhibitor effects by Ibr supports the synergistic activity of I+V. The FD regimen (Confirmed uMRD, placebo arm) of I+V effectively eradicated CLL cells to healthy donor levels and enabled sustained regeneration of normal B cell counts in contrast to ongoing therapy in the remaining tx arms. Combination I+V also allowed for normalization of other critical immune cells, including T cell subsets, classical monocytes, and conventional and plasmacytoid DCs. These data demonstrate promising evidence of immune restoration with FD I+V tx. S145: THE COMBINATION OF IBRUTINIB PLUS VENETOCLAX RESULTS IN A HIGH RATE OF MRD NEGATIVITY IN PREVIOUSLY UNTREATED CLL: THE RESULTS OF THE PLANNED INTERIM ANALYSIS OF THE PHASE III NCRI FLAIR TRIAL P. Hillmen^1,*, A. Pitchford^2, A. Bloor^3, A. Pettitt^4, P. Patten^5, F. Forconi^6, A. Schuh^7, C. Fox^8, N. Elmusharaf^9, S. Gatto^10, B. Kennedy^11, J. Gribben^12, N. Pemberton^13, O. Sheehy^14, G. Preston^15, D. Howard^16, A. Hockaday^2, D. Cairns^2, S. Jackson^2, N. Greatorex^2, N. Webster^17, S. Dalal^17, J. Shingles^17, K. Cwynarski^18, S. Paneesha^19, D. Allsup^20, A. Rawstron^17, T. Munir^21 ^1St James’s University Hospital; ^2Leeds Cancer Research UK Clinical Trials Unit, Leeds Institute of Clinical Trials Research, Leeds; ^3University of Manchester, Manchester; ^4Clatterbridge Cancer Centre NHS Foundation Trust, Liverpool; ^5King’s College Hospital, London; ^6University of Southampton, Southampton; ^7Oxford University Hospital NHT Trust, Oxford; ^8Nottingham University Hospitals NHS Trust, Nottingham; ^9Wales Teaching Hospital; ^10University Hospital of Wales, Cardiff; ^11Leicester Royal Infirmary, Leicester; ^12St Bartholomew’s Hospital, London; ^13Worcestershire Acute Hospitals NHS Trust, Worcester; ^14Belfast City Hospital, Belfast; ^15Aberdeen Royal Infirmary, Aberdeen; ^16Roche, Reading; ^17HMDS, Leeds Cancer Centre, Leeds; ^18University College London, London; ^19Birmingham Heartlands Hospital, Birmingham; ^20Castle Hill Hospital, Hull; ^21UK NCRI CLL sub-group, Leeds, United Kingdom Background: Ibrutinib (I) and venetoclax (V) improve outcome in CLL. I rarely eradicates measurable residual disease (MRD), whereas V (alone or with anti-CD20) can eradicate MRD permitting time limited therapy. Small studies suggest synergy between I and V, as I+V results in MRD negativity in many patients (pts). Aims: The primary aim was to compare the MRD eradication rate between I and I+V. Key secondary aims were IWCLL overall (ORR), complete response (CR) and safety. Methods: FLAIR (ISRCTN01844152), a phase III, randomised, controlled trial for previously untreated CLL requiring therapy by IWCLL criteria. Pts >75 yrs or with >20% 17p del were excluded. FLAIR was adapted in July 2017 to add two arms, I monotherapy and I+V. I was given at 420mg/day. For I+V, V was added after two months of I with dose escalation to 400mg/day over 5 weeks. The duration of therapy (DOT) was defined by MRD with treatment for up to 6 years. MRD was assessed centrally by flow cytometry, MRD negativity was defined as <1 CLL cell in 10,000 leucocytes (IWCLL criteria), was assessed in peripheral blood (PB) and bone marrow (BM) at 9 months post-randomisation, in PB at 12 months and then 6 monthly. When PB was MRD negative, this was repeated after 3 months and then in both PB & BM 3 months later. If PB & BM were negative the time to MRD negativity was calculated (treatment start to first MRD negative PB) and DOT was twice this. The earliest therapy could stop was 2 years post-randomisation. A formal interim analysis was performed when 50% pts in I and I+V arms had reached 2 yrs post-randomisation and a p-value of <0.005 was statistically significant. Results: 523 pts were randomised to I or I+V. We report the interim analysis in the first 274 pts (I [n=138] and I+V [n=136]) reaching 2 yrs post-randomisation from 83 UK Centres from 13/07/17 to 15/03/19. 72.3% male, median age 63 yrs (34.3% >65yo) and 40.9% Binet C. IGHV were available for 256 (93.4%) pts - 48.2% IGHV unmutated (≥98% homology to germline), 45.3% IGHV mutated and 9.1% Subset 2. Hierarchical FISH testing revealed 16.1% 11q del, 19% trisomy 12, 21.9% normal and 36.9% 13q del with 6.2% failed. The arms were well-balanced for all variables. For I+V arm, MRD negativity was achieved within 24 months in BM in 89/136 (65.4%) and PB in 97/136 (71.3%) compared to no pts for I (p<0.0001). MRD negativity for I+V in BM within 24 months was 51/64 (79.7%) for IGHV unmutated and 31/55 (56.4%) for IGHV mutated. At 9 months post-randomisation 49/136 (36%) I+V pts were MRD negative in BM and 56/136 (41.2%) negative in PB compared to 0/138 with I (p<0.0001). ORR at 9 months in 120/136 (88.2%) I+V pts and 119/138 (86.2%) I pts (p=0.6157). At 9 months CR in 81/136 (59.6%) for I+V and 11/138 (8%) for I (p<0.0001). For I+V CR at any time was 93.4%. At 24 months, 54/136 (39.7%) stopped I+V due to meeting MRD stopping criteria. SAEs were reported in 41.5% I+V and 38.2% I pts. Infectious SAEs 14.8% vs 19.9% and cardiac SAEs 11.9% vs 8.1% of pts for I+V & I respectively. Laboratory TLS was reported in 6/136 (4.4%) of I+V pts and none with I. There were no cases of clinical TLS. Most frequent any grade AEs within 12 months of randomisation differing between I+V & I were diarrhoea (52.6% I+V pts, 29.4% I), anaemia (28.9% vs 16.9%), leucopenia (36.3% vs 8.8%), thrombocytopenia (23.7% vs 14%). Leucopenia was the only grade ≥3 SAE reported in >10% of pts (27.4% I+V and 5.1% I). Image: graphic file with name hs9-6-1-g030.jpg [64]Open in a new tab Summary/Conclusion: Ibrutinib plus venetoclax is an effective and well tolerated combination resulting in a high rate of MRD negativity in blood (71.9%) and marrow (65.4%) in the first 2 years of treatment. S146: VENETOCLAX IN PATIENTS WITH CHRONIC LYMPHOCYTIC LEUKEMIA WITH 17P DELETION: 6-YEAR FOLLOW-UP AND GENOMIC ANALYSES IN A PIVOTAL PHASE 2 TRIAL S. Stilgenbauer^1,*, E. Tausch^1, A. W. Roberts^2, M. S. Davids^3, B. Eichhorst^4, M. Hallek^4, P. Hillmen^5, C. Schneider^1, S. Böttcher^6, R. Popovic^7, M. T. Ghanim^7, M. Moran^7, W. J. Sinai^7, X. Wang^7, N. Mukherjee^7, B. Chyla^7, W. G. Wierda^8, J. F. Seymour^2 ^1Division of CLL, Internal Medicine III, Ulm University, Ulm, Germany; ^2Peter MacCallum Cancer Centre, Royal Melbourne Hospital, and University of Melbourne, Melbourne, Australia; ^3Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, United States of America; ^4Department of Internal Medicine, Center of Integrated Oncology Köln Bonn, University Hospital of Cologne, Cologne, Germany; ^5Leeds Teaching Hospitals, NHS Trust, Leeds, United Kingdom; ^6Division of Internal Medicine, Medical Clinic III-Hematology, Oncology and Palliative Medicine, Rostock University Medical Center, Rostock, Germany; ^7AbbVie Inc, North Chicago, IL; ^8The University of Texas MD Anderson Cancer Center, Houston, TX, United States of America Background: Patients (pts) with chronic lymphocytic leukemia (CLL) with del(17p) and/or mutated TP53 have adverse prognosis and are a population of interest for improving outcomes. In a pivotal Phase 2 study ([65]NCT01889186) evaluating venetoclax (Ven; a selective oral BCL2 inhibitor) monotherapy in del(17p) CLL (N=158), overall response rate (ORR) was 77% (median time on study, 26.6 mo; Stilgenbauer. J Clin Oncol. 2018;17:1973). Aims: To report final analyses after 5 y from last pt enrolled, including post-hoc subgroup analyses of prognostic markers. Methods: Adults with R/R or previously untreated del(17p) CLL received Ven 400 mg (via ramp-up) orally daily until progressive disease (PD) or intolerance. Main cohort comprised pts with R/R CLL; safety expansion cohort included pts with R/R or untreated CLL. Primary endpoints were ORR (main) and safety (safety expansion). Peripheral blood (PB) and bone marrow (BM) were analyzed pre-Ven for somatic mutations via targeted next-generation sequencing (NGS). Minimal residual disease (MRD) was assessed by flow cytometry, ASO-PCR, and/or clonoSEQ NGS; undetectable MRD (uMRD) is reported at <10^−4. Results: In all, 158 pts received Ven (main, n=107; safety expansion, n=51). Data cutoff was 15 December 2020. Pts had median 2 (range, 0−10) prior lines of therapy (LOT); 5 had untreated CLL (first-line [1L]). Common mutations (n/N) were TP53 (113/138 [82%]; 35/113 [31%] had >1), SF3B1 (28/137 [20%]) NOTCH1 (22/137 [16%]), ATM (11/120 [9%]), and BIRC3 (6/120 [5%]); 93/115 (81%) had unmutated IGHV and 118/158 (75%) had ≥20% del(17p). Median time on Ven was 27.4 (range, 0–79.3) mo. At study close, 77 pts were alive; 26 remained on Ven. No new safety signals were observed. With 70 mo median follow-up (f/u), investigator-assessed ORR was 77% (95% CI, 70−84); 21% achieved complete remission (CR)/CR with incomplete blood count recovery (CRi). Median duration of response was 39.3 mo (95% CI, 31.1−50.5); 28% had ongoing response at 60 mo. Of 61 evaluable pts with a PB MRD assessment, 25% had uMRD at ~2 y (24−30 mo). Median progression-free survival (mPFS) was 28.2 mo (95% CI, 23.4−37.6; Figure). Median overall survival (mOS) was 62.5 mo (95% CI, 51.7−NR; 5-y PFS and OS rates, 24% and 52%). Of 5 1L pts, 4 were alive and progression-free. In pts with CR/CRi (n=37), mPFS was 62.2 mo (95% CI, 53.1−NR); in pts with partial remission (PR)/nodular PR (n=85), mPFS was 27.6 mo (95% CI, 22.8−38.0). Overall, 98/158 (62%) pts had PD, including 24/158 (15%) with Richter transformation. Of pts with PD (n=98) or whose disease was refractory to Ven (n=11), 73 received another LOT, most commonly ibrutinib (n=41); mOS from ibrutinib initiation was 28.0 mo. In pts with both del(17p) and TP53 mutation (n=111) vs those with either (n=19), ORR was 75% vs 79%; mPFS was 27.4 vs 22.8 mo (P=.8). In pts with mutated (n=22) vs unmutated IGHV (n=93), mPFS was 40.4 vs 26.9 mo (P=.11). No significant difference in ORR, PFS, or OS was observed in pts with vs without ≥20% del(17p), >1 TP53 mutation, NOTCH1, ATM, or BIRC3 mutations; mPFS was shorter in pts with mutated SF3B1 (n=28) vs without (n=109; 16.4 vs 30.2 mo [P=.0071]). Multivariate analysis is ongoing. Image: graphic file with name hs9-6-1-g031.jpg [66]Open in a new tab Summary/Conclusion: At end of study (median f/u, 70 mo), 48% of pts were alive, 24% were progression-free, and 16% remained on Ven, confirming the long-term activity of Ven in this high-risk population with del(17p) CLL and median 2 prior LOT. Except SF3B1 mutation, other adverse features (eg, >1 TP53 mutation, NOTCH1 mutations, unmutated IGHV) did not influence outcomes with Ven treatment in this cohort. S147: PIRTOBRUTINIB, A HIGHLY SELECTIVE, NON-COVALENT (REVERSIBLE) BTK INHIBITOR IN PREVIOUSLY TREATED CLL/SLL: UPDATED RESULTS FROM THE PHASE 1/2 BRUIN STUDY A. R. Mato^1,*, J. M. Pagel^2, C. C. Coombs^3, N. N. Shah^4, N. Lamanna^5, T. Munir^6, E. Lech-Maranda^7, T. A. Eyre^8, J. A. Woyach^9, W. G. Wierda^10, C. Y. Cheah^11, J. B. Cohen^12, L. E. Roeker^1, M. R. Patel^13, B. Fakhri^14, M. A. Barve^15, C. Tam^16, D. Lewis^17, J. N. Gerson^18, A. J. Alencar^19, C. Ujjani^20, I. Flinn^21, S. Sundaram^22, S. Ma^23, D. Jagadeesh^24, J. Rhodes^25, J. Taylor^19, O. Abdel-Wahab^1, P. Ghia^26, S. J. Schuster^18, D. Wang^2, B. Nair^2, E. Zhu^2, D. E. Tsai^2, M. S. Davids^27, J. R. Brown^27, W. Jurczak^28 ^1Memorial Sloan Kettering Cancer Center, New York; ^2Loxo Oncology at Lilly, Stamford; ^3University of North Carolina at Chapel Hill, Chapel Hill; ^4Medical College of Wisconsin, Milwaukee; ^5Herbert Irving Comprehensive Cancer Center, Columbia University, New York, United States of America; ^6Haematology, St. James’s University Hospital, Leeds, United Kingdom; ^7Institute of Hematology and Transfusion Medicine, Warsaw, Poland; ^8Oxford University Hospitals NHS Foundation Trust, Churchill Cancer Center, Oxford, United Kingdom; ^9The Ohio State University Comprehensive Cancer Center, Columbus; ^10MD Anderson Cancer Center, Houston, United States of America; ^11Linear Clinical Research and Sir Charles Gairdner Hospital, Perth, Australia; ^12Winship Cancer Institute, Emory University, Atlanta; ^13Florida Cancer Specialists/Sarah Cannon Research Institute, Sarasota; ^14University of California San Francisco, San Francisco; ^15Mary Crowley Cancer Research, Dallas, United States of America; ^16Peter MacCallum Cancer Center, Royal Melbourne Hospital, and University of Melbourne, Melbourne, Australia; ^17Plymouth Hospitals NHS Trust - Derriford Hospital, Plymouth, United Kingdom; ^18Lymphoma Program, Abramson Cancer Center, University of Pennsylvania, Philadelphia; ^19University of Miami Miller School of Medicine, Miami; ^20Fred Hutchinson Cancer Research Center, Seattle; ^21Sarah Cannon Research Institute, Nashville; ^22Hematology and Medical Oncology, Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York; ^23Robert H. Lurie Comprehensive Cancer Center of Northwestern University, Chicago; ^24Cleveland Clinic, Cleveland; ^25Northwell Health Cancer Institute, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell Health, New Hyde Park, United States of America; ^26Università Vita-Salute San Raffaele and IRCCS Ospedale San Raffaele, Milan, Italy; ^27Dana-Farber Cancer Institute and Harvard Medical School, Boston, United States of America; ^28Maria Sklodowska-Curie National Research Institute of Oncology, Krakow, Poland Background: Covalent BTK inhibitors (BTKi) have transformed the management of chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL), but many patients (pts) will require additional treatment. Pirtobrutinib is a highly selective, non-covalent (reversible) BTKi that inhibits both wild type (WT) and C481-mutated BTK with equal low nM potency. Aims: BRUIN is a phase 1/2 multicenter study ([67]NCT03740529) of oral pirtobrutinib monotherapy in pts with advanced B-cell malignancies who have received ≥2 prior therapies. Methods: Pirtobrutinib was dose escalated in a standard 3 + 3 design in 28-day cycles. The primary objective for phase 1 was to determine the recommended phase 2 dose (RP2D) and the primary objective of phase 2 was overall response rate (ORR); secondary objectives included duration of response, progression-free survival, overall survival, safety and tolerability and pharmacokinetics. Response was assessed every 8 weeks from cycle 3, and every 12 weeks from cycle 13 and was measured according to the iwCLL 2018 criteria, including PR with lymphocytosis (PR-L). Safety was assessed in all pts. Results: As of 27 September 2020, 323 pts with B-cell malignancies (170 CLL/SLL, 61 MCL, 26 WM, 26 DLBCL, 13 MZL, 12 FL, 9 RT, and 6 other) were treated on 7 dose levels (25-300mg QD). Among the 170 pts with CLL/SLL, the median age was 69 (36-88) years. Median number of prior lines of therapies was 3 (1-11). Majority of the CLL/SLL pts had received a prior BTKi (86%), an anti-CD20 antibody (90%), or a chemotherapy (82%). High risk molecular features such as 17p deletion, TP53 mutation, and unmutated IGHV were present in 25% (20/81), 30% (27/91), and 88% (71/81) of pts, respectively. No dose-limiting toxicities (DLTs) were reported and maximum tolerated dose (MTD) was not reached (n=323). 200mg QD was selected as the RP2D. Fatigue (20%), diarrhea (17%) and contusion (13%) were the most frequent treatment-emergent adverse events (TEAEs) regardless of attribution or grade seen in ≥10% of pts (n=323). The most common adverse event of grade ≥3 was neutropenia (10%). Treatment-related hemorrhage and hypertension occurred in 5 (2%) and 4 (1%) pts, respectively. 1% pts discontinued due to TEAEs. 139 CLL/SLL pts were efficacy-evaluable with a median follow up time of 6 months (0.16-17.8+). The ORR was 63% (95% CI 55-71) among the 139 efficacy evaluable pts with 69 PRs (50%), 19 PR-Ls (14%), 45 SDs (32%), and 1 PD (1%), and 5 (4%) discontinued prior to first response assessment. Among the 121 BTKi pretreated pts, the ORR was 62% (95% CI 53-71). Responses deepened over time with an ORR of 86% among the pts with at least 10 months follow-up. ORR was similar in pts who discontinued prior BTKi due to progression (67%), or adverse events or other reasons (52%). Of the 88 responding pts, all except 5 remained on therapy (4 progressed and 1 achieved a PR and electively discontinued treatment to undergo transplant). The longest-followed responding patient had been on treatment for 17.8+ months. Summary/Conclusion: Pirtobrutinib demonstrated promising efficacy in heavily pretreated CLL/SLL pts following multiple prior lines of therapy and in pts with BTK C481 mutations. Pirtobrutinib was well tolerated and exhibited a wide therapeutic index. Updated data, including approximately 100 new pts with CLL and an additional 10 months since the prior data cut will be presented. S148: VENETOCLAX-OBINUTUZUMAB FOR PREVIOUSLY UNTREATED CHRONIC LYMPHOCYTIC LEUKEMIA: 5-YEAR RESULTS OF THE RANDOMIZED CLL14 STUDY O. Al-Sawaf^1 2 3,*, C. Zhang^1, S. Robrecht^1, A. Kotak^4, N. Chang^4, A.-M. Fink^1, E. Tausch^5, C. Schneider^5, M. Ritgen^6, K.-A. Kreuzer^1, B. Chyla^7, B. Eichhorst^1, Y. Jiang^8, S. Stilgenbauer^5, M. Hallek^1, K. Fischer^1 ^1Department I of Internal Medicine and Center of Integrated Oncology Aachen Cologne Bonn Duesseldorf, University Hospital of Cologne, Cologne, Germany; ^2Cancer Institute, University College London; ^3Francis Crick Institute, London; ^4Roche Products Ltd, Welwyn Garden City, United Kingdom; ^5Department III of Internal Medicine, Ulm University, Ulm; ^6Department II of Internal Medicine, University of Schleswig Holstein, Kiel, Germany; ^7AbbVie Inc., North Chicago, IL; ^8Genentech Inc., South San Francisco, CA, United States of America Background: One-year fixed-duration venetoclax-obinutuzumab (Ven-Obi) has demonstrated significant improvement of progression-free survival (PFS) as compared to chlorambucil-obinutuzumab (Clb-Obi) in patients with previously untreated chronic lymphocytic leukemia (CLL) and coexisting conditions in the CLL14 trial. As high rates of undetectable minimal residual disease (uMRD) suggested deep remissions, long-term efficacy data including patients with high-risk disease is of particular interest. Aims: The aim of this report is to provide updated efficacy and safety data from the ongoing follow-up of the CLL14 study with all patients being off study treatment for ≥ 4 years. Methods: Patients with previously untreated CLL and coexisting conditions were randomized 1:1 to 12 cycles of venetoclax with 6 cycles of obinutuzumab or 12 cycles of chlorambucil with 6 cycles of obinutuzumab. The primary endpoint was investigator-assessed PFS. Secondary endpoints included safety, rates of MRD response (measured every 3-6 months up to 9 years after last patient enrolment), time to next treatment (TTNT) and overall survival. Follow-up is ongoing, all patients are off study treatment. Results: Of the 432 enrolled patients, 216 were randomly assigned to receive Ven-Obi and 216 to receive Clb-Obi. With a current median follow-up of 65.4 months (interquartile range 52.6-69.4), PFS remained significantly superior for Ven-Obi compared to Clb-Obi (median not reached [nr] vs 36.4 months; hazard ratio [HR] 0.35 [95% CI 0.26-0.46], p<0.0001). At 5 years after randomization, the estimated PFS rate was 62.6% after Ven-Obi and 27.0% after Clb-Obi. Overall, 52 cases of progressive disease (PD) with 28 required second-line treatments occurred in the Ven-Obi arm and 132 with 86 second-line treatments in the Clb-Obi arm. TTNT was significantly longer after Ven-Obi (5-year TTNT 72.1% vs 42.8%; HR 0.42, 95% CI 0.31-0.57, p<0.0001). In both arms, the majority of next-line therapies were BTK inhibitors (54.3% in the Ven-Obi arm, 47.1% in the Clb-Obi arm). The PFS and TTNT difference was maintained across all risk groups, including patients with TP53 mutation/deletion (5-year PFS 40.6% vs 15.6%; 5-year TTNT 48.0% vs 20.8%) and unmutated IGHV status (5-year PFS 55.8% vs 12.5%; 5-year TTNT 66.2% vs 25.1%). A multivariable analysis indicated 17p deletion and high disease burden as independent prognostic factors for PFS in patients treated with Ven-Obi. Four years after treatment completion, 39 (18.1% of the intention-to-treat population) patients in the Ven-Obi arm still had uMRD (<10^-4 by NGS in peripheral blood), 27 (12.5%) had low (L)-MRD (≥ 10^-4 and < 10^-2) and 41 (19.0%) high (H)-MRD (≥ 10^-2), compared to 4 (1.9%) uMRD, 13 (6.0%) L-MRD and 24 (11.1%) H-MRD in the Clb-Obi arm. Overall, 40 deaths were reported in the Ven-Obi arm (8 PD related) and 57 in the Clb-Obi arm (23 PD related); at 5 years after randomization the estimated OS rate was 81.9% in the Ven-Obi arm and 77.0% in the Clb-Obi arm (HR 0.72 [0.48-1.09], p=0.12). Second primary malignancies were reported in 44 (20.8%) patients in the Ven-Obi arm and 32 (15.0%) in the Clb-Obi arm. No new safety signals were observed. Image: graphic file with name hs9-6-1-g032.jpg [68]Open in a new tab Summary/Conclusion: These data confirm that over 60% of patients who had received 1-year fixed-duration Ven-Obi have remained in remission four years after end of therapy. The majority of patients treated with Ven-Obi still have not required a second line of CLL therapy. Hence, the 1-year Ven-Obi regimen continues to be an effective fixed-duration option for patients with CLL and coexisting conditions, also in the context of high-risk disease. S149: LONG TERM OUTCOMES OF IFCG REGIMEN FOR FIRSTLINE TREATMENT OF PATIENTS WITH CLL WITH MUTATED IGHV AND WITHOUT DEL(17P)/TP53 MUTATION N. Jain^1,*, P. Thompson^1, J. Burger^1, A. Ferrajoli^1, K. Takahashi^1, Z. Estrov^1, G. Borthakur^1, P. Bose^1, T. Kadia^1, N. Pemmaraju^1, K. Sasaki^1, M. Konopleva^1, E. Jabbour^1, N. Garg^2, X. Wang^3, R. Kanagal-Shamanna^4, K. Patel^4, W. Wang^4, S. Wang^4, J. Jorgensen^4, W. Lopez^1, A. Ayala^1, W. Plunkett^5, V. Gandhi^5, H. Kantarjian^1, S. O’Brien^6, M. Keating^1, W. Wierda^1 ^1Leukemia; ^2Radiology; ^3Biostatistics; ^4Hematopathology; ^5Experimental Therapeutics, The University of Texas MD Anderson Cancer Center, Houston; ^6Chao Family Comprehensive Cancer Center, University of California Irvine Medical Center, Orange, United States of America Background: Chemoimmunotherapy with FCR has been an effective treatment for patients (pts) with CLL. Pts with mutated IGHV (IGHV-M) have favorable long-term outcomes after receiving FCR. We designed an investigator-initiated, phase 2 trial with ibrutinib, fludarabine, cyclophosphamide, and obinutuzumab (iFCG) for previously untreated pts with IGHV-M CLL ([69]NCT02629809). We report here long-term outcomes with a median follow-up of 56.8 months. Aims: Investigate the role of combined chemoimmunotherapy and targeted therapy in CLL. Methods: Eligibility included age ≥18, IGHV-M, no del(17p)/TP53 mutation. Pts received 3 courses of iFCG. Pts achieving CR/CRi with undetectable MRD (U-MRD) in bone marrow (4-color flow-cytometry, sensitivity 10^-4) after 3 courses of iFCG received ibrutinib with obinutuzumab (iG) for 3 cycles, followed by ibrutinib monotherapy for 6 months. All other pts received iG for 9 cycles (C4-12). Pts with marrow U-MRD at end of Cycle 12 stop all therapy. Response assessment was per 2008 iwCLL criteria with BM and CT scans every 3 months during the first year. After completion of all therapy, pts were followed by exam, blood counts and peripheral blood MRD every 6 months. NGS MRD (sensitivity 10^-6) was performed in bone marrow in pts with available samples. Results: 45 pts initiated treatment. Median age was 60 [range, 25-71]. 69% had del(13q). After three cycles of iFCG, 39/45 (87%) pts achieved marrow U-MRD. Responses improved with continued therapy with 40/45 (89%) and 41/45 (91%) achieving marrow U-MRD after Cycles 6 and 12, respectively. Overall, 44/45 (98%) pts achieved marrow U-MRD as best response at any time during the study. The 5-year PFS and OS are 97.7% (95% CI 94–100%) and 97.8% (95% CI 94–100%), respectively. No pt had CLL progression or Richter transformation. One pt developed therapy-related MDS; this pt is being monitored for 38+ months without any therapy for MDS with normal blood counts. The sole event noted on both the PFS and OS curve is a pt death from heart failure. 41/45 pts completed 12 cycles of treatment (4 pts came off study prior to C12). All 41 pts achieved marrow U-MRD4 by flow-cytometry and per protocol, all 41 pts discontinued ibrutinib. After a median follow-up of 44.2 mos post-discontinuing ibrutinib, 6 pts had an MRD recurrence (defined as 2 consecutive values of ≥0.01% in peripheral blood by flow cytometry) at a median of 27.2 mos (range, 20.7-49.0 mos) after stopping all therapy. All 6 pts are being monitored with no clinical progression or active therapy; notably, all 6 pts were MRD+ at 10^-6 by marrow NGS after the completion of 3 cycles of iFCG and 4/5 (1 missed sample) were MRD+ at 10^-6 by marrow NGS after the completion of Cycle 12. Of the 16 pts who were marrow NGS MRD+ at 10^-6 after 3 cycles of iFCG, 6/16 had an MRD recurrence in blood in follow-up vs. 0/20 who were NGS MRD negative/indeterminate (p = 0.004). Of the 12 pts who were marrow NGS MRD+ at 10^-6 after Cycle 12, 4/12 had an MRD recurrence in blood in follow-up vs. 1/26 who were NGS MRD negative/indeterminate (p = 0.02). Image: graphic file with name hs9-6-1-g033.jpg [70]Open in a new tab Summary/Conclusion: The iFCG regimen, using only 3 cycles of chemotherapy (as opposed to 6 cycles of chemoimmunotherapy) achieves a very high rate of U-MRD in previously-untreated pts with CLL with IGHV-M CLL. No pt had disease progression with a median follow-up of close to 5 years. The 5-year PFS is 97.7%; this is favorable compared to 5-year PFS of ≈65% with FCR (CLL10), ≈70% with ibrutinib (A041202 trial), and 81% with ibrutinib (RESONATE-2) for IGHV-M CLL. Not unexpectedly, MRD recurrence during follow-up correlated with MRD positivity by NGS during therapy. S150: BIOLOGY AND FUNCTION OF CIRCULAR PCMDT1 IN CHRONIC MYELOID LEUKEMIA IN BLAST CRISIS (CML-BC) A. Prathap Urs^1,*, D. Papaioannou^2, R. Buisson^3, S. Khanal^1, M. Karunasiri^1, S. Kauppinen^4, A. Dorrance^1 5, R. Garzon^1 5 ^1Comprehensive Cancer Center, The Ohio State University, Columbus; ^2NYU Langone Health, Laura and Isaac Perlmutter Cancer Center, New York; ^3Department of Biological Chemistry, University of California, California, United States of America; ^4Department of Clinical Medicine, Aalborg University, Copenhagen, Denmark; ^5Department of Internal Medicine, The Ohio State University, Columbus, United States of America Background: The prognostic and biologic significance of circular RNAs (circRNAs) in patients with acute myeloid leukemia (AML) has been reported. Circular PCMTD1 (cPCMTD1) was among the circRNAs that were prognostic in this disease. Aims: To study the functional role of cPCMTD1 in leukemias. Methods: We used RNase H-recruiting, locked nucleic acid-modified oligonucleotides (gapmers) for knock-down (KD) experiments. Cell cycle analyses were performed with bromodeoxyuridine and 7-actinomycin D staining. Chloro-deoxyuridine (CldU) and iodo-deoxyuridine (IdU) were used for DNA fiber assays. Results: To study whether cPCMTD1 is important for leukemia, we KD-cPCMTD1 using gapmers that specifically degraded the circular transcript without affecting the levels of the linear PCMTD1. Among the 8 AML cell lines that were tested, we found that depletion of the cPCMTD1 led to a decrease in the proliferating fraction and potent G2/M blockade of only K-562 and LAMA-84 CML-BC cells. In contrast, linear PCMTD1-KD did not have any effect on cell cycle. Mass cytometry (CyTOF) experiments validated the cell cycle blockade after cPCMTD1-KD and identified an increase of H2AX phosphorylation. We validated this finding with western blots and could also show that cPCMDT1-KD led to an increase in the phosphorylation of the ATM, ATR, CHK1, DNA-PK and RPA32 proteins. Further, DNA fiber assays detected a global decrease in the length of CldU- and IdU-labeled fibers as well as of the IdU/CldU ratio in the cPCMTD1-KD cells compared to controls. These results were confirmed by COMET assays. Taken together these data indicate that cPCMTD1-KD causes impaired DNA replication in CML-BC. Then, we examined whether we found in CML-BC cell lines was also true for patients. Thus, we measured cPCMDT1 expression in a panel of CML patients in chronic phase (n=15), accelerated phase (n=4) and blast crises (n=7) and we found a significant increase of cPCDMT1 in CML-BC with respect to chronic and accelerated phases. Next, we performed cPCMDT1-KD in CML-BC patient samples and found increase of H2AX phosphorylation. Like the cell line data, cPCMDT1-KD in 3 samples from AML patients had no functional effect. Finally, after confirming that cPCMDT1 is most abundant in the cytoplasm, in silico analysis and polysome profiling experiments indicated that cPCMDT1 encodes for a small protein. We generated a custom antibody that detected the predicted protein of 30kD, which was depleted after cPCMDT1-KD. Immunoprecipitation experiments followed by mass spectrometry analysis using our custom antibody showed a strong and specific enrichment of the BTR complex (BLM, TOP3A and RMI1), which is implicated in DNA replication and repair. cPCMDT1-KD had no effect on the total amount of each of the three proteins but reduced their interaction and the amount of the formed BTR complex. To validate the therapeutic potential of this strategy we developed a CML-BC PDX model by transplanting a CML-BC patient sample into NSGS mice. This PDX is very aggressive, and mice die from disease by 20 weeks. We are currently treating these mice in vivo with lipid nanoparticle anti-cPCMDT1 gapmers. Results will be updated in the meeting. Summary/Conclusion: cPCMDT1 is a circRNA with protein-coding potential that is over-expressed in CML-BC. cPCDMT1 is critical for the proliferation of CML-BC through the regulation of DNA replication and may represent a novel target for therapy. S151: HIGH-THROUGHPUT EVALUATION OF THE POTENTIAL OF CANCER DRUGS TO ENHANCE NATURAL KILLER CELL IMMUNOTHERAPY IN CHRONIC MYELOID LEUKEMIA. P. Nygren^1 2,*, J. Bouhlal^1 2, E. Laajala^1 2, A. Ianevski^3, J. Klievink^1 2, H. Lähteenmäki^1 2, K. Saeed^1 2, D. Lee^4, T. Aittokallio^3 5 6, O. Dufva^1 2, S. Mustjoki^2 7 ^1Hematology Research Unit Helsinki, University of Helsinki and Helsinki University Hospital; ^2Translational Immunology Research Program; ^3Institute for Molecular Medicine Finland, FIMM, HiLIFE, University of Helsinki, Helsinki, Finland; ^4Division of Hematology, Oncology, and BMT, Nationwide Children’s Hospital, Columbus, United States of America; ^5Institute for Cancer Research, Department of Cancer Genetics, Oslo University Hospital; ^6Centre for Biostatistics and Epidemiology (OCBE), Faculty of Medicine, University of Oslo, Oslo, Norway; ^7Hematology Research Unit Helsinki, University of Helsinki and Helsinki University Hospital, Helsinki, Helsinki, Finland Background: Natural killer (NK) cell immunotherapies are promising novel cancer treatments and complete remission has been achieved in patients with relapsed/refractory myeloid leukemias. The significance of NK cells in chronic myeloid leukemia (CML) patients has been highlighted in several studies. CML patients with a higher percentage of mature NK cells have increased relapse-free survival after treatment discontinuation. NK KIR receptor profiles have also been associated with the achievement of remission in response to tyrosine kinase inhibitor (TKI) therapy. Moreover, complex interactions between TKIs and NK cell function have been discovered in CML. For example, dasatinib and imatinib have been suggested to enhance NK cell cytotoxicity through regulation of activating and inhibitory receptors. NK cell responses are usually short-lived and attempts have been made to enhance their function against malignant cells using cytokines, bi- and tri-specific antibodies, and small molecule inhibitors. However, a large-scale drug screening evaluating the combinatorial effects of NK cells and cancer drugs has not been previously conducted. Aims: To evaluate the potential of small molecule cancer drugs to synergize with NK cell immunotherapy in CML and to discover which drugs can enhance or inhibit cytotoxicity in CML cells. Methods: A high-throughput drug sensitivity and resistance testing (DSRT) screen was used to evaluate the effect of 528 investigational and approved cancer drugs on the cytotoxicity of NK cells. CML cells (K562) expressing luciferase and primary expanded NK cells were co-cultured for 24 hours on 384 well drug plates. Target cell viability was measured using a luciferase-based readout, and results were run through a custom pipeline to calculate differential drug sensitivity scores (dDSS) between co-cultured cells and target cells alone, for each drug. Most promising drugs with highest and lowest dDSS were selected for further analysis with single-cell RNA sequencing to determine their mechanism of action. Results: Out of 528 screened drugs, 161 had an inhibitory effect on cytotoxicity (dDSS < -2.5), 325 had no effect, and 42 had an enhancing effect (dDSS > 2.5). The SMAC mimetics birinapant and LCL-161 were the most potent enhancers of NK cell cytotoxicity and had no effect on target cells alone. Individual drugs with other mechanisms, such as PKC activators, were also effective enhancers of NK cytotoxicity. On the contrary, cytotoxicity-inhibiting drugs included previously known immunosuppressive drugs dexamethasone and prednisolone, as well as novel candidates, such as sotrastaurin, a PKC inhibitor. Single-cell RNA sequencing of drug-treated co-cultures revealed the transcriptomic phenotypes of NK cell and target cells under SMAC mimetic treatment. NF-kB target genes such as BIRC3, NFKB2, MHC II genes, and the T/NK cell-recruiting chemokines CXCL9, CXCL10, and CXCL11 were activated in target cells exposed to both birinapant and NK cells, compared to target cells exposed to NK cells alone (p value adj < 0.05). These data suggest that SMAC mimetics in combination with NK cells may induce an immune-inflamed phenotype in addition to sensitizing CML cells to NK cell cytotoxicity. Image: graphic file with name hs9-6-1-g034.jpg [71]Open in a new tab Summary/Conclusion: We discovered novel drug classes, which had both activating and inhibitory effects on the NK cell cytotoxicity against CML cells in vitro. Defining NK cell - drug - CML cell interactions in a high-throughput setting provides a framework for future combination immunotherapies to further improve treatment-free remission in CML. S152: SETD2/H3K36ME3 DEFICIENCY SUSTAINS GENOMIC INSTABILITY AND ENHANCES CLONOGENIC POTENTIAL OF CHRONIC MYELOID LEUKEMIA (CML) PROGENITORS M. Mancini^1,*, S. De Santis^2, C. Monaldi^2, S. Bruno^2, F. Castagnetti^2, G. Gugliotta^1, A. Iurlo^3, M. Cerrano^4, S. Galimberti^5, S. Balducci^5, F. Stagno^6, G. Rosti^7, M. Cavo^8, S. Soverini^2 ^1IRCCS Azienda Ospedaliero-Universitaria di Bologna, Istituto di Ematologia “Seràgnoli”; ^2Dipartimento di Medicina Specialistica, Diagnostica e Sperimentale, Università di Bologna, Bologna; ^3UO Onco-ematologia, Fondazione IRCCS Ca’ Granda - Ospedale Policlinico, Milano; ^4A.O.U. Citta della Salute e della Scienza di Torino, Torino; ^5Clinical and Experimental Medicine, Hematology, University of Pisa, Pisa; ^6Ematologia Universitaria e Trapianto Midollo Osseo, Ospedale Gaspare Rodolico, Catania; ^7Istituto Romagnolo per lo Studio dei Tumori “Dino Amadori” - IRST Srl Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS), Meldola (FC); ^8IRCCS Azienda Ospedaliero-Universitaria di Bologna, Istituto di Ematologia “Seràgnoli”; Dipartimento di Medicina Specialistica, Diagnostica e Sperimentale, Università di Bologna, Bologna, Italy Background: Genomic instability is a hallmark of chronic myeloid leukemia (CML) cells since the chronic phase (CP) of the disease and results in BCR-ABL1 mutations and/or additional genetic and genomic aberrations that may drive resistance to tyrosine kinase inhibitors (TKIs) and progression to blast crisis (BC). Genomic instability is also a feature of CML stem cells and may underlie their persistence. We have recently reported that virtually all CML patients (pts) in BC display loss of function of SETD2, an enzyme that trimethylates histone H3 lysine 36 (H3K36me3). SETD2 is a tumor suppressor implicated in several neoplastic conditions. H3K36me3 is required for homologous recombination (HR) repair of double strand breaks and for DNA mismatch repair (MMR). Aims: We investigated SETD2/H3K36me3 status in CD34+ progenitors of CP CML pts and whether SETD2/H3K36me3 deficiency may play a role in genomic instability in CML models. Methods: CD34+ progenitor cells were isolated from 20 newly diagnosed CP CML pts and screened for SETD2 and H3K36me3 by Western blotting (WB). SETD2-proficient (LAMA84) and -deficient (KCL22) CML cell lines were also studied. SETD2 knock-down and SETD2 forced expression in CD34+ progenitors and cell lines were performed by RNAi and nucleofection, respectively. DNA damage and DNA repair activation were assessed by WB and immunofluorescence (IF). Clonogenic capacity was evaluated by clonogenic assays. Results: Loss of SETD2 protein expression and function (the latter assessed using loss of H3K36me3 as surrogate marker) were detected by WB in the CD34+ cell fraction of 20/20 newly diagnosed CP CML pts, as compared to the corresponding total mononuclear cell fraction or to the CD34+ fraction obtained from a pool of healthy donors. To investigate whether SETD2/H3K36me3 loss impinges on the activation and proficiency of HR, we used UV rays to induce DNA damage in SETD2 siRNA-depleted LAMA 84 (SETD2-proficient) cells. Compared to control cells, cells silenced for SETD2 displayed an increase in the expression of the DNA damage maker γH2AX associated with a loss of RAD51 (HR) and MSH6 (MMR) repair foci. To confirm the role of SETD2 as a tumor suppressor implicated in maintaining genomic stability in CML, we transfected KCL22 (SETD2-deficient) cells with an ectopic SETD2 plasmid. SETD2 forced expression induced more than 50% reduction in cell doubling time and a significant reduction in clonogenic potential. Moreover, SETD2 overexpression was able to restore DNA damage response, as demonstrated by WB and immunofluorescence detection of H2AX phosphorylation, RAD51 (HR) foci, THEX1 (DNA replication) and MSH6 (MMR) observed after UV exposure. In line with the effects of SETD2 deficiency/overexpression observed in cell line models, CML progenitor cells displaying SETD2/H3K36me3 deficiency showed hyper-phosphorylation of H2AX and the overexpression of RAD51 (HR) and MSH6 (MMR) did not result in DNA repair foci. Moreover, forced overexpression of SETD2 restored proliferation control in CD34+ cells from 5 CP CML pts, since after nucleofection clonogenic assays showed a >50% reduction in clonogenic potential. Summary/Conclusion: SETD2/H3K36me3 deficiency is a novel, BCR-ABL1-independent mechanism of genetic instability in CML. SETD2 behaves like a true tumor suppressor in CD34+ progenitor cells of CP CML pts. Supported by AIRC IG 2019 grant (23001) and Italian Ministry of Health, “Bando Ricerca Finalizzata 2016”, project GR-2016-02364880. S153: DECRYPTING THE ROLE OF HSP90Α AND Β ISOFORMS TO OVERCOME RESISTANCE IN BCR-ABL1 LEUKEMIA M. Vogt^1 2,*, N. Dienstbier^1, J. Schliehe-Diecks^1, K. Scharov^1, J.-W. Tu^1, P. Gebing^1, J. Hogenkamp^1, D. Picard^1, T. Lenz^3, K. Stühler^3 4, R. Wagener^1 5, A. Pandyra^1 5, J. Hauer^6, U. Fischer^1 5, A. Borkhardt^1 5, S. Bhatia^1 5 ^1Department of Pediatric Oncology, Hematology and Clinical Immunology, Medical Faculty, Heinrich Heine University Düsseldorf, University Hospital; ^2DSO - Düsseldorf school of oncology; ^3Molecular Proteomics Laboratory, Biological-Medical Research Center, Heinrich Heine University Düsseldorf; ^4Institute of Molecular Medicine, Proteome Research, Medical Faculty and University Hospital, Heinrich Heine University Düsseldorf; ^5German Cancer Consortium (DKTK), partner site Essen/Düsseldorf, Düsseldorf; ^6Department of Pediatrics, Technical University of Munich, School of Medicine, Munich, Germany Background: The BCR-ABL1 fusion protein, which results from the chromosomal translocation t(9;22)(q34;q11) is detected in over 95% of CML, 25% of adult BCP-ALL and 3-5% of childhood BCP-ALL patients. The development of tyrosine kinase inhibitors (TKi) has enabled targeted treatment, however development of resistance is common and the BCR-ABL1 leukemia subtype is still associated with poor prognosis. Among the chaperone proteins involved in maintaining the homeostasis of cancerous cells, heat shock protein 90 (HSP90) have been widely studied, due to its crucial involvement in stabilizing a variety of oncogenic proteins, including BCR-ABL1 and STAT3/5. Therefore, inhibition of HSP90 can be an effective alternative approach to target TKi resistant and BCR-ABL1^T315I mutant expressing cells. In the past, over 17 HSP90 inhibitors (HSP90i) have been evaluated in clinical trials, however associated resistances (e.g. HSR) and dose limited toxicity have thus far precluded their clinical approval. Aims: In mammalian cells, there are two predominant cytosolic isoforms of HSP90, a stress-inducible HSP90α and a constitutively expressed HSP90β isoform. In the past, the distinct roles of the isoforms were studied by applying pan- and isoform-targeted HSP90i. We aimed to decrypt the role of cytosolic HSP90 isoforms using genetically edited models and to identify novel therapeutic vulnerabilities to target therapy refractory BCR-ABL1+ leukemia. Methods: To study the kinetic implications of the resistance mechanism associated with pharmacological targeting of HSP90, we generated a transient knockdown (Kd) and a stable CRISPR-Cas9 based knockout (KO) model of HSP90α/β in BCR-ABL1+ cells. The edited cells were analyzed in ex vivo functional assays (e.g. IF staining and WB) and their transplantation efficiency was determined in NSG mice. Global transcriptomic and proteomic profiling was performed in conjunction with high throughput drug screening (HTDS) to identify therapeutic vulnerabilities upon loss of HSP90α/β. Results: HSP90α/β-KO or -Kd did not affect the expression of other HSP90 paralogues (TRAP1 and GRP94). Only HSP90β-KO cells displayed hyperactive pro-survival HSR. Other reported HSP90 isoform-dependent clients, such as Survivin, CDK4 and CDK6 were also not affected upon KO or Kd of HSP90α/β. As HSP90 is involved in stabilizing, proper folding and subcellular localization of BCR-ABL1 protein, immunofluorescence imaging demonstrated two fold higher abundance of BCR-ABL1 foci (cytoplasmic/nucleocytoplasmic region) in HSP90α KO cells, with hyperactive BCR-ABL1 and downstream signaling (pBCR-ABL and pSTAT5a). When KO cells were transplanted into NSG mice (n=5 mice per arm) the engraftment of the HSP90α-KO cells was significantly reduced with a prolonged overall survival of the animals (19 days, p=0.0083) as compared to HSP90β-KO and control group. To find possible target (s) in order to circumvent resistance associated with the use HSP90i, we performed HTDS with HSP90α/β KO cells, which revealed distinct sensitivities toward certain classes of inhibitors. One of these classes was CDK7i, which was differentially active in HSP90α KO cells. These results can be explained by our global transcriptomic results, which revealed hyperactive androgen receptor signaling in HSP90α KO cells (GSEA), which is known to be a prognostic marker for CDK7 inhibition. Summary/Conclusion: In vivo targeting of HSP90α is a promising approach to target BCR-ABL1+ leukemia cells. Moreover, the combination with CDK7i may help to overcome the resistance associated with clinical use of pan HSP90i. S154: IMATINIB-RESISTANT CLONES ISOLATED FROM A MODEL OF BLAST CRISIS OF CHRONIC MYELOID LEUKAEMIA DIFFER IN MUTATIONS IN BCR::ABL1 AND OTHER CANCER RELATED GENES AND IN THEIR SENSITIVITY TO BH3-MIMETICS A. Laznicka^1 2,*, N. Curik^1 3, V. Polivkova^1, J. Koblihova^1, P. Burda^1 3, A. Dolnikova^3, E. Pokorna^3, C. Salek^1, H. Klamova^1, D. Srbova^1, P. Klener^3 4, K. Machova Polakova^1 3 ^1Institute of Hematology and Blood Transfusion; ^2Second Faculty of Medicine, Charles University; ^3Institute of Pathological Physiology, First Faculty of Medicine, Charles University; ^41st Dep. of Medicine – Dep. of Hematology, First Faculty of Medicine, Charles University, and General University Hospital, Prague, Czechia Background: The treatment of blast crisis of chronic myeloid leukaemia (BC-CML) remains difficult and the outcome of patients is poor. A dysregulated apoptosis leads to survival of malignant cells. BH3-mimetics inhibit antiapoptotic proteins of BCL-2 family. These drugs have been recently introduced in the treatment of chronic lymphocytic leukaemia and acute myeloid leukaemia. This work assumed that these drugs may have a treatment potential for myeloid and/or lymphoid BC-CML. Aims: The aim is to test the sensitivity of imatinib-resistant (IR) clones of KCL-22, model of BC-CML, and primary blast cells to BH3-mimetics and describe a mechanism of sensitivity/resistance using protein analysis of apoptotic pathways. Methods: Isolated IR-clones (n=10) of KCL-22 were characterized by NGS and cultivated with 4µM imatinib (IM) and dilution series of BH3-mimetics (venetoclax – anti-BCL-2, [72]S63845 – anti-MCL-1, A-1155463 - anti-BCL-XL) for 72 hours. IC[50] was determined based on proliferation. Protein expression of BCL-2 family (n=10) was conducted in 4 IR-clones cultivated with IM and with/without BH3-mimetics. For in vivo experiments NOD-SCID-gamma mice (n=16) were subcutaneously injected with 10^6 cells of BCR::ABL1-T315I clone and divided into control group and treated groups. Primary cells of patients with BC-CML (n=4) were exposed to IM and BH3-mimetics for 7 days and LC[50] was estimated. Results: IR-clones of KCL-22 differ in their sensitivity to BH3-mimetics (Table 1). The majority of clones (8/10) including 4 clones with T315I were sensitive to MCL-1 inhibitor. A reduced sensitivity was observed in the T315I clone carrying mutated GATA2 and other clone with mutation Y253H and mutated BCOR. Moreover, two T315I clones were sensitive to venetoclax, but other T315I clones with mutations in other cancer related genes showed insensitivity (n=2) or decreased sensitivity (n=1). Interestingly, two T315I clones with mutations in other cancer related genes insensitive to venetoclax were sensitive to BCL-XL inhibitor and in general, it seems that clones insensitive to venetoclax were sensitive to BCL-XL inhibitor and vice versa. Protein analysis of parental KCL-22IR cells and 4 clones showed a high expression of BCL-2 and BCL-XL and undetectable MCL-1 expression. Except for the T315I clone (B8, Table 1), other analysed clones expressed MCL-1 after exposure to BH3-mimetics. The sensitivity of T315I clone to anti-MCL-1 can be explained by detection of cBAX (an active form of apoptotic effector) after exposure to the inhibitor. The parental cell line showed dissociation of antiapoptotic complex MCL-1/BIM after exposure to MCL-1 inhibitor. The released BIM accelerates the apoptosis as an apoptotic activator. In vivo analysis revealed a decreased tumour growth of xenografted T315I clone (B8) in IM, venetoclax or combined treatment compared to control. An additive effect of dual therapy against monotherapy was not observed. This can be explained by IM inhibition of non-mutated BCR::ABL1 in KCL-22IR carrying 2 Ph chromosomes. LC[50] analysis of primary blasts (4 patients) showed anti-MCL-1 to be the most potent inhibitor. Mutations in other cancer related genes and/or changes in karyotype increased LC[50] for venetoclax and BCL-XL inhibitor or venetoclax, respectively. Image: graphic file with name hs9-6-1-g035.jpg [73]Open in a new tab Summary/Conclusion: The preliminary data of this study showed that the MCL-1 inhibitor [74]S63845 seems to be the most potent BH3-mimetic to induce apoptosis in BC-CML. The combined therapy of TKI and BH3-mimetics should reflect mutation status of BCR::ABL1 and other cancer related genes. S155: EFFICACY AND SAFETY RESULTS FROM ASCEMBL, A PHASE 3 STUDY OF ASCIMINIB VS BOSUTINIB IN PATIENTS WITH CHRONIC MYELOID LEUKEMIA IN CHRONIC PHASE AFTER ≥2 PRIOR TYROSINE KINASE INHIBITORS: WK 96 UPDATE D. Rea^1,*, A. Hochhaus^2, M. J. Mauro^3, Y. Minami^4, E. Lomaia^5, S. Voloshin^6, A. Turkina^7, D.-W. Kim^8, J. F. Apperley^9, J. E. Cortes^10, A. Abdo^11, L. M. Fogliatto^12, D. D. H Kim^13, P. le Coutre^14, S. Saussele^15, M. Annunziata^16, T. P. Hughes^17, N. Chaudhri^18, L. Chee^19, V. García-Gutiérrez^20, K. Sasaki^21, S. Kapoor^22, A. Allepuz^23, S. Quenet^23, V. Bédoucha^23, C. Boquimpani^24 ^1Hôpital Saint-Louis, Paris, France; ^2Universitätsklinikum Jena, Jena, Germany; ^3Memorial Sloan Kettering Cancer Center, New York, United States of America; ^4National Cancer Center Hospital East, Kashiwa, Japan; ^5Almazov National Medical Research Centre; ^6Russian Research Institute of Hematology and Transfusiology, St. Petersburg; ^7National Medical Research Center for Hematology, Moscow, Russia; ^8Uijeongbu Eulji Medical Center, Geumo-dong, Uijeongbu-si, South Korea; ^9Centre for Haematology Imperial College London, London, United Kingdom; ^10Georgia Cancer Center, Augusta, United States of America; ^11Instituto do Câncer do Estado de São Paulo (ICESPSP), São Paulo; ^12Hospital de Clínicas de Porto Alegre (HCPA), Porto Alegre, Brazil; ^13Princess Margaret Cancer Centre, University Health Network, University of Toronto, Toronto, Canada; ^14Charité–Universitätsmedizin Berlin, Berlin; ^15III. Medizinische Klinik, Medizinische Fakultät Mannheim der Universität Heidelberg, Mannheim, Germany; ^16Division of Hematology, AORN Cardarelli, Naples, Italy; ^17South Australian Health and Medical Research Institute and University of Adelaide, Adelaide, Australia; ^18King Faisal Specialist Hospital & Research Center, Riyadh, Saudi Arabia; ^19Peter MacCallum Cancer Center and The Royal Melbourne Hospital, Victoria, Australia; ^20Servicio de Hematología, Hospital Universitario Ramón y Cajal (IRYCIS), Madrid, Spain; ^21The University of Texas MD Anderson Cancer Center, Houston; ^22Novartis Pharmaceuticals Corporation, East Hanover, United States of America; ^23Novartis Pharma AG, Basel, Switzerland; ^24HEMORIO, State Institute of Hematology Arthur de Siquiera Cavalcanti, Rio de Janeiro, Brazil Oncoclínica Centro de Tratamento Oncológico, Rio de Janeiro, Brazil Background: Asciminib is the 1st BCR::ABL1 inhibitor to Specifically Target the ABL Myristoyl Pocket (STAMP). In the ASCEMBL primary analysis, asciminib had superior efficacy and better safety/tolerability vs bosutinib (BOS) in patients (pts) with chronic myeloid leukemia in chronic phase (CML-CP) after ≥2 prior tyrosine kinase inhibitors (TKIs). Major molecular response (MMR) rate at wk 24 was 25.5% on asciminib vs 13.2% on BOS; the difference in MMR rates after adjusting for baseline major cytogenetic response (MCyR) was 12.2% (95% CI, 2.19%-22.30%; 2-sided P=.029). Fewer grade ≥3 adverse events (AEs) and AEs leading to treatment discontinuation occurred on asciminib vs BOS. After a median follow-up of 2.3 years (16.5 months’ additional follow-up since the primary analysis), we report updated efficacy and safety results (cutoff: October 6, 2021). Aims: The key secondary objective was to compare MMR rate at wk 96 on asciminib vs BOS. Methods: Eligible pts provided informed consent, were adults with CML-CP after ≥2 prior TKIs, with intolerance or lack of efficacy per 2013 European LeukemiaNet recommendations. They were randomized 2:1 to asciminib 40 mg twice daily or BOS 500 mg once daily, stratified by baseline MCyR status (Ph+ metaphases ≤35%). Results: 233 pts were randomized to asciminib (n=157) or BOS (n=76). At cutoff, treatment was ongoing in 84 (53.5%) and 15 (19.7%) pts, respectively; the most common reason for discontinuation was lack of efficacy in 38 (24.2%) and 27 (35.5%) pts, respectively. MMR rate at wk 96 (per ITT) was 37.6% on asciminib and 15.8% on BOS, meeting the key secondary objective. The difference after adjusting for baseline MCyR was 21.7% (95% CI, 10.5%-33.0%; 2-sided P=.001). Preplanned subgroup analyses showed that MMR rate at wk 96 was consistently higher with asciminib than BOS in all demographic and prognostic subgroups, including all prior lines of TKI therapy, and regardless of the reason for discontinuation of the last TKI (Figure). At wk 96, more pts on asciminib than BOS had BCR::ABL1^IS ≤1% (45.1% vs 19.4%) (Table). Responses were durable, with a probability (95% CI) of maintaining MMR and BCR::ABL1^IS ≤1% for ≥72 wk of 96.7% (87.4%-99.2%) and 94.6% (86.2%-97.9%), respectively, on asciminib and 92.9% (59.1%-99.0%) and 95.0% (69.5%-99.3%), respectively, on BOS. Median time to treatment failure was 24 months on asciminib and 6 months on BOS. Median duration (range) of exposure was 103.1 (0.1-201.1) wk on asciminib and 30.5 (1.0-188.3) wk on BOS. Despite asciminib’s longer duration of exposure, its safety/tolerability continued to be better than that of BOS (Table). Fewer pts on asciminib than BOS had AEs leading to treatment discontinuation (7.7% vs 26.3%). No new on-treatment deaths were reported since the primary analysis. Most frequent (>10%) grade ≥3 AEs on asciminib vs BOS were thrombocytopenia (22.4%, 9.2%), neutropenia (18.6%, 14.5%), diarrhea (0%, 10.5%), and increased alanine aminotransferase (0.6%, 14.5%). Image: graphic file with name hs9-6-1-g036.jpg [75]Open in a new tab Summary/Conclusion: After >2 years of follow-up, asciminib continued to show clinically and statistically significant, superior efficacy and better safety/tolerability vs BOS. Responses were durable, and MMR was more than double on asciminib than BOS. The difference in MMR rates between the 2 arms increased from 12.2% at wk 24 to 21.7% at wk 96. A higher proportion of pts had BCR::ABL1^IS ≤1%, a milestone response in later lines that is associated with improved long-term survival. These results further support the use of asciminib as a new CML therapy, with the potential to transform standard of care. S156: INTERNATIONAL, PROSPECTIVE STUDY COMPARING NILOTINIB VERSUS IMATINIB WITH EARLY SWITCH TO NILOTINIB TO OBTAIN SUSTAINED TREATMENT-FREE REMISSION IN PATIENTS WITH WITH CHRONIC MYELOID LEUKEMIA F. Pane^1,*, F. Castagnetti^2, L. Luciano^1, A. Russo Rossi^3, E. Abruzzese^4, R. Bassan^5, G. Binotto^6, G. Caocci^7, G. Cimino^8, P. Fazi^9, A. Gozzini^10, M. Lunghi^11, R. Marasca^12, B. Martino^13, M. Bonifacio^14, F. Cavazzini^15, F. Paoloni^16, G. Saglio^17, S. Sica^18, A. Tafuri^19, D. Vallisa^20, M. Vignetti^9, P. Westerweel^21, G. Rosti^22, M. Breccia^23 ^1Dpt. of Clinical Medicine and Surgery, University of Naples, Naples; ^2Department of Experimental, Diagnostic, and Specialty Medicine, University of Bologna, Bologna; ^3Hematology and Transplantation Unit, University of Bari, Bari; ^4Division of Hematology, Ospedale S. Eugenio, Rome; ^5Hematology Unit, Ospedale Dell’Angelo, Venezia-Mestre; ^6Hematology Unit, University of Padova, Padova; ^7Department of Medical Sciences and Public Health, Hematology, Businco Hospital, University of Cagliari, Cagliari; ^8UOC Ematologia e trapianto, S.Maria Goretti Hospital, Latina; ^9GIMEMA Data Center, Fondazione GIMEMA Franco Mandelli Onlus, rome; ^10Department of Cellular Therapy and Transfusional Medicine, AUO Careggi, Florence; ^11Division of Hematology, Department of Translational Medicine, University of Eastern Piedmont, Novara; ^12Section of Hematology, Department of Medical Sciences, University of Modena and Reggio Emilia, Modena; ^13Division of Hematology, Azienda Ospedaliera ‘Bianchi Melacrino Morelli’, Reggio Calabria; ^14Section of Hematology, Department of Medicine, University of Verona, Verona; ^15Hematology Section, Department of Medical Sciences, University of Ferrara, University Hospital Arcispedale S. Anna, Ferrara; ^16Biostatistic Central Office, Fondazione GIMEMA Franco Mandelli Onlus, Rome; ^17Hematology Division, University of Turin, Turin; ^18Ematologia, FONDAZIONE POLICLINICO UNIVERSITARIO AGOSTINO GEMELLI IRCCS, Roma; ^19Department of Clinical and Molecular Medicine, Hematology, Sant’Andrea University HospitalSapienza University of Rome, Rome; ^20Unit of Hematology, Department of Oncology and Hematology, Guglielmo da Saliceto Hospital, Piacenza, Italy; ^21Internal Medicine, Albert Schweitzer Hospital, Dordrecht, Netherlands; ^22Hematology Unit, IRCCS Istituto Romagnolo per lo Studio dei Tumori (IRST) “Dino Amadori”, Meldola; ^23Translational medicine OUC Hematology, La Sapienza university, Rome, Italy Background: Treatment free remission (TFR) is one of the most important goals of the CML treatment but, so far, the best treatment to reach this aim is still undefined. It is widely accepted that a sustained deep molecular remission (DMR) is the pre-requisite to discontinue TKI, and it is expected that in patients who start treatment with imatinib (IMA) and fail early molecular remission (EMR), a switch to a second generation TKI may improve the probability of achieving a DMR Aims: We launched in November 2016 an international, prospective, interventional, randomized, two arms, study to evaluate both the depth of the molecular response and the rate of TFR in newly diagnosed CP-CML patients treated with a second generation TKI (Nilotinib, NIL) or with IMA followed by switching to NIL in absence of optimal response (defined according the ELN 2013 criteria (Clinical Trial number 02602314). Methods: The patients are randomized 1:1 between NIL and IM according to Sokal risk score (high versus intermediate/low risk) and country. All the patients who obtain a residual disease reduction greater than 4.0 logs (MR4.0) within the first three years of treatment and maintain this level of response up to the end of the fourth years of therapy qualify for the discontinuation phase of the study. The study has two primary end-points: a) the rates of molecular response (MR4.5) at 24 months, and b) the rate of patients who remain in sustained treatment free remission (≥MR3.0) without molecular relapse 12 months after entering the TFR phase. The molecular relapse is defined as loss of MMR or confirmed loss of MR3.0 (figure). Results: From November 2016 to January 2021, 457 patients with newly diagnosed CP-CML patients were enrolled into the study and 448 of these (228 and 220 randomized to the NIL and IMA arms, respectively) were evaluable (mean age 54.2 yrs - range 19.4 – 85.8). At baseline, 183 (40.8%), 191 (42.6%) and 72 (16.1%) patients were classified as low, intermediate, or high-risk Sokal, respectively, while 278 (62.3%), 128 (28.7%) and 40 (9.0%) had a low, intermediate, or high-risk ELTS risk score. The median follow-up of the whole cohort of patients is 30.4 mo. Fifthy-six (25.4%) of the 220 patients of the IMA arm did not fulfill the ELN criteria for optimal response within the first 12 mo. of treatment and, according to the protocol, switched to NIL therapy. At the last analysis of the protocol database (February 2022), 59 patients had had stopped the protocol treatment since their decision, death (24), toxicity (23), progression (9), uncontrolled second neoplasia (2) or protocol violation (1), 69 patients had not reach 24 mo. of follow-up and other 15 had missing data. Of the remaining 304 patients, 35 showed non optimal response to therapy. At the 24 mo. of follow-up, 76 of the 322 patients with an available molecular response (23.6%), reached a MR4.5 response that showed a significantly higher frequency within the patients randomized to the NIL arm (48 vs 28; p=0.015) (first primary co-endpoint of the study). Image: graphic file with name hs9-6-1-g037.jpg [76]Open in a new tab Summary/Conclusion: This is the first and, so far, the unique prospective study comparing not only the rate of DMR but, more important, also the rate of TFR according to treatment: a second generation TKI frontline vs IMA frontline followed by the same second generation in case of non-optimal response. The analysis of the first co-primary endpoint indicates that, despite the early switch in the IMA randomized patients, NIL therapy is more effective to induce DMR. Subsequent analysis will clarify whether the higher rates of DMR in the NIL arm may translate into a higher rate of TFR. S157: BCR::ABL1 DIGITAL PCR IDENTIFIES CHRONIC PHASE CML PATIENTS SUITABLE FOR AN EARLY TKI DISCONTINUATION ATTEMPT: A PATIENT-LEVEL META-ANALYSIS C. Kockerols^1,*, S. Dulucq^2, S. Bernardi^3, M. Farina^3, I. Civettini^4, G. Colafigli^5, S. Mori^6, P. Valk^7, F.-X. Mahon^8, C. Gambacorti-Passerini^4 9, F.-E. Nicolini^10, M. Breccia^5, D. Russo^3, P. E. Westerweel^1 ^1Internal Medicine, Albert Schweitzer Hospital, Dordrecht, Netherlands; ^2Laboratory of Hematology, Hôpital Haut Lévêque, University hospital of Bordeaux, Pessac, France; ^3Clinical and Experimental Sciences, Hematology; ^2and Bone Marrow Transplant Center, University of Brescia, ASST Spedali Civili of Brescia, Brescia; ^4University of Milano Bicocca, Monza; ^5Translational and Precision Medicine, Az. Policlinico Umberto I-Sapienza University, Rome; ^6University of Milano-Bicocca, Monza, Italy; ^7Molecular Biology, Erasmus University Medical Center, Rotterdam, Netherlands; ^8Hematology, Institut Bergonié, Bordeaux, France; ^9Hematology, S. Gerardo Hospital, Monza, Italy; ^10Hematology, Centre Léon Bérard and CRCL, Lyon, France Background: Digital PCR (D-PCR) is an emerging technique that delivers a highly accurate BCR::ABL1 quantification, even in CML patients with low residual disease. This is crucial in the context of treatment-free remission (TFR) for the selection of patients who may successfully discontinue TKI therapy. However, it is unclear how its prognostic value relates to time variables such as treatment duration prior to the TFR attempt. Current guidelines suggest aiming for a TKI treatment duration >6 years to increase TFR success rate. Aims: In current analysis, we aimed to assess the prognostic value of BCR::ABL D-PCR in relation to prior TKI treatment duration. Methods: We performed an Individual Patient Data Meta-Analysis (IPD-MA) combining data from different study cohorts in which BCR::ABL1 was assessed by D-PCR prior to TKI discontinuation. Eligibility criteria included age ≥18 years and CML diagnosed in chronic phase. Data of the participating studies were pooled and stratified based on D-PCR and/or treatment duration. BCR::ABL1 D-PCR was dichotomized based on each study-defined prediction cut-off. Strata were assessed for molecular relapse (MolR) with Kaplan-Meier estimates and with cox regression analysis including a frailty term for correction for between-study heterogeneity and including confounding variables: age at diagnosis, gender, Sokal score, TKI generation, treatment duration, DMR duration (model 1) and BCR::ABL1 transcript type (model 2). MolR was defined as a BCR::ABL1 >0,1%IS or a 1-log BCR::ABL1 increase in two consecutive analyses. Patients were censored at last follow-up. Results: For this meta-analysis, data were combined from five cohorts: four published cohorts (STIM2 cohort [Nicolini et al.], n=175; ISAV cohort [Diral et al.], n=107; Bernardi et al., n=111; Colafigli et al., n=50) and 1 unpublished cohort (Dutch cohort; n=40). The pooled dataset comprised 483 patients (Table). A total of 205 patients (42%) experienced MolR with a median time to relapse of 3 months. Median follow-up duration for TFR patients was 27 months. MolR patients had a significantly shorter treatment duration prior to TKI discontinuation (6,7 vs 7,9 years, p=0.006) and more often presented a BCR::ABL1 D-PCR above the study-defined prediction cut-off (34% vs 19%, p<0.001). Interestingly, the median treatment durations were almost identical in patients with a BCR::ABL1 below or above the cut-off (7.0 vs 7.5 years, p = 0.470). The probability of MolR at 24 months was 38% versus 58% for patients with a D-PCR BCR::ABL1 below versus above the prediction cut-off (p <0.001). In the cox regression analysis, the HR of D-PCR BCR::ABL1 below the cut-off for MolR was 0.48 (95% CI 0.35-0.66, p<0.001). Treatment duration and BCR::ABL1 transcript type also remained independent predictors, but TKI generation and DMR duration did not. When stratifying into 4 groups based on the D-PCR result and treatment duration, patients with a TKI treatment for ≥6 years and low D-PCR result had the lowest MolR rate (33% at 24 months, figure). Patients treated <6 years and with a low D-PCR result had a rate of 48% at 24 months, while patients treated <6 years and a D-PCR result above the cut-off had the highest MolR rate of 72% at 24 months. Image: graphic file with name hs9-6-1-g038.jpg [77]Open in a new tab Summary/Conclusion: These combined patient-level data of multiple CML cohorts further support the independent prognostic value of BCR-ABL1 D-PCR for TFR success rate. Importantly, patients with a TKI treatment duration <6 years and a low D-PCR result were found to have a clinically acceptable MolR rate (48%). S158: FINAL RESULT OF TKI DISCONTINUATION TRIAL WITH DASATINIB FOR SECOND ATTEMPT OF TREATMENT FREE REMISSION AFTER FAILING FIRST ATTEMPT WITH IMATINIB: TREATMENT-FREE REMISSION ACCOMPLISHED BY DASATINIB D. Kim^1,*, E. Atenafu^2, D. Forrest^3, I. Bence-Bruckler^4, L. Savoie^5, M.-M. Keating^6, L. Busque^7, R. Delage^8, A. Xenocostas^9, E. Liew^10, P. Laneuville^11, K. Paulson^12, T. Stockley^13, J. Lipton^1, B. Leber^14 ^1Princess Margaret Cancer Centre, Toronto, Canada; ^2Statistics, Princess Margaret Cancer Centre, Toronto; ^3Vancouver General Hospital, Vancouver; ^4Ottawa Hospital Research Institute, Ottawa; ^5University of Calgary, Calgary; ^6Queen Elizabeth II Health Sciences Centre, Halifax; ^7Hôpital Maisonneuve-Rosemont, Montreal; ^8Centre Universitaire d’Hématologie et d’Oncologie de Québec, Quebec; ^9London Health Sciences Centre, London; ^10University of Alberta, Edmonton; ^11McGill University Health Centre, Montreal; ^12CancerCare Manitoba, Winnipeg; ^13Division of Clinical Laboratory Genetics, University Health Network, Toronto; ^14Juravinstki Cancer Centre, Hamilton, Canada Background: The Canadian tyrosine kinase inhibitor (TKI) discontinuation (DISC) trial evaluated if Dasatinib (DA) therapy can lead to a successful second treatment-free remission (TFR2) after failing a Imatinib (IM) DISC for first TFR (TFR1) attempt. We previously reported that 1) The 12-month molecular relapse free survival (mRFS) rate for TFR1 was 58.0%; 2) doubling time (DT) at 2 months after IM DISC correlates with TFR1 failure. Aims: Here, we report the final result of TFR2 rate after DA DISC. The null hypothesis was a TFR2 rate of 17.5% while the alternative hypothesis was a TFR2 rate of 35.0%, and the study was designed to reject our null hypothesis if > 28% of pts remain in TFR2 after DA DISC. Methods: This prospective clinical trial (BMS CA180-543, NCT#02268370) had 3 phases: 1) IM DISC phase, 2) DA rechallenge phase, 3) DA DISC phase. Key inclusion criteria included: 1) CML in chronic phase at original diagnosis, 2) total duration of IM therapy of minimum 3 years, 3) total duration of MR^4.5 or deeper response over 2 years. Molecular relapse was defined as an increase in BCR-ABL qPCR > MR4 on 2 consecutive occasions, or a single increase in BCR-ABL qPCR > MR3. DA treatment was started at 100mg once daily after molecular relapse was confirmed, and continued for at least 12 months after achieving ≥ MR4 until DISC for TFR2. Results: The study was launched on March 2015 and completed all participants’ planned visits as of Feb 2022 with a median follow-up duration of 27.5 months (range 2-51 months). 1) In the IM DISC phase, 58 of 131 pts (44.3%) experienced molecular relapse with a median onset of 3.53 months, thus the remaining 73 pts achieved TFR1 (55.7%): The mRFS rate at 12 months was 56.8% (95% CI, 47.8-64.8 %). Distribution of monthly DT in 6 months is presented in Fig A. 2) In the DA rechallenge phase, out of the 58 pts who failed TFR1, 51 pts received DA. At 3 months, 98.0%, 87.9%, and 75.4% of pts achieved MMR, MR4 and MR4.5. The halving time (HT) after DA re-therapy is summarized in Fig B. Notably, HT at 2 and 3 months was 10.6 and 10.7 days, consistent with rapid reduction of BCR-ABL qPCR in first 3 months. 3) In the DA DISC phase, 35 pts who attained MR4 or deeper response for ≥ 12 months discontinued DA for a TFR2 attempt. Out of 35 pts, only 4 pts (11.4%) has maintained the molecular response at last follow-up, while the remaining 31 pts lost the molecular response with a median 3.65 months. The actuarial mRFS rate at 6 and 12 months was 22.9% (10.8-37.6%) and 10.0% (2.7-23.1%). Monthly DT within 6 months after DA DISC is presented in Fig C. Based on the final result of 11.4% TFR2 rate, we conclude that 12 months’ DA re-therapy could not improve TFR2 rate significantly. 4) For DT after IM DISC, average DT at 2/3 months was much shorter in those lost molecular response (10.6 / 10.7 days) than those who did not (23.5 / 30.1 days). For DT after DA DISC, while average DT was 7.6 and 14.3 days at 2/3 months in overall population, it was much shorter in those lost molecular response within 6 months (3.2 and 13.3 days at 2/3 months). Image: graphic file with name hs9-6-1-g039.jpg [78]Open in a new tab Summary/Conclusion: The final result indicates that re-challenge with DA after TFR1 failure with IM DISC is effective in restoring deep molecular response as most cases rapidly regained at least MR4. However, 12 months’ DA re-therapy does not significantly improve TFR2 rate. Further studies should consider increasing MR4 duration before TFR2 attempt, adding other therapeutics and refining risk factor for TFR2 failure. S159: QUÉBEC CML RESEARCH GROUP ANALYSIS OF TREATMENT PATTERNS IN CHRONIC MYELOGENOUS LEUKEMIA: SWITCHING IS DRIVEN BY INTOLERANCE AND SIMILAR ACROSS TYROSINE KINASE INHIBITORS AND LINES OF TREATMENT L. Busque^1 2,*, M. Harnois^2, N. Szuber^1 2, R. Delage^2 3, L. Mollica^1 2, H. Olney^2 4, P. Laneuville^2 5, S. Sirhan^2 6, G. Cournoyer^2 7, I. Chamakhi^2 8, M. Lalancette^2 3, D. Talbot^2 9, V. Éthier^2 10, P. Desjardins^2 11, S. Assouline^2 6 ^1Hematology, Hôpital Maisonneuve-Rosemont, Université de Montréal; ^2Groupe Québécois de Recherche en LMC-NMP, GQR-LMC-NMP, Montreal; ^3Hematology, Centre Hospitalier Universitaire de Québec (CHUQ), Université Laval, Québec; ^4Hematology, Centre Hospitalier Universitaire de Montréal (CHUM), Université de Montréal; ^5Hematology, McGill University Health Center (MUHC), McGill University; ^6Hematology, Jewish General Hospital, McGill University, Montreal; ^7Hemato-oncology, Hôpital regional de St-Jérôme, St-Jérôme; ^8Hemato-oncology, Hôpital du Sacré-Cœur de Montréal, Université de Montréal, Montreal; ^9Hemato-oncology, Hôpital de la Cité-de-la-Santé, Laval; ^10Hemato-oncology, Centre Hospitalier Universitaire de Sherbrooke (CHUS), Université de Sherbrooke, Sherbrooke; ^11Hemato-oncology, Hôpital Charles-Lemoyne, Université de Sherbrooke, Greenfield Park, Canada Background: In 2001, imatinib was the first tyrosine kinase inhibitor (TKI) approved for the treatment of chronic myelogenous leukemia (CML) which translated into a revolution in the management of this disease. The arsenal of TKIs was progressively reinforced with the addition of other TKIs such as dasatinib, nilotinib, bosutinib, ponatinib and recently asciminib. Two decades of clinical trials have provided critical insight into differential efficacies and specific toxicity profiles of these TKIs supporting the development of guidelines such as the European Leukemia Net (ELN) and the national comprehensive cancer network (NCCN). Real-world evidence studies provide unique and complementary information on treatment patterns, efficacy, side effects and may help identify unmet medical needs of CML patients Aims: This study was aimed at studying frequency of TKI switching, reason for switch, duration of treatment without switching as a function of line of treatment and specific TKI using the real-world data of the Québec registry created in 2009 Methods: Patients with Philadelphia positive (Ph+) CML were recruited with informed consent in the Québec registry. 795 patients were included in this analysis. Data from the registry were extracted July 31^ist 2021. Switching was defined as a change of a specific TKI to another. Reasons to initiate switching were categorized as resistance (primary and secondary) or intolerance (hematological or non-hematological). Standard statistical methods were used to evaluated bivariate and continuous variables. Kaplan-Meier curve were used to evaluate overall survival and survival without switching (JMP® Pro 14.1 software, SAS Institute, Cary, NC, USA). Results: At the time of data collection, the median time of follow-up was 7.5 years. Proportion of switching per line of treatment: 1^st: 357/795 (44.9%); 2^nd: 157/357 (43,9%); 3^rd: 59/157 (37,6%); 4^th:23/59 (39%). The reason for switching (ratio intolerance/resistance) for each line were: 1^st: 1.33; 2^nd: 4.3; 3^rd: 2.4; 4^th: 2.7. 20 patients switched serially across all lines for intolerance and only 3 for serial resistance. Survival without switching was evaluated using a Kaplan-Meier curve by line of treatment and by specific TKI (Figure 1). The survival without switching was similar for patients on imatinib, dasatinib or nilotinib in first line as in second line. In third line, there was no difference between imatinib, nilotinib, dasatinib or bosutinib although numbers of patients were small. We then compared the survival from diagnosis of patients that remained in first or second line versus the patients that reached third or more lines of treatment. Although the mean age at diagnosis and SD were identical between the two groups (54,8 years and SD 15,2) there was a statistically significant difference in survival in favor of patients needing only 1 or 2 lines of treatment (P= 0.0254). Image: graphic file with name hs9-6-1-g040.jpg [79]Open in a new tab Summary/Conclusion: We demonstrate: (i) that switching of TKI is frequent and mainly driven by intolerance in all lines of treatment; (ii) serial intolerance is 6.6 time more frequent than serial resistance suggesting a class effect for intolerance in some patients; (iii) all TKIs have a similar «retention level» in all lines of treatment; (iv) that patients necessitating 3 or more lines of treatment have a survival disadvantage. Our results suggest that one of the most important unmet medical need in CML management is availability of better tolerated drugs S160: MOLECULAR DETERMINANTS OF DISEASE PROGRESSION AFTER HYPOMETHYLATING AGENT THERAPY IN RAS PATHWAY MUTANT CHRONIC MYELOMONOCYTIC LEUKEMIA AT THE SINGLE-CELL LEVEL G. Montalban-Bravo^1,*, F. Ma^2, I. Ganan-Gomez^1, R. Kanagal-Shamana^3, V. Adema^1, N. Thongon^1, H. Yang^1, K. A. Soltysiak^1, C. Bueso-Ramos^3, H. Kantarjian^1, G. Garcia-Manero^1, S. Colla^1 ^1Leukemia, The University of Texas MD Anderson Cancer Center, Houston; ^2Molecular Biology Institute, University of California, Los Angeles; ^3Hematopathology, The University of Texas MD Anderson Cancer Center, Houston, United States of America Background: Most patients (pts) with chronic myelomonocytic leukemia (CMML) have incomplete or transient responses to hypomethylating agent (HMA) therapy. CMML cases driven by mutations in RAS pathway signaling genes or ASXL1 have a higher risk of failure and progression to acute myeloid leukemia (AML). Development of effective alternative therapies has been delayed, owing to an incomplete understanding of how different hematopoietic populations contributes to disease maintenance and progression. Aims: We aimed to dissect the cellular and molecular mechanisms underpinning CMML maintenance and progression in RAS mutant CMML. Methods: We performed single-cell RNA sequencing (scRNA-seq) analysis of lineage-negative (Lin^-) CD34^+ hematopoietic stem and progenitor cells (HSPCs) and BM mononuclear cells (MNCs) isolated from RAS pathway mutant CMML pts (n=5 and 6, respectively) and age-matched healthy donors (HD; n=2 and 3, respectively). CMML samples were obtained at the time of diagnosis and HMA failure. Additionally, we performed scATAC-seq analysis of Lin^-CD34^+ HPSCs isolated at the times of diagnosis and progression (n=1). Results: Our analysis revealed that CMML HSPCs had a predominantly granulomonocytic differentiation route with increased frequencies of myeloid-monocytic progenitors, at the expense of hematopoietic stem cells (HSCs) (Fig 1a), and upregulated expression of genes involved in the oxidative phosphorylation, type I interferon (IFN) and IFNg pathways. Consistent with these results, scRNA-seq analysis of MNCs revealed expanded populations of myelomonocytic progenitors and monocytes and upregulated expression of genes involved in IFNg response and NF-kB activation (Fig 1b), along with upregulation of the NF-kB transcriptional effector BCL2A1 (Fig 1c). Assessment of ligand-receptor interactions using the CellPhoneDB repository identified that CMML monocytes established a high number of cell-cell interactions (n=638) with dendritic cells, NK cells, and HSPCs via chemokines, cytokines, and inhibitory molecules known to induce NF-kB signaling and NK-cell exhaustion. Disease progression was associated with expansion of lympho-myeloid progenitors (LMPPs) (Fig 1d) characterized by the highest levels of IFNg response, NF-kB survival signaling, and cell cycle regulators. scATAC-seq of Lin^-CD34^+ confirmed higher activity of transcriptional factors associated with monocytic differentiation and NF-kB signaling (Fig 1e-f). Accordingly, scRNA-seq analysis of MNCs showed increased frequencies of HSPCs and myelomonocytic precursors, a reduction of T cells (Fig 1f), and emergence of a monocyte population characterized by the highest expression of NF-kB signaling and its effectors MCL1 and BCL2A1. BCL2A1 protein expression at progression was confirmed by immunohistochemistry (Fig 1g). CellPhoneDB analysis identified a high number of cell-cell interactions (n=2978) involving cytokines, chemokines, and surface proteins known to elicit NF-kB activation and immune evasion between expanded monocytes, LMPPs, myelomonocytic precursors, and immune cells during progression. Image: graphic file with name hs9-6-1-g041.jpg [80]Open in a new tab Summary/Conclusion: Our data suggests that CMML is maintained through metabolically active HSPCs, which leads to monocytes’ reprograming and survival through NF-kB signaling activation. We showed that disease progression arises from the expansion of NF-kB dependent immature myeloid progenitors, which leads to therapy resistance and immune evasion. This study has implications for the development of therapies targeting downstream effectors of NF-kB–mediated survival pathway to overcome treatment failure. In vitro validation is ongoing. S161: ZRSR2 AND TET2 MUTATIONS PROMOTE MDS BY DYSREGULATING GENE EXPRESSION AND ABERRANT ALTERNATIVE SPLICING IN MICE C. Garcia-Ruiz^1,*, C. Martínez-Valiente^1, A. Liquori^1, A. Gutiérrez-Adán^2, J. Cervera^3, A. Sanjuan-Pla^1 ^1Hematology Research Group, IIS La Fe, Valencia; ^2Animal Reproduction Department, INIA, Madrid; ^3Genetics Unit, Hospital Universitario y Politécnico La Fe, Valencia, Spain Background: Mutations in splicing factors and epigenetic regulators are the most frequent genetic alterations in patients with myelodysplastic syndromes (MDS). The minor spliceosome factor ZRSR2 and the epigenetic regulator TET2 appear significantly associated in MDS patients. However, the functional impact of such mutations in the hematopoietic system and MDS have been scarcely studied. To address this question, we established a murine model (Zrsr2^m/mTet2^−/−) carrying mutations in both genes, which exhibited signs compatible with MDS disease in mice. However, the molecular disease mechanism has not yet been elucidated. Aims: To interrogate the impact of ZRSR2 and TET2 mutations in gene expression and alternative splicing in the context of hematopoiesis and MDS. Methods: Whole transcriptome sequencing was performed to investigate changes in gene expression and alternative splicing patterns. Lin^-Sca-1^+c-kit^+ (LSK) cells (50,000 per sample) were sorted from pools of 3 mice, and mRNA was sequenced in an Illumina NovaSeq 6000 platform. Differential gene expression was determined using DESeq2 with adjusted P-value ≤ 0.05, log2FoldChange ≥ 0. Functional enrichment from differentially spliced genes was identified with KEGG, and the enrichment cut-off p-value was set as adjusted P-value ≤ 0.05. Relevant alternative splicing events were validated by RT-PCR. Results: RNA-seq analysis of Zrsr2^m/mTet2^−/−cells revealed 2952 differentially expressed genes (DEG), compared to 571 in Zrsr2^m/m and 1203 in Tet2^−/−. From 2952, 1327 were up-regulated and 1625 were downregulated. Interestingly, relevant genes for hematopoietic lineage specification were significantly dysregulated. In particular, genes related to lymphoid lineage (Flt3, Notch1, and Tlr7) were downregulated, while those related to myeloid lineage (Ccl9, Mpo, Ms4a6b, and Prtn3) and megakaryocytic-erythroid lineage (Pdfgrb, Itgb3, Optn, and Tgfbr3) were upregulated. This suggests an early myeloid and megakaryocyte-erythroid priming of LSK towards the production of cells from these lineages in Zrsr2^m/mTet2^−/− mice. Enrichment analysis of DEG using Go enrichment analysis identified ribosome function, inflammation, and migration/motility processes as the most significantly altered. Further, KEGG pathway enrichment pointed ribosome, the MAPK family, and pro-inflammatory pathways as significantly enriched in Zrsr2^m/mTet2^−/− cells. Finally, KEGG pathway enrichment of alternative splicing targets identified the MAPK family and the Fanconi anemia pathway as the most altered targets in Zrsr2^m/mTet2^−/− cells. Importantly, a total of 9 genes related to the MAPK pathway were identified as mis-spliced, from which Dusp1, Tgfbr2, and Fgf11 were validated in this study. Summary/Conclusion: In this study, we broaden our previous report and show that concurrent mutations in Zrsr2 and Tet2 dysregulate normal gene expression and cause aberrant mRNA splicing. Gene expression analysis identified ribosome function, inflammation, and migration/motility as the most altered processes in Zrsr2^m/mTet2^−/− cells. Alternative splicing analysis identified the MAPK and the Fanconi Anemia pathway as key targets of aberrant splicing. All in all, gene expression dysregulation and aberrant mRNA splicing disturb important biological pathways and drive the molecular pathomechanism in Zrsr2^m/mTet2^−/− mice. S162: SOMATIC GENETIC LANDSCAPE IN GATA2 DEFICIENCY PATIENTS L. Largeaud^1 2,*, M. Collin^3, N. Monselet^4, F. Vergez^5, L. Larcher^6, P. Hirsch^7, N. Duployez^8, J. Bustamante^9, C. Bellanné-Chantelot^10, J. Donadieu^11, F. Sicre de Fontbrune^12, M. Nolla^13, C. Fieschi^14, F. Delhommeau^7, E. Delabesse^15, M. Pasquet^13 ^1Clinical Haematology laboratory, Toulouse Hospital; ^2UMR1037, Cancer Research Center of Toulouse, Toulouse, France; ^3Haematology Department, Institute of Cellular Medicine Newcastle University, Newcastle, United Kingdom; ^4Bureau des essais cliniques, Claudius Rigaud Institut; ^5Haematology Department, Toulouse Hospital, Toulouse; ^6Clinical Haematology laboratory, Saint Louis Hospital APHP; ^7Clinical Haematology laboratory, Saint-Antoine Hospital, APHP, Paris; ^8Clinical Haematology laboratory, CHU Lille, Lille; ^9Unité centre d’études des déficits immunitaires, Necker Hospital APHP; ^10Centre de génétique moléculaire et chromosomique, Hôpital Pitié-Salpêtrière, APHP; ^11Haematology Department, Trousseau Hospital, APHP; ^12Service d’hématologie greffe, Saint-Louis Hospital, APHP, Paris; ^13Service d’hématologie et immunologie pédiatrique, Toulouse Hospital, Toulouse; ^14Immunologie clinique, Saint-Louis Hospital, APHP, Paris; ^15Clinical Hematology laboratory, Toulouse Hospital, Toulouse, France Background: Heterozygous germline GATA2 mutations strongly predispose to myeloid malignancies, immunodeficiency, and/or lymphedema. The progression towards haematological diseases seems to be correlated with the acquisition of molecular and cytogenetic abnormalities. Aims: Our study therefore focuses on the biological characterization of the different progression stages of germline GATA2 deficiency patients and their correlation with clinic features. Methods: We describe here a cohort of 78 patients (62 from the French-Belgian cohort and 16 British patients). Molecular analysis by next generation sequencing (targeting 90 genes frequently mutated in myeloid malignancies with a sensitivity = 1%) was performed in 76 patients. Cytogenetic analyses were determined for 76 patients. Results: Median age of our cohort was 21.8 years [6months–62years]. 44 had missense mutations, 32 had null mutations, one intronic and one synonymous mutation. Our results showed a trend toward a younger age at diagnosis in patients harboring null mutations compared to patients with missense mutations (13 vs 17 years, p=0,086) and more chronic infections during follow-up (74% vs 25%, p<0,001). 55 karyotypic abnormalities were identified including monosomy 7 (29%), trisomy 8 (16%) and der (1;7) (9%). Moreover, 141 somatic mutations were identified targeting STAG2 (38%), ASXL1 (13%), SETBP1 (8%), EZH2 (4%) and RUNX1 (3%) genes. To better stratify patients, we defined 3 spectra based on morphological bone marrow analysis: 12 (15%) patients had spectrum 0 (normal bone marrow), 47 (60%) had spectrum 1 (hypoplastic marrow and/or low-grade myelodysplasia (MDS)) and 19 (25%) were classified as spectrum 2 (MDS with excess blasts, AML and CMML). We found a genotype-phenotype correlation: spectrum 1 is more associated with null mutations and spectra 0 and 2 with missense mutations (p=0.023). Patients in spectrum 0 exhibited no acquired molecular and karyotypic abnormalities. Spectrum 2 was enriched with mutations of SETBP1, RAS pathway genes, and RUNX1 as well as more other cytogenetic abnormalities excluding -7, tri8 and der(1;7) (p<0,001). The proportion of monosomy 7 was higher in spectrum 2 than in spectrum 1 without reaching the significance level (47% vs 28%). Conversely, STAG2 mutations were mainly found in spectrum 1. 53 STAG2 mutations were identified in 25 patients (1 to 8 mutations per patient), mostly at low mutated cell fraction (median: 6%). Follow-up of 3 patients with STAG2 clones in spectrum 1 showed no progression toward spectrum 2. Moreover, clonal hierarchy of 3 patients with STAG2 mutations in spectrum 2 showed that STAG2 mutations were subclonal events. Our results suggested that STAG2 mutation did not act as driver and could be related to clonal hematopoiesis. Interestingly, Stag2 KO mice showed an enrichment of GATA2 binding motif (Ochi et al. Cancer Discov 2020) which could lead to an increase of GATA2 activity. We speculate that in GATA2 deficiency syndrome, STAG2 mutations could induce a mechanism of indirect Somatic Genetic Rescue (SGR) by compensating the loss of GATA2 activity induced by the mutated allele. Image: graphic file with name hs9-6-1-g042.jpg [81]Open in a new tab Summary/Conclusion: This study reported that STAG2 mutations were recurrent in GATA2 deficiency clonal hematopoiesis. Some genetic abnormalities are associated with the leukemic transformation stage such as SETBP1, RAS pathway genes and other cytogenetic abnormalities. An improved ability to identify patients with high risk of developing leukemic transformation has the potential to improve clinical outcomes and help clinicians determine the optimal timing of bone marrow transplantation. S163: EFFICACY OF JAK 1/2 INHIBITION IN MURINE IMMUNE BONE MARROW FAILURE E. Groarke^1,*, X. Feng^1, N. Aggarwal^1, A. L. Manley^1, Z. Wu^1, S. Gao^1, B. Patel^1, J. Chen^1, N. Young^1 ^1Hematology Branch, National Heart, Lung, and Blood Institute, Bethesda, United States of America Background: Immune aplastic anemia (AA) is a severe blood disorder characterized by cytotoxic T-lymphocyte mediated stem cell destruction. Therapies including hematopoietic stem cell transplant and immunosuppression are effective but entail costs and risks, and are not effective or possible in all patients. The Janus Kinase (JAK) 1/2 inhibitor ruxolitinib (RUX) suppresses cytotoxic T cell activation and inhibits production of interferon gamma and tumor necrosis factor alpha in models of graft-versus-host disease. Aims: Assess RUX in murine immune AA for potential therapeutic benefit. Methods: In a murine major histocompatibility complex mismatched C67BL/6(B6) to CByB6F1 lymph node (LN) cell infusion AA model, and a C.B10 minor histocompatibility antigen mismatched B6 to C.B10 LN cell infusion AA model, RUX was administered as a food additive (Rux-chow), which achieves therapeutic levels in 48-72 hours after administration. Control BMF mice received chow without RUX. Animals were fed with Rux-chow two days before LN cell infusion as a prophylaxis (BMF+RUXD-2), or two days after LN cell infusion as therapy (BMF+RUXD+2). Recipient mice were either bled and euthanized at day 14 following LN infusion to collect tissues or were kept for 56 days to record animal survival. Blood counts were measured biweekly. Samples for flow cytometry, histology, and gene expression assays were collected. Results: In both AA murine models, RUX attenuated bone marrow hypoplasia, ameliorated peripheral blood pancytopenia, and prevented mortality when used either prophylactically or therapeutically. Treated mice had significantly higher blood counts (neutrophils, red blood cells, hemoglobin, platelets) as well as bone marrow (BM) and RBM (residual BM, excluding T-cells), and cellularity relative to control BMF mice. RUX suppressed infiltration, proliferation and activation of effector T cells in the bone marrow and mitigated Fas-mediated apoptotic destruction of target hematopoietic cells. All mice who received RUX were alive at 56 days while control BMF mice all died (Figure 1). With the exception of two mice with low RBC at day 56, discontinuing Rux-chow at day 28 or day 42 did not affect animal blood counts measurements, nor survival. Gene expression in mice who received RUX revealed downregulated T cell function and JAK/STAT pathway-related genes (Stat1, Stat3, Stat4, Fas, Ly6a, Infg, Gzmb, Gzma, Gzmk, Infgr1, Il2rb, Il2rg, and Lag3). On network analysis of differentially expressed genes in downregulated pathways, Stat1 and Ifn-g genes were at the center of the network, connecting immune responses and cell cycle pathways. When toxicity was assessed, RUX exerted modest suppression of lymphoid and erythroid hematopoiesis in normal and irradiated CByB6F1 mice, but the drug showed impressive clinical efficacy despite this. Image: graphic file with name hs9-6-1-g043.jpg [82]Open in a new tab Summary/Conclusion: RUX showed striking therapeutic efficacy, improving blood counts and prolonging survival in two different BMF murine models. In patients, clonal T cell expansion and activation, IFN-g and TNF-α upregulation, and Fas mediated cell death, lead to severe BM destruction and the development of AA. Our study demonstrated that JAK 1/2 inhibition with RUX produced its suppressive effects by inhibiting T cell activation, reducing inflammatory cytokine secretion, and limiting FasL/Fas-mediated BM destruction. Given these results, JAK 1/2 inhibition with RUX is an exciting and novel potential therapy for BMF patients based on a well characterized mechanism of action. S164: CLONAL DYNAMICS, IMMUNE PHENOTYPES, AND TARGETS OF BONE MARROW-INFILTRATING T CELLS IN ACQUIRED APLASTIC ANEMIA A. Ben Hamza^1,*, C. Welters^1, M. Brüggemann^2, K. Dietze^1, L. Bullinger^1 3, T. H. Brümmendorf^4 5, J. Strobel^6, H. Hackstein^6, K. Dornmair^7, F. Beier^4 5, L. Hansmann^1 3 ^1Department of Hematology, Oncology and Cancer Immunology, Charité - Universitätsmedizin Berlin, Berlin; ^2Department of Medicine II, Hematology and Oncology, University Hospital Schleswig Holstein, Kiel; ^3German Cancer Consortium (DKTK), partner site Berlin, and German Cancer Research Center (DKFZ), Heidelberg; ^4Department of Hematology, Oncology, Hemostaseology and Stem Cell Transplantation, Medical Faculty, RWTH Aachen University; ^5Center for Integrated Oncology, Aachen Bonn Cologne Düsseldorf (CIO ABCD), Aachen; ^6Department of Transfusion Medicine and Haemostaseology, University Hospital of Erlangen, Friedrich Alexander University Erlangen-Nürnberg (FAU), Erlangen; ^7Institute of Clinical Neuroimmunology, Biomedical Center and University Hospital, Ludwig-Maximilian-University Munich, Munich, Germany Background: Acquired aplastic anemia is a hematological disease characterized by hypocellular bone marrow with insufficient hematopoiesis and pancytopenia. High response rates to immunosuppressive therapy with anti-thymocyte-globulin and cyclosporin A suggest a key role for T cells in disease pathogenesis. However, bone marrow-infiltrating clonal T cell expansion, associated immune phenotypes, and T cell targets remain poorly understood. Aims: To defined clonal T cell expansion, associated immune phenotypes, and potential T cell targets in bone marrow specimens of aplastic anemia patients before and after immunosuppressive therapy. Methods: Within bone marrow of 15 patients with acquired aplastic anemia, we determined T cell clone-associated immune phenotypes by multi-parameter flow cytometry single cell index sorting for subsequent T cell receptor (TCR) αβ and cytokine/transcription factor sequencing. T cell clones of 10 patients were monitored before and after immunosuppressive therapy (median time span: 9 months) by TCRβ repertoire sequencing. Twenty-seven TCRs of expanded clones were re-expressed in reporter cell lines to determine recognition of autologous hematopoietic precursor cells. Results: We detected oligoclonal T cell expansion in all patients and 92.9% of all expanded clones were CD8^+, while only 9.6% were CD4^+. Frequencies of dominant T cell clones varied between patients (1.1-20.6% of CD8^+ T cells). Expanded CD8^+ clones were almost exclusively CCR7^- PD1^- TIM3^- CD39^- and frequently expressed cytokines and transcription factors associated with cytotoxic effector differentiation (IFNG, GZMB, PRF1, TBX21). More than 50% of the CD45RA^+ clones (T[EMRA]) were CD57^+, while CD45RA^- clones (T[EM]) were mainly CD28^+ and smaller in size. While T[EM] clones showed similar frequencies in bone marrow and peripheral blood, T[EMRA] displayed higher localization-dependent differences in size. Interestingly, almost all (91.1%) expanded clones with frequencies >1% of all T cells persisted in the bone marrow after immunosuppressive therapy independent of clinical response. Similarly, highly expanded CD4^+ clones, albeit low in numbers, were stable in frequencies before and after therapy. To determine targets of CD8^+ T cell clones with strong expansion, we re-expressed twenty-seven TCRs of eight patients and found five to be specific for immunodominant epitopes of cytomegalovirus. Strikingly, two TCRs recognized hematopoietic progenitor cells expanded from CD34-enriched autologous bone marrow. Summary/Conclusion: We tracked T cell clonotypes and associated immune phenotypes under immunosuppressive therapy at the single cell level. Expanded CD8^+ clones showed cytotoxic effector differentiation and persisted in the bone marrow after immunosuppressive therapy even in situations of therapeutic response, challenging the hypothesis that clinical response follows the decline of highly expanded autoreactive T cell clones. We identified two expanded T cell clones that recognized hematopoietic progenitor populations providing experimental proof of concept that expanded autoreactive T cell clones can be critically involved in aplastic anemia pathogenesis. FB and LH contributed equally. S165: INTEGRATED GENETIC DIAGNOSTICS OF PATIENTS WITH EARLY ONSET OF DE NOVO MYELODYSPLASTIC SYNDROMES. E. Attardi^1,*, L. Tiberi^2, D. Formicola^3, R. Artuso^3, V. Santini^1 ^1MDS Unit, Hematology, AOU Careggi - Department of Experimental and Clinical Medicine; ^2Department of Experimental and Clinical Biomedical Sciences “Mario Serio”, University of Florence; ^3Medical Genetics Unit, Meyer Children’s University Hospital, Florence, Italy Background: Cytogenetic and molecular alterations determine myelodysplastic syndrome (MDS) prognosis and must be evaluated during disease history. Relevance of germline (GM) predisposition in MDS was stressed in WHO classification, but its actual incidence is probably underestimated. Recently, the widespread availability of large-scale genomic sequencing techniques has facilitated the investigation of GM variants, and the identification of novel GM variants predisposing to MDS is expected. In de novo MDS patients (pts) with an early onset of the disease (age ≤55 yrs) when no syndromic signs drive the genetic assessment, whole exome sequencing (WES) analysis of bone marrow (BM) and saliva can constitute an optimal diagnostic procedure, improving the performance of targeted next-generation sequencing (t-NGS) approach. Aims: We adopted integrated diagnostics to define the cytogenetic profile, and WES to investigate possible GM predisposition of a cohort of 30 de novo MDS pts with an unusual young age at disease onset (≤55 yrs). Standard karyotype (SK) was paralleled by low coverage whole genome sequencing (lc-WGS) on plasma cell-free DNA (cf-DNA), as non-invasive screening test for detection of structural chromosome abnormalities. Methods: Patient BM-DNA / peripheral blood DNA (PB-DNA) was analysed by WES and compared to DNA extracted from saliva (S-DNA) samples as control for GM variants. To identify high-confidence MDS variants, we attributed an internal deleteriousness score to define variants, from “of uncertain significance” to “pathogenic” (ACMG guidelines). Variants were suspected as GM where variant allele frequency was > 0.3 and had been confirmed for genes detected in both S-DNA and BM/PB-DNA. To confirm the somatic nature of some variants, deep t-NGS of 27 myeloid custom panel gene was applied in 26/30 MDS cases on BM-DNA. SK and lc-WGS (0.2x) on plasma cf-DNA were assessed in 27/30 pts. Results: The small cohort of “young” MDS pts here analysed (Table 1). The majority of pts, 17/30 (57%) presented at least one high-confidence variant evaluated as GM for MDS. A total of 26 GM variants were recognized. Patients could be grouped on the basis of the type of variant as follows: 11/30 pts (37%) presented variants in genes involved in DNA repair defects and cancer predisposition (ATM, ATR, FANCA, FANCM, PARN, BRCA1, BRCA2, CHEK2); 4/30 (13%) had variants involved in genetic predisposition to myeloid neoplasm - GATA2, ANKRD26 and RBBP6 - the latter classically associated to MPN familial cases; one presented compound heterozygous variants in SBDS; 4 presented variants related to hereditary red blood cell defects. In our cohort, we identified cases with a complex mode of inheritance (4/17) as well as high confidence GM variants in 4 pts in 2 new interesting cancer predisposition genes: RBBP6 and PARN. Lc-WGS of plasma cf-DNA confirmed all the cytogenetic alterations found by SK (8/27) and identified a 21q22.12 deletion involving RUNX1 locus in a patient, not revealed by SK. Image: graphic file with name hs9-6-1-g044.jpg [83]Open in a new tab Summary/Conclusion: By means of an integrated diagnostics, defined as the convergence of diagnostic techniques with advanced information technology (WES, t-NGS, lc-WGS), as a proof of principle, it was possible to fully characterize the genetic asset of a particular subgroup of 30 MDS cases. We also demonstrated that lc-WGS of plasma cf-DNA has an excellent sensitivity allowing to perform periodic cytogenetic evaluations in MDS pts without invasive maneuvers. We showed that, by the use of WES, GM alterations were demonstrated in a small cohort of “young” non-syndromic MDS pts in 57% of cases, a proportion unexpectedly high. S166: MAGROLIMAB IN COMBINATION WITH AZACITIDINE FOR PATIENTS WITH UNTREATED HIGHER-RISK MYELODYSPLASTIC SYNDROMES (HR MDS): 5F9005 PHASE 1B STUDY RESULTS D. A. Sallman^1,*, M. M. Al Malki^2, A. S. Asch^3, E. S. Wang^4, J. G. Jurcic^5, T. J. Bradley^6, I. W. Flinn^7, D. A. Pollyea^8, S. N. Kambhampati^9, T. N. Tanaka^10, J. F. Zeidner^11, G. Garcia-Manero^12, D. Jeyakumar^13, L. Gu^14, A. Tan^14, M. Chao^14, C. O’Hear^14, I. Lal^14, P. Vyas^15, N. Daver^12 ^1Moffitt Cancer Center, Tampa; ^2City of Hope National Medical Center, Duarte; ^3Stephenson Cancer Center, Oklahoma University Health, Oklahoma City; ^4Roswell Park Comprehensive Cancer Center, Buffalo; ^5Columbia University Medical Center, New York; ^6Sylvester Comprehensive Cancer Center, University of Miami Miller School of Medicine, Miami; ^7Tennessee Oncology, Nashville; ^8University of Colorado School of Medicine, Denver; ^9Sarah Cannon Research Institute, Kansas City; ^10University of California San Diego Moores Cancer Center, San Diego; ^11Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill; ^12The University of Texas MD Anderson Cancer Center, Houston; ^13University of California Irvine, Orange; ^14Gilead Sciences, Inc., Foster City, United States of America; ^15Weatherall Institute of Molecular Medicine, University of Oxford, Oxford, United Kingdom Background: Magrolimab is a monoclonal antibody that blocks CD47, a “don’t eat me” signal overexpressed on cancer cells. CD47 blockade by magrolimab induces macrophage-mediated phagocytosis of tumor cells and is synergistic with azacitidine (AZA) in the upregulation of “eat me” signals. A high unmet need exists to build on current standard-of-care AZA frontline therapy to increase efficacy while maintaining a tolerable safety profile in patients (pts) with HR MDS. Aims: To report the final safety/tolerability and efficacy data from a phase 1b trial of magrolimab + AZA in pts with untreated HR MDS ([84]NCT03248479). Methods: Pts with previously untreated intermediate, high, or very high–risk MDS per the Revised International Prognostic Scoring System (IPSS-R) received magrolimab IV as a priming dose (1 mg/kg) followed by ramp-up to a 30-mg/kg QW or Q2W maintenance dose. AZA 75 mg/m^2 was administered IV or SC on days 1-7 of each 28-day cycle. Primary endpoints were safety/tolerability and complete remission (CR) rate. Results: A total of 95 pts (median age, 69 y [range, 28-91 y]) were treated. IPSS-R risk was intermediate, high, or very high in 27%, 52%, and 21% of pts. MDS was therapy related in 22% of pts; 26% had a TP53 mutation, and 62% had poor-risk cytogenetics. Median number of cycles was 6 (range, 1-27). The most common treatment-emergent adverse events (TEAEs) included constipation (68%), thrombocytopenia (55%), anemia (52%), neutropenia (47%), nausea (46%), and diarrhea (44%). The most common grade 3/4 TEAEs included anemia (47%), neutropenia (46%), thrombocytopenia (46%), and white blood cell count decreased (30%). Six pts discontinued treatment due to AEs. The 60-day mortality rate was 2%. Median hemoglobin change from baseline (BL) at first postdose sample was −0.7 g/dL (range, −3.1 to 2.4 g/dL). CR and objective response (OR) rates were 33% and 75%, with 31% of OR-evaluable pts with abnormal cytogenetics at BL having cytogenetic CR. Median time to first OR, duration of CR (DCR), duration of OR, and progression-free survival (PFS) were 1.9, 11.1, 9.8, and 11.6 mo. Overall survival (OS) rates at 12 and 24 mo were 75% and 52%; median OS was not reached (NR) with 17.1 mo of follow-up (figure). In pts evaluated with sequential whole-exome sequencing with a variant allele frequency (VAF) cutoff of 5%, 3 of 3 pts with TP53 mutation who achieved CR had TP53 VAF <5% by cycle 5 day 1. Favorable outcomes were observed both in pts with TP53 mutation (CR rate, 40%; median OS, 16.3 mo) and wild-type TP53 (CR rate, 31%; median OS, NR) (table). Table. Outcomes in patients with previously untreated HR MDS treated with magrolimab + AZA Outcome All patients N=95 ^a Wild-type TP53 n=61 TP53 mutation n=25 OR rate, % ^b 75 79 68 CR rate (95% CI), % 33 (23-43) 31 (20-44) 40 (21-61) Marrow CR rate, % 32 38 20 Stable disease with HI rate, % 11 10 8 Median DCR (95% CI), mo 11.1 (7.6-13.4) 12.9 (8.0-NR) 7.6 (3.1-13.4) Marrow CR rate with HI/any HI, % 17/59 20/61 12/56 Converted to red blood cell transfusion independence, % 14 10 24 Median PFS (95% CI), mo 11.6 (9.0-14.0) 11.8 (8.8-16.6) 11.0 (6.3-12.8) Median OS (95% CI), mo NR (16.3-NR) NR (21.3-NR) 16.3 (10.8-NR) [85]Open in a new tab a ^a 9 patients had missing TP53 status. ^b Defined as CR+PR + marrow CR+SD with HI. HI, hematologic improvement. Image: graphic file with name hs9-6-1-g045.jpg [86]Open in a new tab Summary/Conclusion: Magrolimab + AZA was well tolerated with promising efficacy in pts with untreated HR MDS, including those with TP53-mutated and –wild-type disease. A phase 3 trial of magrolimab/placebo + AZA (ENHANCE: [87]NCT04313881) is ongoing. S167: PREDICTION OF RELAPSE AFTER ALLOGENEIC STEM CELL TRANSPLANTATION USING INDIVIDUALIZED MEASURABLE RESIDUAL DISEASE MARKERS; THE PROSPECTIVE NORDIC STUDY NMDSG14B M. Tobiasson^1,*, T. Pandzic^2, J. Illman^3, L. Nilsson^4, S. Weström^2, K. Sollander^2, E. Ejerblad^5, A. Olsnes Kittang^6, G. Olesen^7, O. Werlenius^8, A. Björklund^9, J. Wiggh^1, C. lindholm^1, F. Lorentz^10, B. Rasmussen^11, J. Cammenga^12, D. Weber^13, D. Grönnås^14, M. Dimitriou^15, S. Kytölä^16, G. Walldin^15, P. Ljungman^9, K. Groenbeck^17, S. Mielke^9, S. E. Jacobsen^15, F. Ebeling^3, L. Cavelier^2, L. Smidstrup Friis^17, I. Dybedal^18, E. Hellström-Lindberg^1 ^1Hematology, Karolinska University Hospital, Stockholm; ^2Department of Immunology, Genetics and Pathology, Science for life, Uppsala, Sweden; ^3Hematology, Helsinki University Hospital, Helsinki, Finland; ^4Hematology, Lund University Hospital, Lund; ^5Hematology, Akademiska sjukhuset, Uppsala, Sweden; ^6Hematology, Haukeland University Hospital, Bergen, Norway; ^7Hematology, Arhus University Hospital, Århus, Denmark; ^8Hematology, Sahlgrenska University Hospital, Gothenburg; ^9Centre for allogeneic stem cell transplantation, Karolinska University Hospital, Stockholm; ^10Hematology, Norrlands University Hospital, Umeå; ^11Hematology, Örebro University Hospital, Örebro; ^12Hematology, Universitetssjukhuset, Linköping, Sweden; ^13Hematology, Odense University Hospital, Odense, Denmark; ^14Institution for environmental medicine; ^15Centre for hematology and regenerative medicine, Institution for medicine, Huddinge, Karolinska Institute, Stockholm, Sweden; ^16HUS Diagnostic Center, HUSLAB, Helsinki University Hospital, Helsinki, Finland; ^17Hematology, Rigshospitalet, Copenhagen, Denmark; ^18Hematology, Oslo University Hospital, Oslo, Norway Background: One third of patients with myelodysplastic syndrome (MDS) relapse after allogeneic stem cell transplantation (HCT). Early detection of impending relapse would enable pre-emptive treatment and potentially reduce relapse risk but is limited by the lack of sensitive markers for measurable residual disease (MRD). We developed a pipeline where patient-specific mutations, as determined by a myeloid next generation sequencing (NGS) panel are tracked using digital droplet PCR (ddPCR). Aims: To evaluate if personalized MRD detection by ddPCR can predict clinical relapse earlier than conventional methods. Methods: The prospective study ([88]NCT02872662) enrolled patients with MDS, MDS/MPN or MDS-AML with < 30% marrow blasts undergoing HCT. Patients were included before HCT, and serial bone marrow (BM) samples were collected every third month post-HCT for 2 years. Peripheral blood (PB) samples were collected monthly. MRD results were not available for the treating physician. Results: We screened 286 pts between 2016 and 2020, whereof 20 were excluded mainly due to lack of genetic aberration or no HCT performed. 266 pts were included from 12 HCT centers. Median age was 64 (18-78) years and 59% were male. Myeloid panel NGS screening identified a median of 2 (0-9) mutations. The most common mutations were TET2 (n=85), ASXL1 (n=73) and SRSF2 (n=59). Median time of follow up was 886 (4-1934) days. Sixty pts relapsed after a median of 189 (53-1281) days and 46 died due to non-relapse mortality after a median of 121 (4-1036) days. Remaining pts (n=160) were in continuous complete remission (CCR) after a median follow-up of 1053 (479-1934) days. Estimated 1 and 2y overall survival was 79%, and 71%, respectively, while estimated 1 and 2y relapse-free survival (RFS) was 75% and 66%, respectively. MRD data was missing in 46 pts; no post-HCT samples available (n=15), no mutation detected (n=14) and difficulties to design ddPCR primers (n=11). 221 pts were available for MRD analysis with a median number of 4 (0-13) and 5 (0-23) samples from BM and PB, respectively. Of 53 clinical relapses with MRD results available, 42 were preceded by pos MRD (>0.1%) with a median of 70 (range 20-425) days between first pos MRD and clinical relapse. For the 11 remaining pts, 8 were inadequately sampled with a median time of 189 (82-397) days between last sampling and clinical relapse. One patient had an extramedullary relapse only. Of 31 pts who died without relapse, 19 were consistently MRD neg, while 5 were borderline positive (MRD > 0.1% and <0.5%) during the first 100 days but negative thereafter. Four MRD+ patients died without clinical relapse. Three pts were initially MRD+ but turned negative, all of which had chronic GVHD (cGVHD). Of 136 CCR patients, 94 were consistently MRD neg; 26 were borderline pos (MRD > 0.1% and <0.5%) during the first 100 days followed by neg samples; 16 were MRD positive (either > 0.5% during the first 100d or > 0.1% after 100d) of which 10 had a transition from pos to neg samples (all had cGVHD); one patient was treated for a molecular relapse detected by clinical routine method (FISH) and five patients were MRD positive at time of last follow-up. MRD used as a time-dependent co-variate was negatively associated with RFS (HR 7.1, p<0.01). Estimated cumulative incidence of relapse and non-relapse mortality 2y after pos MRD was 60% and 7% respectively (see figure). Image: graphic file with name hs9-6-1-g046.jpg [89]Open in a new tab Summary/Conclusion: We report the development of a highly functional personalized MRD pipeline based on patient-specific mutations showing a high sensitivity to predict relapse and relapse-free survival. S168: ERYTHROPOIETIN STIMULATION AGENTS SIGNIFICANTLY IMPROVES OUTCOME IN LOWER RISK MDS. H. Garelius^1,*, A. Smith^2, T. Bagguley^2, A. Taylor^2, P. Fenaux^3, D. Bowen^4, A. Symeonidis^5, M. Mittelmann^6, R. Stauder^7, J. Čermák^8, G. Sanz^9, S. Langemeijer^10, L. Malcovati^11, U. Germing^12, R. Itzykson^13, A. Guerci-Bresler^14, D. Culligan^15, I. Kotsianidis^16, K. Koinig Mag^17, C. van Marrewijk^18, S. Crouch^2, T. de Witte^19, E. Hellström-Lindberg^20 ^1Section of Hematology, Specialist Medicine, Sahlgrenska University hospital, Göteborg, Sweden; ^2Department of Health Sciences, University of York, York, United Kingdom; ^3Service d’Hematologie, Hôpital Saint-Loius, Assistance Publique des Hopitaux de Paris (AP-HP) and Universite Paris; ^7, Paris, France; ^4St.James’s Institute of Oncology, Leeds Teaching Hospitals, Leeds, United Kingdom; ^5Department of Medicine, Division of hematology, University of Patras Medical Scholl, Patras, Greece; ^6Department of Medicine, Tel Aviv Sourasky (Ichilov) Medical Center and Sackler Medical Faculty, Tel Aviv University, Tel Aviv, Israel; ^7Department of Internal Medicine V (Haematology and Oncology), Innsbruck Medical University, Innsbruck, Austria; ^8Dep. of Clinical Hematology, Institute of Hematology & Blood Transfusion, Prague, Czechia; ^9Department of Haematology, Hospital Universitario y Politécnico La Fe, Valencia, Spain; ^10Department of Hematolog, Radboud university medical center, Nijmegen, Netherlands; ^11Department of Hematology Oncology, Fondazione IRCCS Policlinico San Matteo,University of Pavia, Pavia, Italy; ^12Department of Haematology, Oncology and Clinical Immunology, Universitätsklinik Düsseldorf, Düsseldorf, Germany; ^13Service d’Hématologie, Hôpital Saint-Louis, Assistance Publique des Hôpitaux de Paris (AP-HP) and Université Paris; ^7, Pari, Paris; ^14Service d’Hématologie, Centre Hospitalier Universitaire Brabois Vandoeuvre, Nancy, France; ^15Department of Haematology, Aberdeen Royal Infirmary, Aberdeen, United Kingdom; ^16Dep. of Hematology, Democritus University of Thrace Medical School, University Hospital of Alexandroupolis, Alexandroupolis, Greece; ^17Department of Internal Medicine V (Hematology and Oncology), Medical University Innsbruc, Innsbruck, Austria; ^18Department of Haematology, Radboud university medical center; ^19Department of Tumor Immunology - Nijmegen Center for Molecular Life Sciences, Radboud university medical cente, Nijmegen, Netherlands; ^20Department of Medicine, Div. Hematology, Karolinska Institutet, Stockholm, Sweden Background: The EUMDS Registry started in 2008 as a prospective, non-interventional longitudinal study, enrolling newly diagnosed patients with IPSS low or intermediate-1 MDS from 16 European countries and Israel. Aims: The aim of the present analysis was to see how treatment with or without Erythropoietin Stimulating Agents (ESAs) and/or red blood cell transfusions (RBCT) impact overall survival (OS) and quality of life (QoL). Methods: Patient management was recorded electronically every 6 months (“visit”) in a central database, including treatment, transfusions, blood values, and health related quality of life (HRQoL) using the EQ-5D 3-Level index and Visual Analog Scale (VAS). Patients were eligible to be included in the analyses if their hemoglobin was recorded as less than <10 g/dl at a visit. To overcome potential confounding by non-random allocation of ESA treatment, propensity score matching was performed to ensure that treated and untreated patients had similar characteristics. Only patients with comparable propensity scores were included in the analyses to estimate the effects of ESA treatment on outcomes using standard time to event analyses; OS was estimated from the first visit a Hb value of <10g/dl was recorded. OS was examined for patients treated with ESA stratified by their transfusion status prior to commencing ESA treatment (no RBCT, <4 units, ≥4 units). Patients were separated into 4 groups at each clinical visit, depending on the treatment received in the interval leading up to that visit; no ESA nor RBCT, ESA only, ESA and RBCT and RBCT only. HRQoL at each visit according to the treatment status was summarized for patients who had completed a questionnaire at visit 1 and 2; mean values were examined by treatment group. Results: Of 2562 patients registered by November 2021, 2448 were diagnosed before July 2019 and included in the analysis; these patients were divided into two groups: ESA untreated (n=1265) and ESA treated (n=1183). Patients whose Hb remained above 10g/dl were excluded leaving 529 untreated patients and 749 ESA treated; after propensity score matching was applied two comparable groups were produced: ESA untreated (n= 426) and ESA treated (n= 742). Median OS from reaching the eligibility criteria in the ESA treated vs untreated groups were 44.9 and 34.8 months respectively (Fig 1a), giving a clear survival advantage to the ESA-treated group. (p<0.003). In the ESA-treated group, OS was poorer in those who had been transfused prior to commencing ESA (Fig 1b, p<0.001). Fig 1c shows the number of patients at each visit who had been treated with transfusions or ESA; 647/1278 had received neither at visit 1, the figure shows the “flow” of patients by treatment for the first 6 visits. HRQoL was examined for the 695 patients who had completed a questionnaire at both visit 1 and 2 up to visit 6; differences were seen by treatment (Fig 1d). Patients who had received no treatment reported, on average, the highest mean HRQoL, in contrast, patients who had RBCT had the lowest (p<0.001). Image: graphic file with name hs9-6-1-g047.jpg [90]Open in a new tab Summary/Conclusion: This unique large prospective registry study clearly shows a significant survival advantage for lower-risk MDS patients exposed to ESA treatment at onset of anemia (Hb <10g/dL) but before onset of transfusion therapy, strongly supporting recommendations to start ESA treatment early. The effect on patients with an early transfusion need warrants further studies. Moreover, ESA exposure is associated with maintained QoL, while RBCT development with or without ESA exposure is associated with significantly deterioration in QoL. S169: CLINICAL AND MOLECULAR MARKERS FOR PREDICTING RESPONSE TO ROMIPLOSTIM TREATMENT IN LOWER-RISK MYELODYSPLASTIC SYNDROMES A. S. Kubasch^1 2 3,*, A. Giagounidis^2 3 4, G. Metzgeroth^5, A. Jonasova^6, R. Herbst^7, J. M. T. Diaz^8, B. De Renzis^9, K. S. Götze^2 3 10, M.-L. Huetter-Kroenke^11, M.-P. Gourin^12, B. Slama^13, S. Dimicoli-Salazar^14, P. Cony-Makhoul^15, K. Laribi^16, S. Park^17, K. Jersemann^18, D. Schipp^19, K. H. Metzeler^20, O. Tiebel^21, K. Sockel^2 22, S. Gloaguen^2 3, A. Mies^22, F. Chermat^23, C. Thiede^22, R. Sapena^23, R. F. Schlenk^24 25, P. Fenaux^3 23 26, U. Platzbecker^2 3 20, L. Ades^3 23 26 ^1Department of Hematology, Cellular Therapy and Hemostaseology, University Hospital Leipzig; ^2German MDS Study Group (D-MDS); ^3The European Myelodysplastic Syndromes Cooperative Group (EMSCO), Leipzig; ^4Department of Oncology, Hematology and Palliative Care, Marien Hospital, Düsseldorf; ^5Department of Hematology and Oncology, University Medical Centre, Mannheim, Germany; ^61st Medical Department - Hematology, General Hospital, Prague, Czechia; ^7Medizinische Klinik III, Klinikum Chemnitz, Chemnitz, Germany; ^8Department of Hematology and Oncology, CHU de Poitiers, Poitiers; ^9Service d’Hématologie Clinique Adulte, Clermont Ferrand, France; ^10Department of Medicine III, Technical University of Munich, Munich; ^11Department of Hematology, Oncology, and Tumor Immunology, Charité-Universitätsmedizin Berlin, Berlin, Germany; ^12CHU Limoges, Limoges; ^13Service d’Hématologie, Centre Hospitalier d’Avignon, Avignon; ^14University Hospital Bordeaux, Pessac; ^15Centre Hospitalier Annecy-Genevois, Pringy; ^16Centre Hospitalier Du Mans, Le Mans; ^17Department of Hematology, CHU Grenoble, Grenoble, France; ^18GWT-TUD GmbH, Dresden; ^19DS-Statistics, Rosenthal-Bielatal; ^20Department of Hematology, Cellular Therapy and Hemostaseology, Leipzig University Hospital, Leipzig; ^21Institute of Clinical Chemistry and Laboratory Medicine, Medical Faculty, Technical University Dresden; ^22Department of Internal Medicine I, University Hospital Carl Gustav Carus, Technical University Dresden, Dresden, Germany; ^23Groupe Francophone des Myélodysplasies, Paris, France; ^24Department of Internal Medicine V, University Hospital of Heidelberg; ^25NCT-clinical trials office, German Cancer Research Center, Heidelberg, Germany; ^26Hématologie Clinique, Hôpital Saint-Louis, Paris, France Background: In about half of patients with lower-risk (LR) myelodysplastic syndromes (MDS), thrombocytopenia is present at the time of diagnosis, being associated with shortened survival and a higher risk of progression to acute myeloid leukemia (AML). Romiplostim (ROM), a thrombopoietin receptor agonist (TPO-RA), has shown safety and clinical efficacy in prospective trials in LR-MDS. Post-hoc analyses have demonstrated hematologic improvement of platelets (HI-P) after ROM treatment contingent on endogenous thrombopoietin (TPO) levels and platelet transfusion events (PTE) (Sekeres et al. BJH 2014). Aims: The prospective EUROPE multicenter phase 2 trial ([91]NCT02335268) investigated the predictive value of biomarkers like endogenous TPO levels, PTE and molecular markers on the clinical efficacy of single-agent ROM treatment within the `European Myelodysplastic Syndromes Cooperative Group` (EMSCO) network. Patients with IPSS low or intermediate 1 risk were eligible if baseline bone marrow blast count was <5% (central morphology) and platelet count was ≤30 Gpt/L or ≤50 Gpt/L in case of a bleeding history. Methods: According to a previously published model of response to TPO-RA (Sekeres et al. BJH 2014), patients were assigned into two different cohorts at the time of screening based on previous PTE and centrally assessed TPO serum levels (cohort A: TPO<500 ng/l and PTE<6 units/past year; cohort B: TPO>500 ng/l, and/or PTE^36 units/past year). The primary efficacy endpoint was the rate of HI-P according to IWG 2006 criteria lasting for ^38 weeks. ROM was initiated at a dose of 750 μg weekly by subcutaneous injection, the dose was adjusted based on the patient`s platelet counts. Results: From 2015 to 2019, a total of 77 patients were included at 29 different trial sites in Germany, France and the Czech Republic. Regarding the primary endpoint, 32 out of 77 (42%) responded (HI-P) with a numerically higher response rate in cohort A (47%, n=24) vs. cohort B (31%, n=8) (p=0.2953). At 16 weeks of ROM treatment, three (4%) and seven (9%) patients had additional neutrophil (HI-N) and erythroid (HI-E) responses, respectively. None of the patients achieved trilineage responses (HI-P, HI-E and HI-N). Median duration of response was significantly longer for patients in cohort A (351 days) compared to cohort B (315 days) (p=0.006, log-rank-test). Mutated SRSF2 was significantly more frequent in responders (41%) compared to non-responders (16%) (p=0.018, Fisher’s exact test). In patients with an SRSF2 mutation, the probability to achieve HI-P was 65% compared to 33% inpatients with SRSF2 wildtype (Figure 1A). Comparing responders vs. non-responders, we found no significant changes of variant allelic burden of variants detected pre- and post-ROM (Figure 1B). Finally, we developed a response prediction model to ROM therapy with the aim to improve personalized patient stratification in the future. The percentage of correctly predicted HI-P was highest for the model, which included the variables platelet count, SRSF2 mutation status and the hemoglobin level using the threshold of 11.4 g/dl and resulted in an overall accuracy of 70 % for a correct ROM response prediction (Figure 1A). Image: graphic file with name hs9-6-1-g048.jpg [92]Open in a new tab Summary/Conclusion: In conclusion, this prospective study confirms the efficacy and overall safety of ROM in this subgroup of LR-MDS patients with thrombocytopenia. To avoid overfitting of variables and to confirm our results, the here presented response prediction model needs to be validated in an external independent cohort. * U.P. and L.A. contributed equally to this study as senior authors S170: DYNAMIC INTERPLAY BETWEEN TUMOR AND MICRO-ENVIRONMENT DURING MYELOMA DISEASE PROGRESSION M. C. Köse^1,*, I. Bergiers^2, M. Malfait^3, B. Heidrich^4, D. De Maeyer^2, N. Fourneau^2, B. Verbist^2, J. Van Houdt^2, G. Vanhoof^2, R. Verona^4, M. Delforge^5, J. Depaus^6, N. Meuleman^7, J. Van Droogenbroeck^8, P. Vlummens^9, C. J. Heuck^4, N. Bahlis^10, J. Caers^1, T. Casneuf^2 ^1Laboratory of Hematology, University of Liege, Liege; ^2Janssen Research & Development, Beerse; ^3Department of Applied Mathematics, Computer Science and Statistics, University of Ghent, Ghent, Belgium; ^4Janssen Research & Development, Spring House, United States of America; ^5University Hospital Leuven, Leuven; ^6Department of Hematology, CHU UCL Namur, Yvoir; ^7Service d’Hématologie, Université Libre, Brussels; ^8Departmentof Haematology, AZ Sint-Jan Brugge-Oostende AV, Brugge; ^9Department of Clinical Hematology, Ghent University Hospital, Ghent, Belgium; ^10Department of Hematology and Oncology, University of Calgary, Calgary, Canada Background: Multiple Myeloma (MM) is an incurable plasma cell (PC) malignancy that evolves from two premalignant stages: Monoclonal Gammopathy of Undetermined Significance (MGUS) and Smoldering Multiple Myeloma (SMM). The disease progression has been characterized to be driven by intrinsic genomic events in the myeloma cells and by gradual dysregulation of the immune system. Aims: We investigated how the interplay between tumor cells with their microenvironment and the underlying complex and dynamic immune biology evolve during this process. Methods: Single cell multi-omics profiling, including RNA, B-cell receptor (BCR) and antibody barcode-tagged 10x sequencing, was conducted on human bone marrow (BM) aspirates collected at 6 Belgian centers from 4 cohorts: 31 healthy elderly and 28 MGUS, 32 SMM and 32 newly diagnosed MM. Mononuclear cell isolation, freezing and transport to central facilities was optimizedand data were integrated and filtered using Scanpy and Scirpy. The main immune cell types were identified from the RNA and antibody data using SingleR. Further functional subtyping was done using Leiden clustering. Differential pathway expression analysis was performed with Muscat and FGSEA. Results: From the tumor cell transcriptomes, our analyses confirmed the previously documented myeloma molecular hallmarks, such as MYC and IFN-a signaling, cell proliferation, energy metabolism and oxidative phosphorylation. Evidence was found for transcriptomic similarities and within-and between-patient malignant PC transcriptomic heterogeneity, as well as the existence of multiple transcriptomic clones in several patients. We observed a positive correlation between the antigen processing mechanism in the PCs with IFN response, suggesting that this mechanism associates with initiation of the immune recognition and activation against the tumor.The gradually increasing differential gene expression was also observed in the immune microenvironment: dysregulation of signaling pathways initiates early in MGUS and spreads throughout the various cell types surrounding the tumor cells. Cell population shifts were also found. In the CD1C+ DCs, that play a role in cancer immune control, a functional shift was observed that correlated with disease progression towards a more mature and antigen presenting phenotype with higher levels of CD83, HBEGF, MCL1 and CXCL16 as well as increased TNF-a pathway. Similarly, a shift was observed in the macrophage population, toward M1 state showing high IFN response along with expression of MS4A4A, STAT1, TNFSF13B and TRAIL in more severe disease. Interestingly, in the CD8+ T cells, we detected a pre-dysfunctional subpopulation with high expression of GZMK, activation markers CD69, CCL4, CXCR4 and genes associated with T cell pre-dysfunctionality NR4A2, RGS1, TOX and TIGIT, that was found to be associated with progression (Figure). In the CD4+ cytotoxic T cells, a proportion change was observed with more severe disease. Image: graphic file with name hs9-6-1-g049.jpg [93]Open in a new tab Summary/Conclusion: With the atlas of healthy, precursor and active MM patient BM samples, we generated a comprehensive and granular view of the various cell types involved in disease progression and provide evidence for early and gradually increasing immune dysregulation and activation of oncogenic driver pathways. Our data evidence the co-divergence and reciprocal stimulation of transcriptomes of tumor and microenvironment and support the postulation of microenvironment as a central modulator of cancer cell growth, survival and metastasis. S171: MOLECULAR CLUSTERS OF IGM MONOCLONAL GAMMOPATHIES PRESENT DISTINCT BIOLOGIC, IMMUNE AND METABOLIC FEATURES P. Mondello^1,*, J. Paludo^1, J. Novak^1, K. Wenzl^1, S. Jalali^1, J. Krull^1, E. Braggio^2, S. Dasari^3, M Manske^1, J. Abeykoon^1, S. Vivekananda^1, P. Kapoor^1, A. Paulus^4, C. Reeder^2, S. Ailawadhi^4, A. Chanan-Khan^4, R. Kyle^1, M. Gertz^1, Z.-Z. Yang^1, A. Novak^1, S. Ansell^1 ^1Medicine, Mayo Clinic, Rochester; ^2Medicine, Mayo Clinic, Phoenix; ^3Bioinformatics, Mayo Clinic, Rochester; ^4Medicine, Mayo Clinic, Jacksonville, United States of America Background: IgM MGUS and Waldenstrom Macroglobulinemia (WM) represent a wide range of conditions whose management varies from observation to immunochemotherapy. The current classification relies solely on clinical features and does not explain the heterogeneity that exists within each of these conditions. Aims: To shed light on the biology that may account for the clinical differences, we performed the first comprehensive multi-omics analysis of IgM monoclonal gammopathies. Methods: We used bone marrow (BM) CD19^+CD138^+ sorted cells and matched BM plasma from 32 pts (7 IgM MGUS, 25 WM) and 5 healthy controls to perform whole exome sequencing, RNA-seq, proteomic and metabolomic analysis. 7 matched WM and 4 IgM MGUS pts were also evaluated using mass cytometry (CyTOF). Results: Applying principal component analysis to gene expression profiling, most of WM pts clustered together, while a small subset of them grouped separately with MGUS pts, suggesting a biologic dichotomy within WM. The controls formed a group distinct from most WM and MGUS pts. Fig1A We then applied a non-negative matrix factorization consensus clustering to the gene expression data and identified three robust clusters. Cluster 1 (C1) included only pts with WM, cluster 2 (C2) included pts with both WM and MGUS, and cluster 3 (C3) included all normal controls as well as a small number of WM and MGUS pts. Fig1B-C When mutations commonly identified in WM were analyzed, there was no difference among the three groups (excluding controls) in mutation burden of MYD88 L265P and CXCR4. Interestingly, aberrant expression of TNFAIP3 was a distinct feature of C1 as deletion of 6q (which encodes for TNFAIP3) and TNFAIP3 mutations were each significantly enriched in C1 (47%) compared to C2 (0%) and C3 (20%; p=0.04). Individual clusters associated with specific transcriptional signatures and clinical features. While C1 displayed enrichment of RNA processing, downregulation of inflammatory pathways and aggressive clinical behavior, C2 showed increased inflammatory signaling and senescence with indolent clinical behavior. C3 had intermediate features with combined proliferative and antigen response signatures. Fig1D In accordance with transcriptomics, the proteomic hallmark of C1 was upregulation of proteins involved in proliferation (eg AKT, MAPK) and downregulation of inflammatory proteins (eg IL4, IL10) while the opposite was observed in C2. Once more, C3 confirmed intermediate features with combined upregulation of proliferation and inflammatory proteins. The metabolism was rewired towards mitochondrial anabolism in C1 and C3, while towards glycolysis in C2. Accordingly, C1 and C3 showed undetectable concentrations of 3-hydroxybutyric acid as opposed to C2 which had increased levels of lactic acid, as end products of fatty acid oxidation and glycolysis respectively. Next, we explored whether C1 and C2 displayed a distinct immune profile. tSNE analysis showed that CD4^+ T cells were more abundant in C2 compared to C1 while the opposite was observed for CD8^+ T cells. Among CD4^+ T cells, activated follicular helper (T[FH]; p=0.02) and regulatory (T[reg]; p=0.008) cells were predominantly expressed in C2. In contrast, C1 showed a higher expression of senescent T effector memory (T[EM]) cells. Fig1E In support of this, SPADE clustering analysis identified three clusters including T[FH], T[reg] and T[EM] cells. Image: graphic file with name hs9-6-1-g050.jpg [94]Open in a new tab Summary/Conclusion: We have identified three molecular clusters in IgM monoclonal gammopathies with distinct clinical, proteomic, metabolomic and immune features, suggesting a potential biologic classification that may have therapeutic implications. S172: FUNCTIONAL SUBSETS OF PLASMA CELLS ASSOCIATED WITH AMYLOID PRODUCTION AND X. Wang^1 2 3,*, H. Han^1 3, J. Sun^3 4, Q. Wang^2, X.-M. Gao^1 3, K.-N. Shen^1 3, L. Zhang^1 3, Y. Zhao^2, X.-X. Cao^1 3, M. Qian^5, Y. Chen^2, J. Li^1 3 ^1Department of Hematology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College; ^2The State Key Laboratory of Medical Molecular Biology, Department of Biochemistry and Molecular Biology, Institute of Basic Medical Sciences, School of Basic Medicine, Chinese Academy of Medical Sciences and Peking Union Medical College; ^3State Key Laboratory of Complex Severe and Rare Diseases; ^4Department of Pathology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College; ^5School of Mathematical Sciences, Peking University, Beijing, China Background: Light chain amyloidosis (AL amyloidosis, AL) is a life-threatening plasma cell dyscrasia characterized by misfolded monoclonal immunoglobulin light chain production by pathogenic bone marrow plasma cells (BMPCs), leading to irreversible damage in multiple organs. Until now, the pathogenic BMPCs in AL haven’t been elucidated, let alone the functional roles of BMPC subsets. Aims: We aimed to identify intra- and inter-individual heterogeneities in BMPCs, especially those related to AL development, light chain production, organ tropism, and chemotherapy response. Methods: Here, we conducted single-cell RNA sequencing and image flow cytometry analysis of BMPCs from patients with AL (n=3), compared with that from monoclonal gammopathy of undetermined significance (MGUS) (n=2) and healthy controls (n=21). All the subjects provided informed consent. Results: We identified 7 functional subsets of BMPCs in AL, MGUS or healthy controls (Fig. 1a). These subsets had over-lapping but distinct functions, including DNA repair, cell proliferation, drug response, osteoclast differentiation, and immunoglobulin production (Fig. 1b). Subsets enriched in AL showed up-regulation of amyloidosis-associated genes (Fig. 1c), such as the amyloid-beta binding protein-encoding Apolipoprotein E (APOE), and showed plasmablastic morphology (Fig. 1d). Subsets defined by aberrant light chain production up-regulated the pathways related to neutrophil degranulation, transportation to and modifications in the Golgi apparatus, and asparagine N-linked protein glycosylation. High expression of Cyclin D1 (CCND1), CD79A, and V-Set Pre-B Cell Surrogate Light Chain 3 (VPREB3) were observed in the predominant subset of AL predicted sensitive to venetoclax, while Cyclin D2 (CCND2), S100 Calcium Binding Protein A6 (S100A6), Cystatin C (CST3) were up-regulated in that of AL predicted resistant (Fig. 1e). In an independent cohort of bulk RNA sequencing (n=29), clinical subgroups of patients with AL were defined by the proportion of functional subsets, including 2 major subgroups that were consistent with the aforementioned AL with differential sensitivity to venetoclax (Fig. 1f). Co-expression of CCND1, CD79A, and VPREB3, and a larger predominant subset, were in the venetoclax-sensitive subgroup. The up-regulation of CCND2 and amyloidosis-associated genes, including S100A6 and CST3, were in the venetoclax-resistant subgroup (Fig. 1g). Image: graphic file with name hs9-6-1-g051.jpg [95]Open in a new tab Summary/Conclusion: These functional subsets of BMPCs provide mechanistic insights into AL pathogenesis, light chain production, and chemotherapy response. These results suggest potential avenues for the exploration of clinical subgroups among AL patients. S173: T-CELL ACTIVATION AND MYELOMA CELL KILLING CONFIRM THE MODE OF ACTION OF RG6234, A NOVEL GPRC5D T-CELL ENGAGING BISPECIFIC ANTIBODY, IN PATIENTS WITH RELAPSED/REFRACTORY MULTIPLE MYELOMA I. Dekhtiarenko^1,*, I. Lelios^2, J. Attig^2, N. Sleiman^2, D. Lazzaro^2, I. Clausen^2, N. Gräfe^3, H.-J. Helms^2, E. Schindler^2, S. Belli^2, T. Fauti^1, J. Eckmann^3, P. Umana^1, W. Jacob^3, M. Schneider^2, C. Hasselbalch Riley^4, M. Hutchings^4, S.-S. Yoon^5, Y Koh^5, S. Manier^6, T. Facon^6, S. J. Harrison^7, J. Er^7, F. Volzone^8, A. Pinto^8, C. Montes^9, E. M. Ocio^9, A. Alfonso-Pierola^10, P. Rodríguez Otero^10, F. Offner^11, A. Guidetti^12, P. Corradini^12, C. Titouan^13, C. Hulin^13, C. Touzeau^14, P. Moreau^14, R. Popat^15, S. Leong^15, R. Mazza^16, C. Carlo-Stella^16, A.-M. E. Bröske^3 ^1Roche Pharma Research and Early Development, Roche Innovation Center Zurich, Zurich; ^2Roche Pharma Research and Early Development, Roche Innovation Center Basel, Basel, Switzerland; ^3Roche Pharma Research and Early Development, Roche Innovation Center Munich, Penzberg, Germany; ^4Rigshospitalet, Copenhagen, Denmark; ^5Seoul National University College of Medicine, Seoul, South Korea; ^6Lille University Hospital, Lille, France; ^7Peter MacCallum Cancer Center and Royal Melbourne Hospital, and Sir Peter MacCallum Department of Oncology, University of Melbourne, Melbourne, Australia; ^8Istituto Nazionale dei Tumori, Fondazione Pascale, IRCCS, Napoli, Italy; ^9Hospital Universitario Marques de Valdecilla (IDIVAL), Universidad de Cantabria, Santander; ^10Clinica Universidad de Navarra, Navarra, Spain; ^11Universitair Ziekenhuis Gent, Gent, Belgium; ^12Istituto Nazionale dei Tumori, Milano, Italy; ^13CHU de Bordeaux, Bordeaux; ^14CHU de Nantes, Nantes, France; ^15University College London Hospitals NHS Foundation Trust, London, United Kingdom; ^16Department of Biomedical Sciences, Humanitas University and Department of Oncology and Hematology, IRCCS Humanitas Research Hospital, Milano, Italy Background: RG6234 is a novel T-cell engaging bispecific antibody targeting G protein-coupled receptor 5D (GPRC5D) with a unique 2:1 format. GPRC5D is highly expressed on multiple myeloma (MM) cells and concurrent binding of RG6234 to GPRC5D and CD3 on T cells results in immunological synapse formation and potent T-cell directed tumor cell killing. An ongoing Phase 1 dose-escalation study ([96]NCT04557150) is investigating the safety, clinical activity, pharmacodynamics (PD), and pharmacokinetics of RG6234 monotherapy in patients (pts) with relapsed/refractory MM (RRMM). Clinical activity was observed during dose escalation, and the safety profile was manageable (Riley et al. EHA 2022). Aims: Here, we present preliminary clinical biomarker data highlighting PD effects after intravenous (IV) administration that confirm the mechanism of action and high potency of RG6234. Methods: Exploratory biomarker analyses included data from pts treated with RG6234 doses ranging from 0.006mg to 4.8mg in dose escalation. RG6234 was administered as an IV infusion under a step-up dosing regimen, reaching the target dose not later than 2 weeks after the priming dose. Peripheral biomarkers were evaluated using whole blood flow cytometry (n=28), plasma cytokine Protein Simple ELLA (n=33), and plasma sBCMA Protein Simple ELLA (n=26). MM cells were assessed at baseline and on-treatment by bone marrow (BM) aspirate flow cytometry and by BM biopsy CD138/CD8 immunohistochemistry. Informed consent was obtained from participating pts. The clinical cut-off date for the current analysis was January 31, 2022. Results: PD changes were observed in peripheral blood at all tested doses. Cytokines (IFNg, TNFa, CXCL10, IL6, IL10, IL2, IL8) and sCD25 peaked at 4 to 24 hours (h) following the first administration, while cytokine peak magnitudes decreased at subsequent administrations. Cytokine release was followed by transient reduction of circulating T cells at 4 h after infusion, with partial recovery of peripheral T-cell counts by Day 8 after first administration. Elevation of sCD25 and IFNg in plasma (~3.2 and 33 median fold change from baseline, respectively) together with increase in T-cell proliferation (~4 median fold increase of Ki67+CD8+ T-cells) within 72 h post first infusion indicated T-cell activation. Analysis of a limited number of paired baseline and on-treatment BM biopsies (n=13) revealed that the density of CD8+ tumor-infiltrating T cells increased upon treatment in responders, indicating T-cell recruitment towards the tumor. RG6234 induced rapid depletion of MM cells, as demonstrated by a decrease of sBCMA in the plasma of responding pts already 8 days after the first administration (median 33.5% reduction from baseline in responding pts; n=15). Moreover, at the end of Cycle 1, the majority of pts (14/15) had <1% of MM cells in BM based on flow cytometry readout. GPRC5D expression was detected at baseline in all pts with evaluable bone marrow aspirate and >20 detectable MM cells (n=16). Updated exploratory biomarker data will be presented. Summary/Conclusion: Cytokine release, T-cell activation, BM infiltration, and MM cell depletion are early PD changes seen after treatment with RG6234 and precede clinical responses. These PD changes indicate that RG6234 leads to T-cell engagement in the BM of pts with RRMM and clearly demonstrate rapid and effective T-cell mediated anti-MM activity. S174: HIGH LEVEL OF CIRCULATING TUMOUR DNA AT DIAGNOSIS CORRELATES WITH DISEASE SPREADING AND DEFINES MULTIPLE MYELOMA PATIENTS WITH POOR PROGNOSIS M. Martello^1,*, A. Poletti^1, D. Bezzi^2, E. Borsi^1, B. Taurisano^1, V. Solli^1, S. Armuzzi^1, I. Vigliotta^1, G. Mazzocchetti^1, I. Pistis^3, L. Pantani^3, S. Rocchi^1, K. Mancuso^1, P. Tacchetti^3, I. Rizzello^1, M. Cavo^1, E. Zamagni^1, C. Nanni^2, C. Terragna^3 ^1IRCCS - Azienda Ospedaliero Universitaria di Bologna - Department of Experimental, Diagnostic and Specialty Medicine - University of Bologna; ^2IRCCS - Azienda Ospedaliero Universitaria di Bologna - Department of Nuclear Medicine; ^3IRCCS - Azienda Ospedaliero Universitaria di Bologna, Bologna, Italy Background: Multiple Myeloma (MM) is a plasma cell (PC) disorder characterized by the presence of skeletal involvement at the time of diagnosis, as detected by MRI and/or FDG PET/CT, in most of the patients. The patchy nature of the disease is probably related to the ability of fitter clones to spread into peripheral blood reaching distant sites, where favourable microenvironment conditions might promote clones’ seeding. Recently, cell-free DNA (cfDNA) has been proven to resume the heterogeneity of spatially distributed clones. However, it has to be determined to which extent cfDNA correlates with disease distribution and its possible implications with patients’ outcome. Moreover, the potential of cfDNA to track the evolutionary dynamics and the heterogeneity of MM, possibly anticipating the emergence of therapy resistant residual cells, remains to be confirmed. Aims: Aim of this study is to quantitatively and qualitatively evaluate cfDNA at diagnosis and during follow-up in correlation with imaging data to possibly define the integration of this approach with molecular bone marrow (BM) and whole-body residual disease assessment. Methods: A total of 88 newly diagnosed MM patients were screened at baseline with 18F-FDG PET/CT, and molecularly assessed by Ultra Low Pass-Whole Genome Sequencing (ULP-WGS). In a subgroup of 22 patients, cfDNA was monitored monthly, whereas PET/CT was reassessed after induction therapy to evaluate metabolic tumor response. For each pts, ULP-WGS was used to characterize both the neoplastic PC clone(s) in the BM (gDNA) and the cfDNA from peripheral blood. Data were analysed by ichorCNA and Clonality R packages. Results: At diagnosis, the cfDNA tumor fraction (TF) was significantly lower as compared to gDNA TF [median (M) TF: 4.4 vs. 59.7%, respectively]. Nevertheless, high cfDNA TF levels (> 4.4% cfDNA TF values; range: 4.4-84.3%) correlated with high gDNA TF levels (>65.7% gDNA TF values; range: 65.7-96.7%). This observation was further confirmed by a significant correlation between cfDNA TF and the percentage of BM CD138/CD38 positive plasma cells (r=0.47; p<.0001). Interestingly, patients with high cfDNA TF at baseline were more likely to present with extramedullary disease (EMD) and a higher number of focal lesions, and also featured a more active tumour metabolism, as compared to pts with low TF (EMD 4/44=9% vs. 1/44=2.3%, p=ns; M n. PET lesions: 1.7 vs. 2.5, p= 0.003; SUVmax: 5.2 vs. 9.6, p = 0.01). Despite an overall concordance between cfDNA and BM genomic profiles (80/88=90.9%), those patients with high cfDNA TF showed more frequently evidence of spatial heterogeneity, as highlighted by a substantial divergence of copy number alterations profiles between cfDNA and gDNA (7/8=87.5% of pts with divergent profiles). Finally, high cfDNA TF at diagnosis predicted for poorer prognosis as compared with low cfDNA TF (PFS at 20 months: 67% vs. 86%, p=0.05; OS at 20 months: 90% vs. 100%, p=0.04). Interestingly, after bortezomib-based induction therapy, imaging data and cfDNA TF levels were concordant; notably, in 4/10 (40%) patients with median SUVmax 5.8 and >2 PET lesions, cfDNA TF was still detectable (>1% TF), thus corroborating a possible role for cfDNA in monitoring response to therapy. Summary/Conclusion: In conclusion, patients with high cfDNA TF displayed imaging data that overall suggested a higher propensity to a metastatic spread of the disease, which finally correlated with poorer prognosis. Future studies will be addressed to exploit the use of cfDNA in disease monitoring. Acknowledgements: AIRC IG2019, AIL BOLOGNA ODV