Sideroblasts Research Articles

A Public Neoantigen Produced By the SF3B1 K700E Mutation Is a Bona Fide T Cell Target in AML/MDS

Myelodysplastic syndrome (MDS) is a life-threatening disease in which malignant bone marrow cells fail to differentiate, resulting in anemia, cytopenia, and increased risk of progression to acute myeloid leukemia (AML). T cell-based therapies have great potential in treating AML/MDS, given that allogeneic stem cell transplantation (aSCT) can be curative in such settings. aSCT is associated with the graft-vs-leukemia (GvL) effect, in which T cells within the donor graft recognize and target antigens expressed by leukemia cells. Unfortunately, aSCT is also associated with considerable morbidity and mortality, as donor T cells may also recognize antigens expressed by non-malignant tissues and kill them, leading to graft-vs-host disease (GvHD). A T cell-based therapy that targets an AML/MDS-exclusive antigen has the potential to eliminate leukemic cells while sparing healthy tissues. Most known antigens expressed on AML/MDS are also expressed by healthy myeloid cells. An ideal T cell target would be 1.) only expressed by leukemia cells, including leukemic stem cells (LSCs), 2.) expressed by all leukemia cells, and 3.) required for continued leukemia persistence. Somatic mutations in genes encoding spliceosomal machinery are common in MDS and have the potential to generate such neoantigens. In particular, a conserved missense mutation encoding a K700E substitution in SF3B1 is present in about 20% of patients with MDS, and approximately 80% of patients with MDS with ringed sideroblasts (MDS-RS). Moreover, SF3B1 K700E-mutant MDS has been proposed to be a distinct disease entity in which the SF3B1 mutation is a necessary, disease-initiating event. Further, SF3B1 K700E mutations are broadly detected across the spectrum of myeloid neoplasms from clonal hematopoiesis (CH) to relapsed/refractory AML and are associated with worse overall survival (OS) in the latter. We demonstrate that the SF3B1 K700E mutant protein produces a leukemia-restricted T cell target. NetMHCpan4.1 binding algorithm predicted that a nonameric SF3B1 K700E-derived peptide could bind HLA-A*02:01, the most common HLA-A allele. A similar peptide derived from the wild type (WT) SF3B1 sequence was also predicted to bind HLA-A*02:01 with similar affinity. We identify potent TCRs from the T cell repertoires of healthy HLA-A*02:01 positive individuals, which effectively discriminate between WT and K700E peptides. We further show that the SF3B1 K700E: HLA-A*02:01 is a bona fide T cell antigen, as it is naturally processed by the proteasome and presented on the surface of SF3B1 K700E mutant cell lines and primary patient samples. Primary human CD8 T cells that were retrovirally engineered to express an SF3B1 K700E:HLA-A2 specific TCR are activated by, and kill SF3B1 K700E mutant cancer cells, while sparing SF3B1 WT cells. Antigen dose response experiments demonstrate that this TCR has an approximately 4 order of magnitude specificity for the K700E peptide over the WT peptide, and amino acid substitution scanning experiments tiling the mutant peptide suggested no other naturally occurring HLA-A*02:01 restricted peptides from the human proteome that could activate the TCR. The EC 50 of the K700E:A0201-specific TCR in peptide pulsing experiments was <1nM, which is comparable to the potency observed in viral antigen-specific TCRs. In summary, we report that the SF3B1 K700E mutation yields an HLA-A*02:01 restricted AML/MDS-specific neoantigen which can be effectively targeted with TCR-T cells. We identified a highly potent TCR that eliminates cancer cells while sparing mutation negative cells. These data support further development of this TCR as a potential therapeutic for MDS and AML.

Improved Overall Survival Among Luspatercept Responders and Predictors of Response

Introduction Luspatercept (luspa) is a first-in-class erythroid-maturation agent approved for red blood cells transfusion dependent (RBC-TD) lower risk myelodysplastic syndrome (LR-MDS) with ring sideroblast (RS) based on MEDALIST clinical trial. We aimed to explore impact of luspa on overall survival (OS) and predictors of response in a large cohort of pts treated with luspa, collected from Moffitt Cancer Center (MCC) and from Fondazione Italiana Sindromi Mielodisplastiche (FISiM). Methods We evaluated all pts with LR-MDS-RS treated with luspa from MCC cohort since FDA approval, and from a multicenter observational trial of FISiM for pts treated on a compassionate use program . Baseline RBC transfusion burden (TB) was defined as: non-transfusion dependent (NTD) (0 units in 8 weeks prior luspa), low TB (LTB) (1-5 units/8 weeks ) and high TB (HTB) (≥ 6 units/8 weeks). A hematological response (HI) was defined as an (i) objective Hgb increase of > 1.5 g/dl in NTD, and (ii) RBC-TI with Hgb increase of 1.5 g/dl, or RBC-TI without Hgb 1.5 g/dl increase, or >50% reduction in RBC TB among RBC-TD. Somatic mutation (MT) data were available for 137 pts. Results Between January 2020 and March 2023, 331 pts were treated with Luspa (Table-1). The median age was 69 years; 308 pts (93%) had MDS-RS, 86,1% were intermediate, low or very low risk MDS by R-IPSS and 82% (105/128) were moderate low (ML), low (L) or very low (VL) by IPPS-M. SF3B1 MT was detected in 86.3% (120/139) with a median VAF of 37.7. The mean Hgb level was 7.97 g/dl and 93.7% were RBC-TD. Majority had prior erythroid stimulating agents (ESA) (93.6%). In the MCC cohort 42.9% (52/121) had prior hypomethylating agents (HMA) therapy, and 31.4% (38/121) had prior lenalidomide (len). With a median follow up of 14.6 mos from starting luspa treatment, the median OS was not reached (NR) among responders to luspa compared to 24.5 mos for non-responders (p <.001) (Figure-1). The median OS benefit was more pronounced among pts with RBC-TI and Hgb> 1.5 g/dl increase compared to other categories of response. In the NTD group at baseline, the median OS among luspa responders was NR compared to 22.7 mos among non-responders (P< .001). In the LTB group at baseline median OS was NR as well among luspa responders compared to 29.9 mos in non-responders (p=.002) and finally among HTB group at baseline, the median OS among luspa responders was 35.4 mos compared to 21.2 mos in non-responders (p=.007). In multivariable analysis adjusting for IPSS-M and baseline RBC-TD, response to luspa was independently associated with improved OS, HR .39 (95% CI .18-.83), p= .015 The overall HI rate to luspa was 40.6% (134/330) (Table-1), 85% of pts needed dose escalation, HI correlated with baseline RBC-TB where 80% (16/20) NTD, 60.2% (65/108) LTB and 41.8% HTB (77/184) achieved HI respectively, p< .001. By IPSS-M, there was a trend for lower HI with higher risk disease (100% in VL, 60.3% in L, 36.7% in ML, 56.3% in MH, 16.7% in H risk pts, p=.051). There was a mild trend for higher response among SF3B1 MT pts compared to wild type, 54% (64/119) versus 42% (8/19), p=.34.The segregation of SF3B1 MT cases into three independent groups based on IPSS-M, showed that HI rate was better in SF3B1 β ( SF3B1 and any gene from BCOR, BCORL1, NRAS, RUNX1, SRSF2 or STAG2) and SF3B1 α (as any other mutant) than SF3B1 5q(concomitantpresence with isolated del5q): 50% (7/14), 58.8% (57/97), and 0% (0/5) respectively, p=.03. Among SF3B1 5qpts, 4 had been treated with len before luspa and only 1 had achieved an HI with prior len. Among pts with SF3B1 MT, we observed a trend for higher response in pts with the hotspot K700E, 62.7% (37/59) vs 47.4% (27/57) in non-K700E, p=.097. In pts with SF3B1 MT, divided by median VAF, we observed a trend for better HI with VAF was >38, 63.5% (33/52) vs 45.3% (24/53), p=.062. Among responders, the median duration of response was 14.1 mos. As of last follow up, 85% pts (191/224) discontinued treatment. Reason of discontinuation was lack of response in 126 pts, loss of response in 26 patients and adverse events in 6 pts. The most common reported adverse events were fatigue (n=10), GI symptoms (n=4), shortness of breath and arthralgia (1 each). Conclusions Our RWD in the largest luspa treated cohort demonstrates OS benefit to luspa response. Low baseline RBC-TB dependency, lower risk IPSS-M, SF3B1 MT, SF3B1-K700E mutation and SF3B1 α co-mutations correlated with higher response rates. Our data emphasizes the impact of RBC-TI in LR-MDS.

Morphologic and Genetic Features of Unclassifiable Chronic Myeloid Neoplasms: Five-Year Experience of a Canadian Provincial Referral Center

Introduction: The challenge of diagnosing and definitively classifying myeloproliferative neoplasms (MPNs) and myelodysplastic/myeloproliferative neoplasms (MDS/MPNs) remains a conundrum in the field of hematology. While the World Health Organization has outlined diagnostic criteria for various categories, there remain cases that elude specific classification, falling under the umbrella of MPN, unclassifiable (MPN-U) and MDS/MPN, unclassifiable (MDS/MPN-U). In this study, we aimed to compare the morphological and genetic profiles of cases classified as MPN-U or MDS/MPN-U against more definitively classified cases (including chronic myelomonocytic leukemia (CMML), essential thrombocythemia (ET), MDS/MPN with ring sideroblasts and thrombocytosis (MDS/MPN-RS-T), and primary myelofibrosis (PMF), including prefibrotic PMF (pre-PMF)). Methods: Morphologically unclassifiable myeloid neoplasm cases referred for pathologist review to our laboratory in the largest tertiary care center in British Columbia, Canada were identified through pathology database query over a five-year period. Bone marrow materials were reviewed and reanalyzed with standardized collection of qualitative and semiquantitative data by two hematopathologists. Morphological parameters studied included the myeloid to erythroid ratio, bone marrow cellularity, lineage dysplasia, megakaryocyte histotopography, bone marrow blast frequency, fibrotic changes, and bone marrow vascularity, among others. Clinical data, cytogenetic findings, and results of a targeted sequencing panel assessing a set of approximately 50 genes recurrently mutated in myeloid malignancies were collected by electronic medical record review. A cohort of contemporarily collected myeloid neoplasms with definitive pathologic diagnoses was reviewed for comparison. Statistical analyses were performed using python3. Results: The cohort comprised 29 cases classified as MPN-U, 12 cases of MDS/MPN-U, and 5 cases each of ET, PMF, pre-PMF, CMML, and MDS/MPN-RS-T. MPN-U cases were further subcategorized based on the apparent reason for difficulty in classification, including: overlapping features with another MPN; late stage/overt fibrosis; early stage without fully-developed morphologic features; and coexisting atypical morphologic findings. MPN-U with overlapping features had highly variable reticulin fibrosis, while, as expected, late stage entities had a high degree of reticulin fibrosis and early stage entities showed less overt fibrosis (Figure, Left, demonstrating select morphologic features of evaluated chronic myeloid neoplasms). MPN-U as a whole appeared to show less prominent fibrosis-associated findings than MDS/MPN-U and to, in general, show a greater degree of neovascularization, similar to changes observed in cases with PMF. Splicing factor and epigenetic modifier mutations appeared to be overrepresented in MDS/MPN-U relative to MPN-U, notably excluding MPN-U with atypical morphologic features (Figure, Right, demonstrating the distribution of pathogenic variants identified in evaluated myeloid neoplasms; results are indicated as presence (1) or absence (0) of a variant for each gene). Notably, no CALR mutations were identified in the MDS/MPN-U category or early stage / atypical morphology MPN-U, though they were enriched in overtly fibrotic MPN-U entities. Discussion/Conclusion: In this cohort we observed, as expected, significant overlap in cases classified as MPN-U and MDS/MPN-U with better delineated entities, though patterns of morphologic change and mutational spectra suggest potential directions for further investigation of additional criteria that may guide accurate classification of these neoplasms. These findings reflect the heterogeneity of MPNs described in the literature and the complexity of their genetic underpinnings, particularly in the context of concurrent dyshematopoiesis; suggest enhanced bone marrow vascularization may be an informative differential diagnostic criterion in evaluation of these neoplasms, though further study is required; and emphasize the need for broader genetic testing and perhaps further subtyping of MPN-U cases for a more nuanced analytical approaches for these disorders.

Downregulation of SLC7A11 in MDS-SF3B1 Patients Induces a High Sensitivity of Erythroid Precursors to Ferroptosis

Introduction Myelodysplastic syndromes with SF3B1 mutation (MDS- SF3B1) represent a subtype of hematopoietic stem cell disorders characterized by bone marrow erythroid dysplasia and ring sideroblasts. They are associated with patient-impacting anemia and iron overload. Adequate iron levels are crucial for heme synthesis during erythropoiesis. An elevated iron level can enhance cell sensitivity to ferroptosis, a non-apoptotic iron-dependent form of cell death caused by lipid peroxidation. Recent studies have demonstrated that normal erythroblasts (EBs) are resistant to ferroptosis, expressing highly Glutathione Peroxidase 4 (GPX4), a ferroptosis suppressor, which requires Glutathione (GSH) as a cofactor. We hypothesize here that a loss of resistance to ferroptosis could occur in MDS- SF3B1 EBs, contributing to ineffective erythropoiesis Methods and results We first used UT7-EPO and HUDEP-2 as cell models of in vitro erythropoiesis to study the effects of SF3B1 knockdown (KD) using a shRNA lentiviral-based strategy. In both cell lines, we observed after SF3B1 KD a higher level of ferroptosis, as assessed using BODIPY 581/591 C11 staining, both at basal state and after exposure to Ras-Selective Lethal 3 (RSL3), which degrades GPX4 and L-Buthionine- (S, R)-sulfoximine (BSO), which inhibits the cysteine import and subsequently GSH synthesis. In EBs derived from CD34 + cells, SF3B1 KD induced a delayed differentiation together with an increased level of ferroptosis, and a decreased level of GSH. Since GSH production requires import of Cystine through the Solute Carrier family 7 member 11 ( SLC7A11) transporters, we quantified this transporter by Western Blot and found it decreased in comparison to control. The use of γ-Glutamyl-Cysteine, which bypasses SLC7A11 for GSH synthesis, decreased ferroptosis both at basal state and after RSL3 or BSO exposure.In normal EBs, GPX4 inhibition is not enough to induce a high ferroptosis. We hypothesized that, after SF3B1 KD, the decreased GSH lowered the “trigger” for ferroptosis, but that an additional stimulus was required. Free iron appeared to be a good candidate since (i) it plays the central role in ferroptosis, (ii) an excess of free iron is observed in MDS- SF3B1 EBs. We exposed EBs after SF3B1 KD to free iron. We observed increased intracytoplasmic reactive oxygen species level (ROS), measured using the H2DCFDA probe and increased membrane lipid peroxidation as well as accumulation of intracytoplasmic iron assessed by FerroOrange probe while it decreased rapidly after a peak in the control conditions. Quantifying ferroportin revealed a decreased expression after SF3B1 KD. α-Tocopherol, that reverts lipid peroxidation and N-Acetyl-Cysteine, that decreased intracellular ROS, restored ferroportin expression, iron excretion andoxidative stress. At last, we used primary bone marrow cells from 3 MDS- SF3B1 patients and confirmed the phenotype observed in other cell models characterized by (i) an excess of basal ferroptosis, (ii) a decreased SLC7A11 and ferroportin expression and (iii) an accumulation of intracellular free iron in comparison with normal EBs. Conclusion In summary, we show that SF3B1 KD or mutation in primary cells from patients, impaired the natural resistance of EBs to ferroptosis together with a defect in erythroid differentiation and an increased mortality, potentially contributing to anemia. The heightened susceptibility to ferroptosis is associated with downregulation of SLC 7A11 expression, leading to a decreased GSH synthesis. Moreover, the high level of ROS is associated with a decreased ferroportin expression, leading to accumulation of intracellular free iron and providing a permanent “fuel” for membrane lipid peroxidation that could be reverted by ferroptosis-blocking agents such as Vitamin E or Deferasirox. Both showed promise in restoring erythroid differentiation, preventing EB cell death, and may serve as an interesting therapeutic strategy for treating MDS- SF3B1 patients.

Combined Myeloid and Lymphoid Malignancies in 13 Elderly Patients

Disclosures Myeloid malignancies such as myelodysplasia (MDS) often evolve from age related clonal hematopoiesis (CH) with progression of variant allele frequencies. The Philadelphia negative myeloproliferative neoplasms (MPN) polythemia vera (PV), essential thrombocythemia (ET), and myelofibrosis (MF) are associated with mutations in JAK2, CALR, or MPL. Multiple myeloma (MM) and monoclonal gammopathies (MGUS) are felt to be age related, and the malignant cell is of lymphoid origin. Burkitt's Lymphoma is associated with EBV exposure and dysregulation of c-myc. However the etiology of most lymphoproliferative disorders (LPD) is not defined, but immune dysregulation may be causative. The combination of myeloid and lymphoid malignancies are not recognized to be a common occurrence; however within a three year period 13 patients with combined myeloid and lymphoid malignancies were detected in a single community practice. DATA Five patients had combined myeloid and lymphoid malignancies diagnosed concurrently. Three had a lymphoid disorder and then developed a myeloid disorder. Five had a myeloid disorder followed by a lymphoid disorder. Three patients had multiple myeloma (MM) and two had MGUS. Initially all patients responded to standard therapies, One patient at 87 years died from end stage renal disease likely related to her myeloma. Another patient with CLL and PV died at 92 years from an epithelial malignancy. Concurrent diagnoses Age at diagnosis (years) CLL ET 76MGUS MDS 77MM MDS 78MM MDS (RSBA) 85 5) DLBCL MF. 60 LPD/MM followed by a Age at initial diagnosis (years) and time to second diagnosis myeloid disorder. 1) MZL MF 58 15 months 2) MM ET 83 7 months 3) CLL PV 86 4 years Myeloid disorder followed by a LPD/MGUS MDS DLBCL 76 3 monthsMDS MGUS 78 30 monthsTP CLL 15 60 yearsET MyF 77 65 monthsET CLL 60 15 years (CLL chronic lymphocytic leukemia, DLBCL diffuse large B cell lymphoma, MF myelofibrosis, MyF mycosis fungoides, MZL mantle zone lymphoma, RSBA ringed sideroblastic anemia, TP inherited thrombocytopenia) JAK2 mutations were found in three of the four ET patients and in the one PV patient. One MF patient and one ET patient were negative for the JAK2 and CALR mutations. The mutation status of the other MF patient was not known. At diagnosis the patient with inherited thrombocytopenia was suspected to have a myeloid disorder. Combined myeloid and lymphoid malignancies is not well defined and is a seemingly uncommon event. Yet an Italian study demonstrated that MPN patients with JAK2 mutations have a 2.8-3.4 fold increased risk of developing a LPD compared with an age matched population (1). The linkage is unclear. Immune senescence may be important for the development of LPD observed in the elderly. Also long term latency of many years between the JAK2V617F mutation acquisition and MPN presentation has been documented (2). The occurrence of both myeloid disorders and LPD is not well recognized. Future studies may provide insights into the mechanisms underlying these disorders. Increasing genetic complexity and immune senescence associated with aging likely plays a major role. The limitations of this study are its small size, single community investigator, retrospective clinical nature, and the lack of gene sequencing and other molecular studies in most if not all of the patients. This observation was noted over three years in a single community practice which strongly implies that combined occurrence may be more common than recognized and suggests that further investigation is warranted. Boxer none, Denes none

Successful Management of D816V Mutated c-KIT Systemic Mastocytosis with Chronic Imatinib: A Case Report

First characterized in France in 1936, systemic mastocytosis (SM) is a rare disorder by an excessive proliferation and accumulation of abnormal mast cells in extracutaneous tissues (Nettleship et al., 1869). The release of large quantities of vasoactive mediators by these cells results in allergic- and anaphylactic-type reactions in one or more organ systems. Given the similarity of symptoms between allergic reactions and SM and the rarity of the condition, SM is often missed and misdiagnosed (Schwaab et al., 2020). Once considered a subtype of myeloproliferative neoplasms, mastocytosis was reclassified in the 2016 revision to the World Health Organization (WHO) due to advances in understanding of the disorder (Arber, et al., 2016). Presented below is a case of a severe systemic mastocytosis with an initially positive D816V mutated c-kit successfully managed with long-term imatinib use alongside intermittent bone marrow biopsies (BMB) to monitor what is now an indolent systemic mastocytosis. 69 year-old female with a past medical history of paroxysmal atrial fibrilation, hashimoto's thyroiditis, metabolic syndrome, and gastro-esophageal reflux (GERD) presents to clinic for routine management of what was previously severe systemic mastocytosis (SSM) and now is indolent systemic mastocytosis with the help of imatinib. To help control her ISM her medications include a daily histamine-1 blocker (ranitidine), a daily histamine-2 blocker (diphenhydramine), and a twice-daily histamine-1 antagonist (cetirizine). Originally evaluated in 2005 with a bone marrow biopsy (BMB) and found to have a Asp816Val (D816V) c-kit mutation, she was placed on imatinib with significant clinical improvement. This seemed to support that the D816V mutation may cause constitutional activation of the c-kit transmembrane protein and thus sensitivity to imatinib therapy. This despite the FDA labeling for imatinib as only to be used in SM patients without the common D816V mutation or for whom the mutational status is unknown (Reference ID: 4957210, 2001). Interestingly, after being placed on imatinib her repeat bone marrow biopsies in 2011 and 2015 indicated that the D816V mutation was no longer present in her CD117 positive mast cells but continued to benefit from imatinib therapy. This finding is suggestive of two unique mast cell populations as imatinib continued to help maintain a low tryptase level and decrease the severity of reactions. When reactions did occur, she treated them at one point with a cocktail of an histamine-2 blocker (ranitidine), histamine-1 blocker (diphenhydramine), and a mast cell stabilizer (cromolyn) and without requiring the use of epinephrine. As a part of her longitudinal observation, she received an additional BMB in 2022 that revealed a minimal residual monocytosis that persisted with less than 10% nucleated cells, an age-appropriate trilinear hematopoiesis, decreased iron stores without ringed sideroblast formation. Flow cytometry was indicative of a typical female karyotype with the population of aberrant mast cells that is consistent with the diagnosis of ISM. Importantly, no abnormalities were detected in the FLT3, IDH1, IDH2, NPM1 genes which, if abnormal, could obfuscate the apparent efficacy of imatinib treatment in the management of SM. Given the dearth of mastocytosis in the BMB and the absence of ongoing end-organ damage, the patient meets the definition of ISM. After the most recent BMB the patient altered her daily regimen due to gastric upset and found she could take of her histamine-1 blocker (diphenhydramine) once daily along with a twice daily regimen of her imatinib (split into two smaller daily doses), histamine-1 antagonist (cetirizine), her histamine-2 antagonist (famotidine), and citalopram (selective serotonin reuptake inhibitor or SSRI) with improved symptomatic control to this day. This case serves to illustrate the diagnostic and therapeutic challenges associated with SM, as well as the potential therapeutic use of imatinib in patients with D816V outside of its original FDA indication. Further characterization of the effect of imatinib on D816V mutated SM may warrant the consideration of expanding the FDA indications for imatinib as its role and efficacy outside of solely D816V negative c-kit MS.

The ELEMENT-MDS Trial: A Phase 3 Randomized Study Evaluating Luspatercept Versus Epoetin Alfa in Erythropoiesis-Stimulating Agent-Naive, Non-Transfusion-Dependent, Lower-Risk Myelodysplastic Syndromes

Background and Significance: Anemia is a common and debilitating symptom in patients with myelodysplastic syndromes (MDS), frequently leading to dependence on red blood cell (RBC) transfusions. Transfusion-dependent (TD) patients have worse overall survival and quality of life, and repeat transfusions are associated with complications, including iron overload. Erythropoiesis-stimulating agents (ESAs) are the standard treatment for anemia in patients with lower-risk (LR)-MDS, but many patients are ineligible for or refractory to ESAs, and responses are limited. Luspatercept, a late-stage erythroid maturation agent, is approved to treat TD patients with LR-MDS after ESA failure. In the phase 3 COMMANDS study, in ESA-naive patients with LR-MDS requiring transfusions, a significantly higher proportion of patients receiving luspatercept than epoetin alfa achieved RBC transfusion independence for ≥ 12 weeks with concurrent hemoglobin (Hb) increase ≥ 1.5 g/dL (weeks 1-24) (Platzbecker U, et al. Lancet 2023. doi: 10.1016/S0140-6736(23)00874-7). The potential of luspatercept to improve moderately severe chronic anemia and reduce RBC transfusion dependence has not yet been investigated in the majority non-transfusion-dependent (NTD) LR-MDS patient population. Preventing transfusion dependence is crucial for improving LR-MDS patient outcomes. The aim of this actively enrolling, phase 3, randomized, multicenter study is to compare the safety and efficacy of luspatercept versus epoetin alfa in reducing progression to RBC transfusion dependence in ESA-naive, NTD adult patients with anemia due to LR-MDS. Study design and methods: The ELEMENT-MDS study includes a 5-week screening period, a 96-week treatment period, an extension phase, and a post-treatment follow-up period (Figure). Target enrollment is 360 patients with anemia due to Revised International Prognostic Scoring System (IPSS-R) Very low-, Low-, or Intermediate-risk MDS. Eligible patients are ≥ 18 years of age with a confirmed MDS diagnosis, baseline serum erythropoietin (sEPO) levels ≤ 500 U/L, and symptomatic anemia defined as a score of moderate or worse on ≥ 1 Patient Global Impression of Severity (PGI-S) item (fatigue, shortness of breath, weakness, or dizziness). Participants must be RBC transfusion-independent per International Working Group 2018 criteria (no RBC transfusions within 16 weeks prior to randomization) and have baseline Hb ≤ 9.5 g/dL. Patients will be excluded if they have a del(5q) cytogenetic abnormality, unclassifiable MDS or myelodysplastic/myeloproliferative neoplasms, anemia due to reasons other than MDS, bleeding disorders, low absolute neutrophil or platelet counts, uncontrolled hypertension, previous acute myeloid leukemia diagnosis, or history of recent malignancies, among other reasons. Participants will be randomized 1:1 to receive luspatercept or epoetin alfa. Stratification factors include baseline sEPO levels, ring sideroblast status, and IPSS-R risk category. Patients will receive luspatercept once every 3 weeks at a starting dose of 1.0 mg/kg (escalation up to 1.75 mg/kg allowed) or epoetin alfa once a week at a starting dose of 450 IU/kg (escalation up to 1050 IU/kg allowed). The primary endpoint will compare the proportion of patients becoming TD (defined as requiring ≥ 3 RBC units/16 weeks) during any continuous 16-week interval in weeks 1-96. The key secondary endpoint is a Hb increase ≥ 1.5 g/dL sustained for ≥ 16 weeks during weeks 1-48. Additional secondary endpoints include time to RBC transfusion dependence, transfusion-free survival, time to achieve modified hematologic improvement-erythroid response, transfusion independence ≥ 24 weeks, and quality of life assessment. Safety endpoints include type and severity of adverse events and relationship to treatment, overall survival, and progression to high-risk MDS and acute myeloid leukemia. Summary: The results of this phase 3 study will determine the efficacy and safety of luspatercept as a therapy for ESA-naive NTD patients with LR-MDS, who have limited treatment options for anemia. The study is registered at ClinicalTrials.gov (NCT05949684) and EudraCT (2022-500430-29-00).

Clinical and laboratory features of SF3B1-mutated myeloproliferative neoplasms

Objective: To explore the clinical and laboratory characteristics of SF3B1 gene mutations in myeloproliferative neoplasms (MPN) patients. Methods: The clinical data of 273 MPN patients who were diagnosed MPN and treated in the Second Hospital of Tianjin Medical University from November 2017 to March 2023 were retrospectively analyzed. There were 133 males and 140 females, with a median age M(Q1,Q3)of 56(46, 67) years. The molecular biology and cytogenetic characteristics were detected by second-generation sequencing (NGS) and R+G banding techniques, and the clinical and laboratory characteristics of patients with SF3B1 gene mutation were analyzed. Results: SF3B1 gene mutations were found in 13 patients (4.8%, 13/273).The types of SF3B1 mutations included missense (92.3%, 12/13) and nonsense mutations (7.7%, 1/13).Compared to the non-mutant cohort, patients in SF3B1 mutant cohort had older ages [68(51, 76) vs 56(45, 66)years,P=0.025], higher proportion of splenomegaly [46.2%(6/13) vs 15.8%(41/259),P=0.014]and secondary tumor [23.1%(3/13)vs 3.8%(10/260), P=0.018]with higher proportion of bone marrow blast [0.5%(0, 1.5%) vs 0(0, 0.5%),P=0.002] and lower hemoglobin[(104±36) vs (137±40) g/L,P=0.004] and hematocrit [31%(22%, 40%) vs 41%(35%, 52%),P=0.003]. All of the 10 patients in the SF3B1 mutant cohort whose ring sideroblast (RS) could be evaluated showed no RS formation. The overall survival, thrombosis-free survival and leukemia free survival of MPN patients in SF3B1 mutant cohort were 4.0 (2.0, 6.0), 2.0 (0.5, 4.5) and 4.0 (2.0, 6.0) years, respectively, while patients in the non-mutant cohort were 6.0 (3.0, 10.0), 5.0 (1.0, 8.0), 6.0 (3.0, 10.0) years, respectively, there were no statistical significance between two groups (Z=3.69, 1.66, 2.05, all P>0.05).The secondary tumor free survival of SF3B1 mutant cohort patients was 4.0 (2.0, 6.0) years, which was lower than that of non-mutant cohort patients [5.5 (3.0, 10.0) years, Z=18.18, P<0.001). Conclusions: MPN patients with SF3B1 gene mutations are older, more prone to splenomegaly and secondary tumors. They also have a higher proportion of bone marrow blast, lower hemoglobin and hematocrit, and show no RS formation.

Erythroid Differentiation Enhances RNA Mis-Splicing in SF3B1-Mutant Myelodysplastic Syndromes with Ring Sideroblasts.

Ring sideroblast isolation combined with state-of-the-art multiomics identifies survival mechanisms underlying SF3B1-mutant erythropoiesis and establishes an active role for erythroid differentiation and ring sideroblasts themselves in SF3B1-mutant myelodysplastic syndrome pathogenesis.

Early-Onset Myelodysplastic Syndromes (MDS) with Ring Sideroblasts (RS) without SF3B1 Mutations in Adults: Enrichment with Germline Variants in Genes Responsible for Congenital Sideroblastic Anemias

Introduction: Acquired mutations in SF3B1 gene in myeloid neoplasms are classically associated with the presence of RS in Pearl Stain, ineffective erythropoiesis, and favorable prognosis. Nevertheless, in up to 20-25% of adults diagnosed with MDS with RS (WHO 2017), mutations in SF3B1 are not found, and molecular grounds for the presence of RS remain to be ascertained. The objective of our study was to investigate whether the presence of RS could be associated with germline variants in genes responsible for congenital sideroblastic anemia (CSA). Methods: Patients diagnosed with de novo MDS between 16-60 years of age without previous organ dysfunction were recruited from 32 centers of GEMSD since 2016. Whole exomes were sequenced using HiSeq4000-NovaSeq6000-Illumina, paired tumor-germline samples. Mean depth was 100x, with 150 million reads per sample and quality Q30a&amp;gt;95%. Variants were analyzed using a bioinformatics pipeline: filtering intronic and synonymous variants and those with a population frequency &amp;gt;1%. The mutational state of SF3B1 was determined by Sanger Sequencing and Next Generation Sequencing (NGS). Germline variants were categorized according to American College of Molecular Genetics (ACMG) criteria. The list of genes related to CSA explored in this study is shown in Table 1. Results: Among 239 cases of adults diagnosed with early-onset MDS (mean age at diagnosis: 48 years, range 16-60), 58 (24%) patients presented with RS in bone marrow (mean RS: 28%). Of these 58 patients, acquired mutations in SF3B1 were not found in 32 (55%). Nine out of these 32 (25%) harbored a germline variant (two variants in one case) in genes responsible for CSA (Table 2): SLC25A38 (n=2), STEAP3 (n=2) FECH, ALAS2, GLRX5, SLC19A2, TRNT1 and IARS2. This frequency was statistically higher than in the SF3B1 and RS mutated group (n=23), with only one case with a germline variant in a CSA gene (p=0.013). It was also higher than in the non-RS cases (n=181), with only one case in this group (p&amp;lt;0.001). Using the Fisher's exact test, commonly used to perform enrichment, the odds ratios were also significant (p=0.033 and p&amp;lt;0.0001), respectively. Among patients with RS, those carrying a CSA gene germline variant were younger (43 vs. 54 years, p=0.04), had a higher rate of neutropenia (1.8 vs. 2.6 x 10E9/L, p=0.02), and thrombocytopenia (151 vs. 259 x 10E9/L, p=0.03) than MDS-RS patients with SF3B1 mutated. Furthermore, MDS-RS with a germline variant in CSA genes had a lower mean percentage of RS than patients who acquired the mutation in SF3B1 (14% vs. 38%; p=0.003). Conclusions: In our series, the frequency of MDS-RS without SF3B1 mutations is higher in early-onset adult MDS than the one reported in MDS in advanced age. Whole exome analysis allowed us to describe, for the first time, a significant enrichment of variants in genes causing CSA in young adults with RS and without acquired mutation in SF3B1.

Luspatercept Modulates Inflammation in the Bone Marrow, Restores Effective Erythropoiesis/Hematopoiesis, and Provides Sustained Clinical Benefit Versus Epoetin Alfa (EA): Biomarker Analysis from the Phase 3 COMMANDS Study

Introduction :Anemia is common in patients (pts) with myelodysplastic syndromes (MDS) due to ineffective erythropoiesis, characterized by increased dysplastic erythroid precursors (EPs) and abortive differentiation in the bone marrow (BM). Insufficient red blood cells (RBCs) lead to chronic anemia and heart failure due to oxygen deprivation. Luspatercept is approved to treat anemia after erythropoiesis-stimulating agent (ESA) failure in pts with transfusion-dependent, lower-risk MDS (LR-MDS). In the phase 3 COMMANDS trial (NCT03682536), 58.5% of ESA-naive luspatercept treated pts with LR-MDS experienced RBC transfusion independence for ≥ 12 wk with concurrent mean hemoglobin (Hb) increase ≥ 1.5 g/dL during wk 1-24, vs 31.2% of EA-treated pts. Aims :To differentiate the mechanism of action of luspatercept from EA and determine its correlation with clinical benefit in the COMMANDS study. Methods :Cytomorphology assessments were performed on BM aspirates at baseline, wk 24 and 48. Complete blood count, Hb, and reticulocytes (retics) were measured in peripheral blood. BM mononuclear cells (BMMC) were subjected to bulk RNA-seq. Serum cytokine analysis was performed at screening and wk 24 using Rules-Based Medicine's Human Inflammation MAP ® v.1.1 and a custom Human MAP ELISA panel. Results :In ITT population, EPs, retics, and Hb were increased vs baseline in the luspatercept arm at wk 24 ( P &amp;lt; 0.001) and wk 48 ( P &amp;lt; 0.005). EPs, retics, and Hb were increased in the EA arm at wk 24 ( P &amp;lt; 0.01), but this was not sustained to wk 48. Subgroup analysis showed similar trends for ring sideroblast (RS)+ pts in both arms. While RS− pts in the EA arm showed similar trends to the ITT population, luspatercept RS− pts showed trends of EP expansion post-wk 24 only. Retics increased significantly in RS− luspatercept pts at wk 24 ( P &amp;lt; 0.005). Hb increased more significantly at wk 24 with EA ( P &amp;lt; 0.003) but at wk 48 with luspatercept ( P &amp;lt; 0.005).Gene set enrichment analysis of BMMC revealed enrichment of early, mid, and late EP genes at baseline favored response to luspatercept, but enrichment of mid and late EP genes was unfavorable for EA (Figure A). Significant downregulation of REACTOME IL-1 signaling (normalized enrichment score [NES] = −2.27; adjusted [Adj] P &amp;lt; 0.0001 [vs EA = not significant (ns)]), BIOCARTA IL-6 pathway (NES = −1.99; Adj P = 0.0025 [vs EA = ns]) and IFNα response (NES: −2.26; Adj P &amp;lt; 0.0001 [vs EA = ns]) was observed with luspatercept. BIOCARTA IL-1R pathway was more significantly reduced with luspatercept (NES = −1.78; Adj P = 0.0073) compared with EA (NES = −1.79; Adj P = 0.0172), as was IL-10 signaling (luspatercept NES = −2.39; Adj P &amp;lt; 0.001 vs EA NES = −1.82; Adj P = 0.0125). TGF-β, apoptosis, and spliceosome pathways down regulation was also observed with luspatercept. At wk 24, luspatercept responders had a median 640 pg/mL decrease ( P = 0.0035) of N-terminal pro-brain natriuretic peptide (NT-proBNP), produced by the heart, compared with a median 360 pg/mL decrease in EA arm responders ( P = 0.37). At wk 24, non-responders had a median increase of 775 pg/mL ( P = 0.0032) in the EA arm and 31.5 pg/mL ( P = 0.95) in the luspatercept arm. Conclusions : Multiplatform analysis of primary BM samples from a phase 3 trial of luspatercept revealed novel response-associated biomarkers. Compared with EA, luspatercept led to sustained increase of EPs and retics over 48 wk with a concomitant Hb increase. RS− pts showed Hb accumulation peaked at wk 24 in the EA arm. Conversely, luspatercept in RS− pts significantly increased retics at wk 24 and EPs past wk 24 with concomitant peaking of Hb at wk 48, suggesting gradual and sustained clinical benefit. Unlike EA, which plays a role in early differentiation of hematopoietic stem cells (HSCs), luspatercept acts on different erythroid stages, leading to EP expansion and maturation. Furthermore, downregulation of IL-1, IL-1R, IFNα, IL-6 pro-inflammatory (Figure B), and IL-10 anti-inflammatory signaling with luspatercept suggests modulation of the BM inflammatory environment in MDS to restore HSC function for normal hematopoiesis (Wang J, et al. Front Immunol 2022;13:951937). Additionally, decreased or static levels of NT-proBNP with luspatercept suggest a positive effect in mitigating cardiac damage. These novel insights obtained from analysis of a large controlled clinical trial mechanistically differentiate the superior clinical benefit of luspatercept from EA in treating anemia due to MDS.

Improved Quality of Life upon Shifting from Darbepoetin to Luspatercept in Patients with Low-Risk Myelodysplastic Syndrome (MDS)

Introduction: Low-risk MDS is characterized by refractory anemia due to Dyserythropoiesis 1. Standard of care treatment is mainly Erythropoiesis stimulating agents (ESA) or blood transfusion or both 1. Response to ESA is achieved in some patients whose erythropoietin (EPO) level is less than 500 mu/ml 2. Luspatercept is a red cell maturing agent that may lead to an increase in Hb of up to 3 g/dL 3 in responders. Response to ESA can be enhanced with the addition of Luspatercept 4. Hereby we are reporting successful shifting of two patients from Darbepoietin to Luspatercept with significant increase in Hb level and improvement in quality of life. Case 1: A 62-year-old man without previous illness presented in April 2018 with progressive symptomatic anemia. His body weight was 107 kg, and his physical examination was unremarkable other than pallor and tachycardia. The laboratory investigation revealed WBC, 5.62 x 10 3/uL, RBC 2.2 x 10 3/uL, Hb 7.8 g/dL, MCV 110.4 fl, platelets 706 x 10 3/uL, and reticulocytes 1%. Blood film showed reduced RBC count with numerous macrocytes and no other abnormalities. Bone marrow revealed 80% cellularity with Dyserythropoiesis, 16% ring sideroblasts, and 2% blasts. The Ferritin level, B12 levels were elevated (table). EPO level was elevated but less than 500 mlU/ml and the cytogenetic analysis showed normal karyotype. However his molecular profiling demonstrated SF3B1 and TET2 mutations. At the time of diagnosis, the Revised International Prognostic Scoring System (IPSS-R) was 2.0, indicating low risk MDS. He was given darbepoetin 300 mcg SC weekly for four weeks without any improvement, then darbepoetin dose was escalated to 500 mcg weekly with a good response. Hb level remained at 8.5 - 9.5 g/dl for three years. The patient's physical performance improved moderately, with an increase in walking distance on flat surfaces from 20 to 35 minutes. In August 2022, he was switched to Luspatercept 100 mcg SC on a three-weekly schedule. Within three months, his hemoglobin level increased to 10.8 g/dl and he experienced significant improvement in physical performance and mental well-being, along with a significant reduction in fatigue, enabling him to increase his walking distance to 76 minutes. Case 2: A 77-year-old man with type 2 diabetes mellitus and mild renal impairment who developed an unexplained progressive anemia. His body weight was 79 kg, and his physical examination was unremarkable other than pallor. His laboratory investigations showed WBC 4.5 x 10 3/uL, RBC 1.5 x 10 6/uL, Hb 5.2 g/dL, MCV 96 fl, platelets 706 x 10 3/uL, and reticulocytes 0.8 %. Blood film showed reduced RBC count, normocytic anemia, and no other abnormalities. Bone marrow revealed 60% cellularity with Dyserythropoiesis, 17% ring sideroblasts, and 1% blasts. His baseline serum creatinine was 1.8 mg/dl, and HbA1C 10.0. Two units of packed red blood cells were transfused every 2- 4 week to keep his Hb level around 8 g/dL. Darbepoetin 80 mcg SC weekly for four weeks,150 mcg weekly for 4 weeks and 300 mcg weekly for further 4 weeks without any response. Eventually, he responded to 500 mcg of Darbepoetin per week. As a result, his Hb increased to the range of 8.6-9.4 g/dl for one year, with moderate improvements in performance status. Darbepoetin was switched to Luspatercept 75 mcg three-weekly in January 2023, resulting in an increase in his hemoglobin level to 10.3 g/dl, with significant improvement in his energy level and overall well-being. Discussion: Blood transfusion and ESA have traditionally been used as standard treatments for low-risk MDS. Recently, Luspatercept in combination with ESA has shown enhanced effectiveness in elevating Hb levels. Both of our patients had significant increases in Hb levels along with a noticeable improvement in quality of life because of Luspatercept alone. Conclusion: Following the switch from Darbepoetin to Luspatercept, there was a significant improvement in Hb level and in quality of life in our 2 patients. This finding needs to be confirmed with larger studies.

Clinico-Genetic and Prognostic Analyses of 635 Patients with Myelodysplastic Neoplasms Based on the 2022 World Health Organization Classification

Myelodysplastic neoplasms (MDS) are a heterogeneous group of clonal myeloid neoplasms characterized by dysregulated hematopoiesis. In the past several years, major advances in molecular technologies and the development of next generation sequencing have deepened our understanding of MDS pathobiology. Revisions of the existing classification were warranted. Under the aegis of the International Agency for Research on Cancer, the 2022 World Health Organization (WHO) classification was released. The 2022 WHO proposed new categories including MDS with biallelic TP53 inactivation (MDS-bi TP53), MDS, hypoplastic (MDS-h) and MDS with fibrosis (MDS-f). In addition, the novel risk scoring system-Molecular International Prognostic Scoring System (IPSS-M) has been established. There are limited data that explore the application of the 2022 WHO classification of MDS in light of the IPSS-M. In this study, a cohort of 635 patients diagnosed as having primary MDS based on the 2022 WHO criteria was retrospectively reviewed. We aimed to elucidate the differences in genetic features and clinical outcomes among patients with different MDS subtypes, based on the 2022 WHO classification. The survival impact of allogeneic hematopoietic stem cell transplantation (allo-HSCT) was also evaluated. TruSight myeloid sequencing panel and the HiSeq platform were used to analyze the alterations of 54 myeloid neoplasm-relevant genes. Comparison of the genetic profiles of patients with different subtypes of MDS showed that patients with MDS-low blasts (LB) and those with MDS with LB and ring sideroblasts (RS) exhibited comparable mutation profiles, while patients with MDS- SF3B1 had different mutation pattern from that of these two subtypes. Among patients with MDS-increased blasts-1 (IB1), MDS-IB2 and MDS-f, patients with MDS-IB2 had higher frequencies of NRAS (9% vs. 2%, P=0.019) and mono-allelic TP53 (7% vs. 1%, P=0.031) mutations, and a trend of more TET2 (19% vs. 10%, P=0.059) mutations than those with MDS-IB1. In addition, patients with MDS-f exhibited lower incidence of TET2 (0% vs. 19%, P=0.020), and STAG2 mutations (8% vs. 27%, P=0.050) as compared to those with MDS-IB. Furthermore, patients with MDS-bi TP53 had specific concurrent genetic alterations with few co-occurring mutations which differed from those found in patients with MDS-IB or MDS-f. We performed pairwise survival comparisons among each subtypes of MDS (Figure 1). In low-risk MDS, patients with MDS-h had significantly longer median leukemia-free survival (LFS) and overall survival (OS) (185.5 months for median LFS and OS) compared with those with MDS-LB and RS [38.3 months for median LFS ( P=0.013) and OS ( P=0.010)], while had a similar outcome compared to those with MDS-LB [170.2 months for median LFS ( P=0.152) and OS ( P=0.146)], and MDS- SF3B1 [114.6 months for median LFS ( P=0.549) and OS ( P=0.528)]. For patients with increased blasts or bi TP53 mutations, those with bi TP53 inactivation had the worst outcomes (median LFS: 3.9 months and OS: 4.5 months, all P&amp;lt;0.001) compared to those with MDS-f (median LFS: 16.6 months and OS: 17.7 months), MDS-IB2 (median LFS: 10.0 months and OS: 17.7 months), and MDS-IB1 (median LFS: 25.6 months and OS: 31.4 months). Patients with MDS-f had similar survival to those with MDS-IB2 and IB-1 in terms of LFS (16.6 vs. 25.6 months, P=0.271) and OS (17.7 vs. 31.4 months, P=0.194) . Multivariable analysis showed that IPSS-M ( P&amp;lt;0.001), old age (HR for LFS: 1.023; for OS: 1.032, P&amp;lt;0.001), and the 2022 WHO categorization ( P&amp;lt;0.001) could predict LFS and OS independently and that HSCT could improve LFS (HR, 0.492, P=0.001) (Table 1). Subgroup analysis of the impact of transplantation using time-dependent Cox regression revealed that allo-HSCT could prolong both LFS and OS in patients with MDS-IB1 (HR for LFS: 0.293, for OS: 0.350, P&amp;lt;0.001), LFS alone (HR: 0.534, P=0.049) in patients with MDS-IB2, and a trend of LFS (HR: 0.305, P=0.078) in patients with MDS-f. However, allo-HSCT failed to remedy the dismal outcomes for patients with MDS-bi TP53. In conclusion, based on the evidence gathered from this cohort analysis, the 2022 WHO classification promote efficient segregation of this heterogeneous disease, with focus on the differences in molecular features and prognoses across different disease subtypes, leading to accurate MDS diagnosis and effective risk-adapted treatment.

GDF11/SMAD Regulated Splicing of GATA1 Is Associated with Response to Luspatercept in Lower-Risk Myelodysplastic Syndromes (LR-MDS)

Introduction: Luspatercept is an erythroid maturation agent that promotes late-stage erythropoiesis in patients with LR-MDS. While it is known that luspatercept, a modified activin receptor ligand trap, binds to GDF11 and Activin B, the mechanism by which these ligands trigger a decrease in red blood cell (RBC) formation is unknown. Luspatercept was approved by the FDA following the phase 3 MEDALIST study, but biomarkers of response are currently unknown. Results:In a large cohort of primary MDS CD34+ samples, GDF11 expression was significantly higher in MDS samples than in age-matched healthy controls (183 MDS samples vs 17 controls, P &amp;lt; 0.05). GDF11 was detected in primary MDS sera and was substantially elevated in the refractory anemia (RA) and RA with ring sideroblasts (RARS) MDS subgroups. Activin receptor 2B (ACVR2B), the receptor of GDF11, was also overexpressed in MDS samples, and its expression was confirmed by immunohistochemistry (IHC) in the bone marrows of primary MDS patients. GDF11/ACVR2B activates SMAD2, and we found that the overexpression of SMAD2 in CD34+ cells was associated with more severe anemia in MDS patients. RNA-seq and SMAD2 ChIP-seq analyses were then performed to determine the downstream effects of GDF11 in erythroid progenitors. Human CD34+ primary cells were utilized to generate BFU-E stage erythroid progenitors. We discovered that GDF11 promotes an abnormal erythroid gene expression program characterized by increased expression of pro-apoptotic and early erythroid differentiation-related genes. SMAD2 binding was detected in regulatory regions of early erythroid (including GATA1), proinflammatory, and apoptotic transcripts using ChIP-seq analysis with p-SMAD2. In a zebrafish model, administration of GDF11 caused a decrease in erythroid cells, which was rescued by treatment with luspatercept (Wobus et al. Leukemia 2021 35(10):2936-47). Luspatercept inhibited the phosphorylation and activation of SMAD2 in response to GDF11 stimulation in primary hematopoietic cells and leukemia cell lines. Importantly, luspatercept could reverse the growth-inhibiting effects of MDS-derived sera on primary hematopoietic progenitors. Finally, to identify biomarkers associated with clinical responses, we analyzed phase 3 MEDALIST study samples collected before and after treatment with lupatercept. RNA-seq analysis was performed on bone marrow samples from responders (R; n = 14) and non-responders (NR; n = 9) (Figure). Analysis of pre-treatment/screening bone marrow transcriptomes revealed no significant differences between R and NR in global gene expression. Analysis of splice isoforms between the two cohorts revealed strikingly greater differences. Most differential splicing events were observed in NR (702 in R vs 5232 in NR). The most prevalent splicing modifications were exon skipping events, and most functional pathways affected by differential exon usage were RNA splicing related. GATA1, the master erythroid transcription factor, was spliced differently in R at baseline, which was an intriguing observation. Response to luspatercept was found to be associated with exon 2 skipping, resulting in a shortened isoform of GATA1. R had a greater proportion of the short GATA1 isoform than NR (12% in R versus 6% in NR; P = 0.04). We analyzed ChIP-seq data from primary erythroid cells to determine the effect of GDF11/SMAD2 activation on GATA1 splicing. GDF11 stimulation resulted in SMAD2 binding to the second intron of GATA1. CRISPR-mediated deletion of the SMAD2 binding region in primary erythroid progenitors increased expression of the GATA1 long isoform. Next, we performed a CRISPR-mediated deletion of exon 2 of GATA1, resulting in a predominance of GATA1 short isoform expression and a decrease in erythroid output. In CRISPR-modified primary erythroid cells with a predominant short GATA1 isoform (exon 2-deleted cells), luspatercept was able to reverse the inhibition induced by GDF11. Conclusions: These results demonstrate that MDS cases that respond to luspatercept are associated with a relatively high level of expression of functionally deficient GATA1 short isoform. We further demonstrate that GDF11 stimulates the production of the GATA1 short isoform via SMAD2 binding to intron 2 of the GATA1 gene. The regulation of a GATA1 splicing isoform by GDF11/SMAD2 is associated with clinical responses to luspatercept, according to these novel findings.

A Pilot Multicentric Prospective Study on the Usefulness of Peripheral Blood Immunophenotyping As Erythroid Maturation Response Marker during Luspatercept Treatment in Patients Affected By ß-Thalassemia Major and SF3B1+ Myelodysplastic Syndrome

Introduction Erythropoiesis is a multi-step maturation process; luspatercept proved to be successful on ineffective erythropoiesis acting on the second phase of erythroid maturation reducing the hyper expression of transforming growth factor beta (TGF-beta) ligands via the Smad 2/3 signaling pathway, thus ameliorating the aberrant late stage erythroid maturation featured in the myelodysplastic syndrome (MDS) and thalassemia. MEDALIST and BELIVE studies showed efficacy and safety of Luspatercept in MDS and Thalassemic patients respectively, but, regarding erythroid maturation, a significant correlation has been revealed only between hematological response and the reticulocyte value or medullar erythroblasts counts. In order to avoid for patients extra scheduled bone marrow aspirate , in particular in thalassemic patients, in our pilot study we propose to explore, for the first time, the role of peripheral blood immunophenotyping (IF) as reproducible and more affordable erythroid maturation's marker respect to the medullar immunophenotyping described in the previous published papers. Our hypothesis is that peripheral IF could represent a new tool to detect the ameliorate of inefficacy erythropoiesis during Luspatercept therapy, acting as “minimal residual disease” in patients in response or that loss the response to Luspatercept. Patients, Material and Methods Approval by local Ethic Scientific Commission has be completed. Patients will be enrolling by Hematology Unit and by Hemoglobinophaties Unit from 2 different Institution in Italy (“A. Businco” Oncologic Hospital and “A. Cao” Hemoglobinopathies Hospital). Inclusion criteria: Patients must be affected by MDS with ringed sideroblasts (with or without SF3B1 mutation) very low, low and intermediate revised IPSS stage or affected by ß-thalassemia major; aged ≥ 18 years and able to understand and sign an informant consent; not renal or hepatic failure; EF &amp;gt; 50%. Duration of the study will be scheduled for 1 year. Prevision of 30 patient enrolled in the first 1 year of study. Primary endpoint is to confirm the role of peripheral blood immunophenotyping (IF) as reproducible and affordable erythroid maturation's marker in these setting of patients. Secondary endpoint is to evaluate the role of peripheral IF as “minimal residual disease” in particular during the loss of Lusplatercept response and also the assessments of iron homeostasis in responder's patients. Patients will be assessed for r-IPSS and NGS (only for MDS), endogenous erythropoietin, full blood count and reticulocyte, iron status, transfusion burden in the previous 24 weeks and peripheral IF (centralized in the laboratory of Hematology Unit). All patients will collect these samples at baseline timing and after 12,24,36 and 48 weeks. Peripheral IF is tested by 8-color EuroFlow antibody panels define progressive terminal stages of erythroblasts' maturation (10 health samples have been used as control). Hematological response and Luspatercept's dose escalation are considered as for MEDALIST study criteria for MDS and BELIVE study for thalassemia, as well adverse events. Results At the time of this report 26 patients have been enrolled in the two institution (10 MDS and 16 thalassemic patients). Baseline patients' characteristics are reported in TAB 1 and baseline peripheral blood erythroid maturation (based on centralized IF) are reported in TAB 2 Conclusion Our pilot study aims to explore, for the first time, the role of peripheral blood immunophenotyping (IF) as reproducible and affordable erythroid maturation's marker during Luspatercept therapy in Thalassemic and MDS patients.

Clonal Hematopoiesis-Related Mutations Are Associated with Favorable Clinical Benefit Following Luspatercept Treatment in Patients with Lower-Risk Myelodysplastic Syndromes: A Subgroup Analysis from the Phase 3 COMMANDS Trial

Introduction: Clonal hematopoiesis of indeterminate potential (CHIP) is a common age-related condition arising from somatic mutations in hematopoietic stem cells that leads to mutant cell clonal progeny in the blood (variant allele frequency [VAF] &amp;gt; 2%) and is found in 10%-30% of people &amp;gt; 70 years of age. CHIP-related mutations (eg, DNA methyltransferase 3 alpha [ DNMT3A], tet methylcytosine dioxygenase 2 [ TET2], ASXL transcriptional regulator 1 [ ASXL1], splicing factor 3b subunit 1 [ SF3B1]) are associated with increased inflammation and elevated risk of developing hematologic malignancies, including myelodysplastic syndromes (MDS)/acute myeloid leukemia, and are an independent risk factor for cardiovascular complications. Recently, patients with MDS were reported to have high rates of cardiovascular co-morbidities including elevated levels of N-terminal pro-brain natriuretic peptide (NT-proBNP), a marker of cardiovascular and all-cause mortality. Currently, no treatments targeting CHIP mutations or associated complications are available. Luspatercept is currently approved for anemia in adult patients with lower-risk MDS (LR-MDS) with ring sideroblasts who require red blood cell (RBC) transfusions (≥ 2 RBC units/8 weeks) after erythropoiesis-stimulating agent (ESA) treatment failure. Recently, the phase 3 COMMANDS trial (NCT03682536) reported superiority of luspatercept over ESAs in reducing transfusion burden in ESA-naive patients with LR-MDS. Aims: We investigated the association between the most frequent CHIP-related mutations ( DNMT3A, TET2, ASXL1, SF3B1) and clinical outcomes in patients treated with luspatercept from the COMMANDS trial. Methods: Genomic DNA was isolated from bone marrow (BM) mononuclear cells, and 36 myeloid-specific somatic gene mutations were identified by targeted next-generation sequencing (400X; sensitivity 3%; Munich Leukemia Laboratory information system) at baseline. Hematologic parameters (eg, hemoglobin [Hb], complete blood counts), transcriptome data (bulk BM RNA sequencing [RNA-Seq] data) and cytokines/chemokines (peripheral blood) were assessed at baseline and week 24. Results: Overall, 145 patients in the luspatercept arm of the COMMANDS trial were included in the current analyses, of whom 85% (123/145) (Figure 1A) had the most frequent CHIP-related mutations (eg, DNMT3A, TET2, ASXL1, and SF3B1) with a VAF of 3%-50%. Modest positive correlations were observed between DNMT3A, TET2 and SF3B1 baseline VAF and monocyte percentage ( P = 0.024, P = 0.015, and P = 0.005, respectively). Baseline Hb levels were lower in patients with frequent CHIP-related mutations than in those with less frequent/other somatic mutations. Luspatercept treatment significantly reduced anemia from baseline in patients with CHIP-related mutations (primary endpoint response 62%; n/N = 76/123). Concomitant with the Hb increase (∆ = +1.0 mg/dL; P &amp;lt; 0.001), we observed improvements in other cell counts, including white blood cell counts (∆ = +0.7 × 10 9 cells/L; P = 0.01). In addition, luspatercept treatment in patients with CHIP-related mutations was associated with downregulation of inflammatory gene signatures (Figure 1B) and proinflammatory cytokines/chemokines, including hepcidin (∆ = −43.9 ng/mL; P &amp;lt; 0.001). Furthermore, we observed upregulation of an anti-inflammatory regulator (growth/differentiation factor 15) following treatment. Most importantly, significant downregulation of NT-proBNP levels (∆ = −330 pg/mL; P &amp;lt; 0.05) was observed in patients with LR-MDS who received luspatercept. Conclusions: This retrospective subgroup analysis of patients in the COMMANDS trial with the most frequent CHIP-related mutations ( DNMT3A, TET2, ASXL1 and SF3B1) demonstrated novel effects of luspatercept in clonal hematopoiesis-related consequences including improvements in anemia, cytopenias, and reduced inflammation. Most importantly, NT-proBNP and elevated hepcidin levels were significantly downregulated in responders following luspatercept treatment. These results warrant evaluation of luspatercept in patients with high-risk CHIP mutations including clonal cytopenia of undetermined significance and patients with anemia of inflammation.

Long-Term Evaluation of Luspatercept in Erythropoiesis-Stimulating Agent (ESA)-Intolerant/Refractory Patients (pts) with Lower-Risk Myelodysplastic Syndromes (LR-MDS) in the Phase 3 MEDALIST Study

Background: Pts with LR-MDS receiving luspatercept in the phase 3 MEDALIST trial previously reported durable, long-term red blood cell (RBC) transfusion independence (RBC-TI) (Fenaux P, et al. J Clin Oncol 2022;40[suppl 16]. Abstract 7056). MEDALIST pts were eligible to enroll in a follow-up study to evaluate long-term efficacy and safety. Aims: To report long-term efficacy and safety outcomes in ESA-refractory/intolerant pts with LR-MDS treated with luspatercept in the MEDALIST trial with an additional 26 mo of follow-up from the final MEDALIST data cutoff. Methods: Eligible pts were ≥ 18 y of age; IPSS-R-defined Very low-, Low-, or Intermediate-risk MDS with ring sideroblasts; were refractory, intolerant, or unlikely to respond to ESAs, and required regular RBC transfusions. Pts were randomized 2:1 to luspatercept (starting dose 1.0 mg/kg, with titration up to 1.75 mg/kg) or placebo. Efficacy outcomes include RBC-TI ≥ 8 wk and ≥ 16 wk. Cumulative duration of response was determined by Kaplan-Meier (KM) analysis. Rates of longest single and cumulative duration of RBC-TI ≥ 1 y and exposure-adjusted incidence rates (EAIR) of treatment-emergent adverse events (TEAEs) and AEs of special interest (AESIs) were calculated. Rates of high-risk (HR)-MDS and acute myeloid leukemia (AML) progression were calculated from LR-MDS diagnosis to HR-MDS/AML diagnosis, or to last HR-MDS/AML follow-up date for pts who did not progress. Results: As of January 2, 2023, 11 pts remained on luspatercept treatment; 8 pts randomized to luspatercept and none from placebo continued in post-treatment follow-up. The median (range) follow-up time was 39.9 (2.8-76.0) mo for luspatercept arm pts and 38.7 (1.7-68.6) mo for placebo arm pts. The median (range) duration of treatment was 50.9 (5.9-332.9) wk for luspatercept pts and 24.0 (9.0-103.0) wk for placebo pts. Seventy-five of 153 (49.0%) luspatercept pts achieved RBC-TI ≥ 8 wk during the entire treatment period and 53/75 (70.7%) had ≥ 2 response periods with a median cumulative duration of 110.14 wk (95% confidence interval [CI], 53.71-154.14; Figure A). Overall, 31/75 (41.3%) luspatercept pts had the longest RBC-TI duration ≥ 1 y, and 44/75 (58.7%) had a cumulative RBC-TI duration ≥ 1 y. In total, of the 48/153 (31.4%) pts receiving luspatercept who achieved RBC-TI ≥ 16 wk during the entire treatment period, 32/48 (66.7%) had ≥ 2 response periods with a median cumulative duration of 129.29 wk (95% CI, 79.86-240.43). Grade 3/4 treatment-related TEAEs were reported in 14/153 (9.2%; EAIR, 5.3 per 100 pt-y [PYs]) and 3/76 (3.9%; EAIR, 6.6 per 100 PYs) luspatercept and placebo pts, respectively; TEAEs of any grade leading to permanent treatment discontinuation were reported in 24/153 (15.7%; EAIR, 8.6 per 100 PYs) luspatercept and 6/76 placebo pts (7.9%; EAIR, 13.0 per 100 PYs). The most common TEAEs of any grade reported in &amp;gt; 5% of luspatercept pts were fatigue (47/153 [30.7%]; EAIR, 22.1 per 100 PYs; vs placebo, 11/76 [14.5%]; EAIR, 27.1 per 100 PYs), diarrhea (47/153 [30.7%]; EAIR, 23.5 per 100 PYs; vs placebo, 8/76 [10.5%]; EAIR, 18.4 per 100 PYs), asthenia (41/153 [26.8%]; EAIR, 18.8 per 100 PYs; vs placebo, 9/76 [11.8%]; EAIR, 21.5 per 100 PYs), peripheral edema (40/153 [26.1%], EAIR, 17.6 per 100 PYs; vs placebo, 13/76 [17.1%]; EAIR, 32.0 per 100 PYs), and back pain (38/154 [24.7%]; EAIR, 16.8 per 100 PYs; vs placebo, 5/76 [6.6%]; EAIR, 11.7 per 100 PYs; Figure B). During the entire treatment period, 26/44 (59.1%), 9/23 (39.1%), and 12/86 (14.0%) pts who received ≤ 1 mg/kg, 1.33 mg/kg, and 1.75 mg/kg luspatercept, respectively, reported fatigue of any grade. No new progression to HR-MDS or AML events have been reported since the last data cutoff (Santini V et al. Blood 2022;140[suppl 1]:4079-4081). Conclusions: Pts with LR-MDS who received luspatercept for &amp;gt; 2 years longer than in the original MEDALIST study, continued to experience sustained periods of RBC-TI with more than half of pts experiencing cumulative RBC-TI for ≥ 1 year. The long-term safety profile of luspatercept is consistent with previous shorter-term reports, and AEs were mostly lower grade with rates of fatigue decreasing with increasing luspatercept dose. These data demonstrate the long-term efficacy and safety of luspatercept in pts with LR-MDS intolerant/ineligible for ESA treatment.

Efficacy and Safety of Luspatercept Versus Epoetin Alfa in Erythropoiesis-Stimulating Agent (ESA)-Naive Patients (Pts) with Transfusion-Dependent (TD) Lower-Risk Myelodysplastic Syndromes (LR-MDS): Full Analysis of the COMMANDS Trial

Introduction: ESAs are an established treatment (tx) for pts with TD LR-MDS and endogenous serum erythropoietin (sEPO) levels ≤ 500 U/L; however, eligible pts often do not respond, or the response duration is limited. Luspatercept is approved in the US and EU to treat anemia due to LR-MDS after ESA failure. The preplanned interim analysis of the phase 3 COMMANDS trial (NCT03682536), comparing luspatercept with epoetin alfa in ESA-naive TD pts with anemia due to LR-MDS (with or without ring sideroblasts [RS]) showed for the first time the superiority of another therapy over ESAs in improving red blood cell transfusion independence (RBC-TI) rates (Platzbecker U, et al. Lancet 2023. doi:10.1016/S0140-6736[23]00874-7). Here we report the full efficacy and safety analysis of the COMMANDS trial. Methods: Eligible pts were ≥ 18 y of age, had Revised International Prognostic Scoring System-defined LR-MDS with &amp;lt; 5% bone marrow blasts, sEPO &amp;lt; 500 U/L, and were TD (received 2-6 RBC U/8 wk for ≥ 8 wk before randomization). Pts were randomized 1:1 to luspatercept (1.0-1.75 mg/kg) subcutaneously (SC) Q3W, or epoetin alfa (450-1050 IU/kg) SC Q1W for ≥ 24 wk and stratified by baseline RBC transfusion burden (&amp;lt; 4 vs ≥ 4 RBC U/8 wk), RS status (RS+ vs RS−), and sEPO (≤ 200 vs &amp;gt; 200 U/L). The primary endpoint was the achievement of RBC-TI ≥ 12 wk with a concurrent mean hemoglobin (Hb) increase ≥ 1.5 g/dL (wk 1-24). Key secondary endpoints (wk 1-24) included achievement of RBC-TI ≥ 12 and for 24 wk and hematologic improvement-erythroid (HI-E) ≥ 8 wk. Other endpoints included duration of RBC-TI ≥ 12 wk, progression to acute myeloid leukemia (AML), and safety. Results: As of Mar 31, 2023, 182 pts were randomized to luspatercept and 181 to epoetin alfa, with a median (range) tx duration of 51.3 (3-196) and 37.0 (1-202) wk and median (range) follow-up of 17.2 (1-46) and 16.9 (0-46) months, respectively. The primary endpoint was achieved by110 (60.4%) pts in the luspatercept arm versus 63 (34.8%) in the epoetin alfa arm ( P &amp;lt; 0.0001). Subgroup analysis of the primary endpoint response showed that response rates achieved with luspatercept versus epoetin alfa, respectively, were greater for SF3B1-mutated and non-mutated pts (Table), greater for pts with baseline sEPO ≤ 200 U/L and with baseline sEPO &amp;gt; 200 to &amp;lt; 500 U/L (Table), and greater for RS+ pts (Table). The response rates were comparable between tx arms for RS− pts (Table). RBC-TI ≥ 12 wk was achieved by 124 (68.1%) and 88 (48.6%) pts in the luspatercept and epoetin alfa arms, respectively, RBC-TI for 24 wk by 87 (47.8%) and 56 (30.9%) pts, and HI-E ≥ 8 wk by 135 (74.2%) and 96 (53.0%) pts. The median (95% CI) duration of RBC-TI ≥ 12 wk was longer with luspatercept versus epoetin alfa (128.1 wk [108.3-not estimable (NE)] versus 89.7 wk [5.9-157.3]; HR, 0.534; Figure), and longer for clinically relevant subgroups, including RS+ and RS−. The median duration of tx was longer in the luspatercept arm compared with the epoetin alfa arm (51.3 versus 37.0 wk). Five (2.7%) and 6 (3.3%) pts in the luspatercept and epoetin alfa arms, respectively, progressed to AML. Overall, 178 (97.8%) and 165 (92.2%) pts receiving luspatercept and epoetin alfa reported tx-emergent adverse events (TEAEs) of any grade; 107 (58.8%) and 88 (49.2%) pts reported grade 3/4 TEAEs, respectively. The most common any-grade TEAEs in either arm (≥ 10% of pts) were diarrhea (17.6% luspatercept vs 14.0% epoetin alfa), COVID-19 (14.8% vs 15.6%), asthenia (13.7% vs 16.2%), and anemia (12.1% vs 10.6%). TEAEs of interest were reported by 105 (57.7%) and 81 (45.3%) pts receiving luspatercept and epoetin alfa, respectively, with asthenia including fatigue, malaise, and lethargy (30.8% vs 24.6%), hypertension (15.9% vs 9.5%), malignancies and premalignant disorders (14.3% vs 12.8%), kidney toxicity (8.8% vs 6.7%), and injection site reactions (6.6% vs 2.2%), occurring in &amp;gt; 5% pts. In both tx arms, rates of on-tx deaths (8.2% for luspatercept and 7.2% for epoetin alfa) and post-tx deaths (13.2% and 13.4%) were similar. Conclusions: Results of this full analysis confirm the findings from the interim analysis; RBC-TI duration and erythroid responses achieved with luspatercept are superior compared with epoetin alfa. Luspatercept safety results were consistent with previous MDS studies. These data show that luspatercept could represent a new standard of care for pts with TD LR-MDS.

Clinical Text Reports to Stratify Patients Affected with Myeloid Neoplasms Using Natural Language Processing

Background: The availability of multimodal patient data, such as demographics, clinical, imaging, treatment, quality of life, outcomes and wearables data, as well as genome sequencing, have paved the way for the development of multimodal clinical solutions that introduce personalized or precision medicine. The clinical report is an information layer that contains relevant information about the disease in addition to the patient's point of view. Natural language processing (NLP) is a branch of artificial intelligence (AI) and its pre-trained language models are the key technology for extracting value from this data layer. Aims: This project was conducted by GenoMed4all and Synthema EU consortia, with the aim to: 1) Build an AI language model specific for the hematology domain. 2) Use NLP technology to extract relevant information from clinical reports and perform unsupervised stratification of patients, in order to 3) demonstrate that the clinical report is earlier access to data relative to disease clinical phenotype and biology and provide important information for patient stratification and prediction of clinical outcomes. Methods: To translate text sentences into numerical embeddings, we implemented bidirectional encoder representations from transformers (BERT) framework. To learn text representations and correlations within data, we performed domain-adaptation by fine-tuned pre-trained model on hematological clinical reports of patients with myeloproliferative neoplasms (MPN), myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML). Patient stratification was performed by HDBSCAN clustering on text embedding encoded by BERT (HematoBERT). Clusters validation was performed by assessing patients' diagnosis and survival probability. Finally, we compared domain-tuned HematoBERT vs pre-trained non-contextualized models. Results: We implemented HematoBERT based on the bert-base-multilingual-uncased version of BERT. Training data were hematological text reports of 1,328 patients. During fine-tuning, texts were tokenized, then we randomly replaced 15% of the tokens with masked tokens, training the model to predict them. We performed stratification using clinical reports from a validation cohort of 360 patients. We identified 7 clusters, defined according to similar words in meaning that were placed in a specific topic. We extracted the most important words and concepts for each cluster (topic) and we summarized them into effective descriptions for each group of patients. Two clusters included MDS patients with excess blasts, and without excess blasts with ring sideroblasts and del5q (n=69, n=115). One cluster included patients with excess blasts and MDS/MPN (n= 33). Two clusters included MPN patients with primary and secondary myelofibrosis, and MPN patients most including subjects affected with polycythemia vera and essential thrombocythemia (n=35, n=46). Two clusters included patients with AML from MDS and therapy-related AML, and patients with de novo AML (n=22, n=42). Clinical validation was performed based on the diagnosis and survival probability of patients assigned to clusters. Patients' diagnoses were compatible with the cluster assignment (Figure 1). Frequency of gene mutations (as assessed by targeted Next-Generation Sequencing) among different clusters reflected the well-known genotypic-phenotypic associations in MDS, MPN and AML. Kaplan-Maier curves indicated significative risk stratification in clusters in terms of survival probability (Figure 2), similar to stratifications performed on clinical and genomic data. Finally, we evaluate the domain adaptation by comparing the model to other pre-trained non-contextualized ones. Pseudo perplexity score (PPS), accuracy and F1 score were calculated to quantify how good the models are when they see new data, predicting the next word given the context of the sentence. HematoBERT obtained high PPS, accuracy and F1 scores, outperforming the other models also trained on generic clinical domains. Conclusion: Domain-adapted language models are able to understand contexts and correlations in documents. HematoBERT can be used to extract relevant features from clinical reports. This data layer is relevant to perform disease stratification of patients based on clinical and genomic information and could be integrated into next-generation multimodal models of personalized medicine.

STAG2 Somatic Mutations Are Associated with Specific Dysplastic Megakaryocytic and Myeloid Cell Features in Myelodysplastic Syndrome

Introduction: Although myelodysplastic neoplasms/myelodysplastic syndromes (MDS) are driven by genetic mutations, their diagnosis relies on morphologic evaluation of abnormal hematopoiesis in the bone marrow. Only a small number of genetic abnormalities define bone marrow morphologic features in MDS, such as those harboring SF3B1 mutations and deletions of chromosome 5q. We hypothesized that there are additional genetic alterations and dysplastic morphologic features in MDS that have not been well-described in the literature. We assessed genetic morphologic associations between 50 commonly mutated genes and 10 morphologic features in a cohort of MDS bone marrows with a high degree of dysplasia. We validated and extended these findings in an independent cohort of myeloid neoplasms. STAG2 mutations were strongly associated with specific dysplastic megakaryocytic and myeloid cell features in both cohorts. Methods: We reviewed bone marrow core biopsies and aspirates from 272 MDS clinical cases (2013-2014) and selected 89 for further study based on the presence of prominent dysplastic features. Each case was independently evaluated by two pathologists for the following dysplastic features: small hypolobated megakaryocytes, widely separated megakaryocyte nuclei; abnormal myeloid nuclear segmentation, myeloid hypogranulation, Auer rods; ring sideroblasts, erythroid binucleation, irregular erythrocyte nuclei, erythroid nuclear cytoplasmic dyssynchrony, and erythroid karyorrhexis. Targeted next generation sequencing was performed using a 50 gene panel comprising the most common somatic myeloid mutations. For each dysplastic feature, univariate risk analysis was used to identify genetic mutations for use in multivariable analysis. Significant associations were validated through blinded morphologic review in an independent cohort of 155 bone marrow biopsies of myeloid neoplasms, comprised of 83 myelodysplastic syndromes, 33 myelodysplastic/myeloproliferative neoplasms and 39 acute myeloid leukemias, diagnosed between 2014 and 2019 and selected for mutations identified in the initial discovery cohort ( STAG2, RUNX1, SRSF2, ASXL1 and SETBP1) . Results: Pairwise analyses demonstrated broad patterns of genetic associations with myeloid/megakaryocytic dysplasia versus erythroid dysplasia (Panel 1). In multivariable analysis, STAG2 and SRSF2 mutations were significantly associated with megakaryocytes with widely separated nuclei ( STAG2, OR = 25.5, 95% CI [4.17, 493.5], P = 0.003; SRSF2, OR = 10.1, 95% CI [2.07, 75.0], P = 0.008). Mutations in STAG2 and SETBP1 were significantly associated with abnormal myeloid nuclear segmentation ( STAG2, OR = 7.08, 95% CI [1.94, 28.0], P = 0.003; SETBP1, OR = 12.2, 95% CI [1.29, 267.4], P = 0.04). Furthermore, STAG2 mutations were significantly associated with myeloid cell hypogranulation (OR = 12.7, 95% CI [3.10, 86.3], P = 0.002). Because this initial cohort was relatively small, which led to wide confidence intervals suggesting estimation instability, we validated and extended these findings in an independent cohort of myeloid neoplasms. In the validation cohort (enriched for mutations in STAG2, SRSF2, and SETBP1), STAG2 mutations were significantly associated with separated megakaryocyte nuclei (univariate OR = 4.18, 95% CI [1.98, 8.80], P = 0.0002), abnormal myeloid nuclear segmentation (univariate OR = 3.47, 95% CI [1.47, 8.22], P = 0.008), and myeloid cell hypogranulation (multivariable OR = 5.28, 95% CI [2.18, 13.4], P = 0.0003) (Panel 2). In contrast, morphologic associations with SRSF2 and SETBP1 mutations were not validated. Conclusion: STAG2 genetic alterations are strongly associated with the presence of separated megakaryocyte nuclei, abnormal myeloid nuclear segmentation and myeloid cell hypogranulation in MDS. Our data suggest that STAG2 may affect pathways involved in nuclear shape control and secondary granule formation. Panel 1: Pairwise associations between gene mutation and bone marrow morphology in MDS. N, number. Panel 2: Multivariable (black) and univariate (white) analysis in two independent cohorts showing significant gene associations with megakaryocytes with separated nuclei, abnormal myeloid nuclear segmentation, and myeloid cell hypogranulation. CI, confidence interval.