Jain, Akriti Gupta

6501 Background: Hepcidin, a central regulator of iron homeostasis, is pathologically elevated in patients with myelofibrosis (MF) and anemia. Chronic hepcidin elevation limits iron availability for red blood cell (RBC) production and contributes to anemia onset and severity. DISC-0974 is an investigational, first-in-class, monoclonal antibody that blocks hemojuvelin, a co-receptor in the bone morphogenetic protein-signaling pathway driving hepcidin expression. Methods: RALLY-MF (NCT05320198) is an ongoing Phase 2, open-label study conducted in the US. The primary aim is to evaluate DISC-0974 efficacy in participants with MF and anemia. Secondary aims include evaluating safety, pharmacokinetics (PK), and pharmacodynamics (PD). Eligible participants are ≥18 years of age with primary or secondary MF and hemoglobin (Hgb) <10 g/dL or RBC transfusion requirement. Participants enroll into 3 cohorts based on RBC units transfused in the 84 days prior to screening: non-transfusion dependent (nTD, 0 units and Hgb <10 g/dL), low transfusion burden (TD low, 1-2 units), high transfusion burden (TD high, 3-12 units). Stable dose of concomitant Janus kinase inhibitor (JAKi) is allowed. DISC-0974 is administered subcutaneously at 50 mg monthly for up to 6 doses, with escalation to 75 mg for inadequate or lost response. Primary endpoints include major hematologic response defined as transfusion independence (TI) during any consecutive 16 weeks (TD low) or 12 weeks (TD high), or a mean Hgb increase of ≥1.5 g/dL from baseline for ≥12 weeks (nTD). Secondary endpoints include PK/PD markers of iron regulation and safety assessments. Data were summarized using descriptive statistics. Results: At data cut (16 Oct 2025), 47 participants enrolled in the nTD (n=30), TD low (n=10), and TD high (n=7) cohorts, with 53% on concomitant JAKi therapy. DISC-0974 led to sustained hepcidin reduction and iron mobilization across cohorts. DISC-0974 resulted in meaningful hematologic responses among evaluable participants: 50% of nTD participants achieved a mean Hgb increase of ≥1.5 g/dL for ≥12 weeks, 71% of TD low participants achieved TI for ≥16 weeks, and 67% of TD high participants experienced a ≥50% reduction in transfusion burden, with follow-up ongoing. Major hematologic response rates were 50% for participants on and off concomitant JAKi therapy (n=18 and n=16, respectively). DISC-0974 was associated with meaningful improvement in FACIT-Fatigue scores for nTD and TD low participants. Serious adverse events (AEs, n=9) and ≥Grade 3 AEs (n=16) were considered unrelated to DISC-0974. There were no early withdrawals due to AEs. Conclusions: DISC-0974 was well tolerated and shows a favorable safety profile. Anemia response and fatigue scores improved across patients with MF and anemia, including those receiving JAKi therapy. These data validate hepcidin reduction as a promising approach to treating anemia in MF. Clinical trial information: NCT05320198 .

Abstract Introduction: Despite advent of multiple tyrosine kinase inhibitors (TKIs) and newer treatment methods for AML and ALL, the outcomes of patients (pts) with chronic myeloid leukemia in blast phase (CML-BP) remain poor. Consensus guidelines on treatment of CML-BP are scarce. A recent study from the European leukemia network blast phase registry on 240 pts reported marked heterogeneity in treatment patterns and lack of an accepted standard. Hence, we undertook this study through the H Jean Khoury Cure CML Consortium (HJKC3) to describe the treatment patterns and outcomes of pts with CML-BP in the US. Methods: This is a multi-institution, retrospective study. IRB approval was obtained at each institution and data sharing agreements were executed. All pts diagnosed with CML-BP based on modified MD Anderson Criteria (≥30% blasts in the blood or bone marrow or extramedullary (EM) disease) from 2010 to 2025 were included. Categorical variables were compared using Chi-square test. Overall survival (OS) was calculated from time of diagnosis of CML BP to death from any cause. Event free survival (EFS) was calculated from date of best response to event (death or relapse). Results: Of the 283 pts with CML-BP, 96 (33.9%) were female and 187 (66.1%) were males. At the time of diagnosis, 183 (64.6%) had CML-CP, 21 (7.4%) pts had CML-AP and 79 (27.9%) pts had de-novo CML-BP. For the 204 pts that were initially diagnosed with CML-CP or CML-AP, the median time to progression to CML-BP was 22.5 months (1-207 months). Median age at diagnosis of CML-BP was 50 years (range 16-86 years). EM disease was present in 78 (27.6%) pts (38 myeloid sarcoma, 30 central nervous system and 10 skin). Blasts were of myeloid phenotype in 168 (59.4%), lymphoid in 103 (36.4%) and mixed in 9 (3.2%) pts. The bcr-abl protein was p210 in 229 (80.9%), p190 in 8 (2.8%) and p230 in 2 (0.7%) pts. ABL kinase domain mutation was seen in 42.1% (80/190) pts, with T315I being the most common (n=32), followed by E255K (n=16). Next generation sequencing was performed in 117 pts. The most common mutations were ASXL1 (n=27), RUNX1 (n=14), BCOR (n=4), DNMT3A (n=4) and BCORL1 (n=3). Of the 283 pts, 269 underwent CML-BP directed therapy. 63 (23.4%) pts were treated with TKI monotherapy, 184 (68.4%) pts were treated with a combination of chemotherapy (chemo) and TKI and 22 (8.2%) pts were treated with chemo alone. Response rate after CML-BP directed therapy was 75.3% (n=213). Best response was return to CML-CP in 49 pts, CHR with PCyR in 18 pts, CCyR in 31 pts, MMR in 31 pts, MR4 in 18 pts, MR4.5 in 19 pts and MR5 in 47 pts. Rate of response with 2G and ponatinib was significantly better than imatinib (85.3% and 77.3% vs 63.6%; p=0.04). 149 (55.4%) pts proceeded to alloSCT. Of the 213 pts that responded, 59 (27.7%) relapsed. At a median follow-up of 47 months, the median OS for all pts was 19.7 months (95% CI: 16.3-40.2 months), with 5-year OS rate of 38% (95% CI: 32-45%). Pts with myeloid phenotype had a significantly lower OS compared to lymphoid and mixed phenotype (Median OS 14.9 vs 45.9 and 31.1 months respectively; p=0.02). OS of pts that were treated with a combination of TKI and chemo was significantly longer compared to pts treated with TKI or chemo alone (52.9 vs 16.1 or 7.5 months; p&lt;0.0001). Dasatinib (n=99; 34.9%) and ponatinib (n=97; 34.3%) were the most used TKIs followed by imatinib (n=23; 8.1%), nilotinib (n=21; 7.4%), bosutinib (n=8; 2.8%) and asciminib (n=3; 1.1%). Pts treated with ponatinib had a numerically longer OS compared to 2G-TKI and imatinib, however this was not statistically significant (40.1 vs 30.6 and 20.3 months; p=0.52). EFS for pts treated with ponatinib was significantly longer compared to 2G-TKI or imatinib (25.9 vs 17.6 and 15.9 months; p=0.03). In a landmark analysis, pts that underwent alloSCT within 6 months of diagnosis had a significantly longer OS compared to those who did not (69.3 vs 16.0 months; p=0.006). Conclusion: CML-BP continues to have poor prognosis. Treatment is heterogenous involving various combinations of TKI and chemo. In our analysis, treatment with 2G-TKI or ponatinib led to better responses. CML-BP with myeloid phenotype had worse OS. Treatment with ponatinib was associated with a numerically longer OS and significantly longer EFS compared to 2G-TKI and imatinib. In addition, treatment with chemo in combination with TKI and alloSCT led to improved OS.

Abstract Background Hepcidin, a central regulator of iron homeostasis, is pathologically elevated in patients with myelofibrosis (MF) and anemia. Chronic hepcidin elevation limits iron availability for red blood cell production and contributes to the onset and severity of anemia, for which there is a large unmet need for safe and effective treatments. DISC-0974 is an investigational, first-in-class, monoclonal antibody that blocks hemojuvelin, a co-receptor in the bone morphogenetic protein-signaling pathway driving hepcidin expression. In studies involving healthy volunteers (Novikov, et al, J Clin Pharmacol 2024:64(8):953-962) and participants with MF (Gangat et al, Blood 2024;144(Suppl1):657), DISC-0974 dosing resulted in reduced serum hepcidin levels and increased serum iron, with a favorable safety profile. Clinically meaningful improvements in anemia were also observed: 50% of non–transfusion-dependent (nTD) participants achieved a mean hemoglobin increase of ≥1.5 g/dL for at least 12 weeks, while 80% of participants with low transfusion burden (TD low) and 40% of participants with high transfusion burden (TD high) achieved transfusion independence (TI) for ≥16 and ≥12 consecutive weeks, respectively. Aims This Phase 2, open-label, multiple-ascending dose study (NCT05320198) assesses efficacy, safety, pharmacokinetics (PK), and pharmacodynamics (PD) in participants with MF. Methods Participants were enrolled into 3 primary cohorts (n=30 per cohort), categorized by the units of red blood cells transfused in the 84 days prior to screening: nTD (hemoglobin &lt;10 g/dL with 0 units), TD low (1-2 units), and TD high (3-12 units transfused). Additionally, an exploratory cohort of participants receiving concomitant treatment with momelotinib or pacritinib (n=10) was included. Eligible participants were ≥18 years of age with hemoglobin &lt;10 g/dL or a transfusion requirement meeting World Health Organization diagnostic criteria for primary or secondary MF. A stable dose of concomitant hydroxyurea and/or Janus kinase (JAK) inhibitor (ruxolitinib and fedratinib for primary cohorts; momelotinib and pacritinib for exploratory cohort) was allowed. Major exclusion criteria included liver iron concentration ≥7 mg/g dry weight (TD high only) and nutritional, genetic, infectious, or autoimmune anemia. DISC-0974 was administered subcutaneously at a dose of 50 mg once monthly for a total of 6 doses. Dose escalation to 75 mg was required for inadequate or loss of response. Primary endpoints included evaluating hematologic response defined for TD low and TD high as TI during any 16- and 12-week consecutive period, respectively, and for nTD as a mean hemoglobin increase of ≥1.5 g/dL from baseline for at least 12 weeks. Secondary endpoints included PK/PD markers of iron regulation and safety assessments. Data were summarized using descriptive statistics. Results As of 01 July 2025, enrollment of the exploratory cohort of momelotinib was complete (n=10). Baseline hepcidin values in the exploratory cohort were similar (mean (SD) of 83.6 (62.8) ng/mL) to baseline hepcidin values from the DISC-0974 Phase 1b study (mean (SD) of 85.8 (71.0) ng/mL, n=35), and sustained hepcidin reduction was observed after DISC-0974 dosing. Early hematologic responses observed after 1 to 3 doses in the exploratory cohort were comparable to those reported in the Phase 1b study (Gangat et al, Blood 2024;144(Suppl1):657). Based on these initial data, the protocol was amended to allow concomitant momelotinib or pacritinib in all cohorts. Preliminary results, including baseline characteristics, PD, and hematologic responses across all cohorts, will be presented.

Abstract Somatic mutations in hematopoietic stem and progenitor cells (clonal hematopoiesis) are known to alter the immune milieu and can result in autoimmune disorders including VEXAS (Vacuoles, E1 enzyme, X-linked, Autoinflammatory, Somatic) syndrome. There are currently no curative strategies for CH-linked autoimmune disorders besides allogenic hematopoietic stem cell transplantation. Recent work has shown a high prevalence of autoimmune syndromes among patients with IDH1/2 mutant myeloid neoplasms (MN). We thus hypothesized that IDH1/2-mutant CH might drive a subset of autoimmune disorders. To further characterize a possible association between IDH-mutant CH and autoimmune disease, we used data from two observational cohort studies, the UK Biobank (N=454,218) and All of Us (N=414,830). We observed 264 individuals with IDH1/2-mutant CH. IDH1/2 CH was associated with autoimmune disease (OR=1.8, p=4e-3), and, when compared to other CH genes with at least 100 carriers, showed the strongest effect size. When comparing autoimmune subtypes, we observed significant associations with polymyalgia rheumatica (OR=4.0, p=1e-3) and giant cell arteritis (OR=8.1, p=1e-6). IDH1/2 CH carriers had elevated CRP levels compared to other CH-carriers (p=1e-2) and non-carriers (p=8e-3). IDH1/2 CH was associated with lower neutrophil counts and higher monocyte counts. While most (68%) had blood counts outside the normal range, only a minority (11%) had severe cytopenias that met criteria for CCUS. Since IDH1/2 CH is associated with both autoimmune disease and blood count abnormalities, we sought to define the frequency, clinical presentation, and outcomes of IDH1/2-associated autoimmune disease among individuals with bone marrow biopsy confirmed CCUS. We used data from three CCUS cohorts: first, a multi-institutional cohort of CCUS patients with systemic autoinflammatory disease (SAID) who were UBA1-negative, second, a hospital-based cohort at Mayo Clinic, and third, a cohort of IDH1 carriers enrolled on a phase II clinical trial of ivosidenib (NCT05030441). Within 193 SAID patients who were free of myeloid neoplasm at the time of recruitment, IDH1/2 pathogenic variants were the most observed CH driver, occurring in 2% of individuals. These individuals (N=4) all had recurrent fevers and dermatologic manifestations, requiring chronic glucocorticoid treatment with 75% having inflammatory arthritis requiring multiple Disease-Modifying Antirheumatic Drugs (range 1-7, median=5). Within the Mayo Clinic cohort, 275 patients with CCUS were serially ascertained. Among those 8 (2.9%) were IDH1-mutant of whom 50% had autoimmune disease (three with seronegative rheumatoid arthritis and one with Sweet syndrome). At last follow up (median 16 months), 2 (25%) of 8 patients had progressed to MN. Within our open-label decentralized trial of ivosidenib (IDH1 inhibitor) in CCUS, we enrolled 20 patients, of whom 5 (25%) had a diverse array of pre-existing autoimmune disorders, including two cases of seronegative rheumatoid arthritis and one case each of Sweet syndrome, polymyalgia rheumatica, and Sjögren's syndrome. Underlying the clinical presentation of all five cases was episodic joint pain, with half also experiencing episodic dermatologic manifestations. The majority (58%, 11/19) of the IDH1mt-CCUS patients had an elevated hsCRP (&gt;=5; range 6.91–46.2) including all four with co-occurring autoimmune disorders (range 7.97–46.2). Out of the 20 enrolled patients, 17 received at least 8 weeks of therapy, of which 9 had elevated hsCRP at baseline. Most of these (7/9) had normalization of hsCRP (falling below 5 mg/L) while on treatment (median time to hsCRP normalization of 58 days). Among the four patients with autoimmune disease, all had subjective improvement in their autoimmune symptoms following treatment with ivosidenib. In conclusion, we show that IDH1/2 CH is associated with autoimmune disorders and is particularly enriched among IDH1/2 CH carriers with co-occurring cytopenias. IDH1/2-associated autoimmune disease can range from classic rheumatic conditions such as rheumatoid arthritis to more severe VEXAS like presentations. IDH inhibition demonstrates efficacy for improving inflammatory manifestations within an ongoing clinical trial. Together these results nominate IDH mutant CH as a driver of autoimmune diseases with therapeutic implications.

Abstract BACKGROUND Ivosidenib is an oral inhibitor of mutant isocitrate dehydrogenase type I (IDH1) currently approved for the treatment of IDH1 mutated acute myeloid leukemia and myelodysplastic syndrome (MDS). Hematopoietic stem and progenitor cells carrying preleukemic mutations, referred to as clonal hematopoiesis (CH), are the origin of myeloid neoplasms (MN). Individuals who have cytopenias and CH, but do not meet diagnostic criteria for a hematologic malignancy, are classified as having Clonal Cytopenias of Undetermined Significance (CCUS), a condition associated with a high risk of progression to MN. While CCUS patients (pts) can have multiple mutations, they show lower mutational burden on average compared to MN. Thus, we hypothesized that treatment with ivosidenib would improve cytopenias in IDH1-mutant CCUS. METHODS; We performed a decentralized phase II single-arm study to characterize the safety and efficacy of ivosidenib in IDH1-mutant CCUS (NCT0503044). Drug was shipped to pts with local oncologists performing standard-of-care clinical exams supplemented by study team lead interviews for adverse event monitoring. Research labs were drawn by local providers or through Quest. QT corrected interval (QTc) was monitored through a mobile electrocardiogram device (AliveCor). Eligible patients were aged ≥ 18 years and had confirmed IDH1-mutant CCUS with at least 6 months of unexplained cytopenias (hemoglobin&lt;10 g/dL, platelets&lt;100 x109/L or absolute neutrophil count &lt;1.8 x109/L). The primary endpoint was hematologic response of at least 8 weeks in duration based on the IWG 2006 criteria for MDS. RESULTS 20 pts with IDH1-mutant CCUS have been enrolled. Most were male (70%) with a median age of 66 (range 41-79). Overall, 74% had a self-reported race of non-Hispanic white, 11% of Hispanic white, 5% of Asian and 11% of black or African American. Most pts presented with neutropenia (95%) including some with co-occurring thrombocytopenia (21%). One pt presented with anemia and thrombocytopenia without neutropenia. Prior to treatment, the average number of co-mutations beyond IDH1 was 2 (range 0-3), with most pts (74%) harboring a co-occurring spliceosome gene mutation. The IDH1 variant allele fraction (VAF) was lower in the blood (mean=19%, range 3-35%) than in the bone marrow (mean=31%, range 18-43%) in most (93%) patients. Out of the 20 enrolled pts, 17 received at least 8 weeks of therapy at the time of our data cutoff (7/29/2025) and were included in the efficacy analysis. Grade ≥3 adverse events (AEs) occurred in five patients (26%) including two with grade ≥3 treatment-related adverse events (pancytopenia and sepsis in the first pt, rash in the second pt). The most common all-grade treatment-related adverse events were nausea (n=5, 26%), prolonged QTc (n=3, 16%), anemia (n=3, 16%), and diarrhea (n=3, 16%). There were no instances of reported differentiation syndrome. Overall, 82% of pts achieved a hematologic response (14/17) including improvement in neutropenia (88%) and thrombocytopenia (75%). With a median follow-up of 16 months (range 0.6-24 months), the median duration of hematologic response was not reached. Among responders, 86% had a continued response at one year. Two pts progressed to MDS after 4 and 17 months of treatment characterized by recurrence of cytopenias and the emergence of new clones. A surveillance bone marrow biopsy was performed at 18 cycles for the 8 pts who remained on treatment with none showing disease progression. Most pts (90%, N=9/10) treated for at least 12 months had a decrease in IDH1-mutant CH VAF over time and 40% (N=4/10) showed IDH1 mutation clearance (VAF &lt;2%) with a median time to clearance of 9 months. Among those with decreasing IDH1 clonal burden, all had a corresponding decrease in other co-occurring mutations present at baseline with one responder showing emergence of a newly detectable CHEK2 mutation without disease progression. CONCLUSIONS Decentralized trials represent a feasible solution for the conduct of clinical trials in rare hematologic neoplasms and precursor conditions. Ivosidenib was well-tolerated and induced durable hematologic remissions in IDH1-mutant CCUS. Most pts showed a reduction in clonal burden with mutation clearance commonly seen with long-term treatment demonstrating that ivosidenib modifies the natural history of IDH1-mutant CCUS.

Introduction: Despite advent of multiple tyrosine kinase inhibitors (TKIs) and newer treatment methods for both AML and ALL, the outcomes of patients (pts) with chronic myeloid leukemia in blast phase (CML-BP) remain poor. Consensus guidelines on treatment of CML-BP are scarce. A recent study from the European leukemia network (ELN) blast phase registry reported outcomes of 240 CML-BP pts. They also reported marked heterogeneity in treatment patterns and lack of an accepted standard. Hence, we are undertaking this study through the H Jean Khoury Cure CML Consortium (HJKC3) to study the treatment patterns and outcomes of pts with CML-BP in the US. Methods: This is a multi-institution, retrospective study. IRB approval was obtained at each institution separately and data sharing agreements were executed. All pts diagnosed with CML-BP based on modified MD Anderson Criteria (≥30% blasts in the blood or bone marrow or extramedullary (EM) disease) from 2010 to 2024 were included. Categorical variables were compared using the Chi-square test. Overall survival (OS) was calculated from the time of diagnosis to death from any cause. Survival probabilities were computed using the Kaplan-Meier method and compared with the log-rank test. Results: Of the 72 pts with CML-BP included in this analysis, 29 (40.3%) were female and 43 (59.7%) were males. At the time of diagnosis, 36 (50%) had CML-CP (chronic myeloid leukemia in chronic phase), 11 (15.3%) pts were diagnosed with CML-AP (accelerated phase CML) and 25 (34.7%) pts with de-novo CML-BP. For the 47 pts that were initially diagnosed with CML-CP or CML-AP, the median time to progression to CML-BP was 26 months (1-207 months). Median age at diagnosis of CML-BP was 54 years (range 21-86 years). EM disease was present in 17 (23.6%) pts (7 myeloid sarcoma, 5 central nervous system and 5 skin) with 3 pts presenting with EM disease only. Blasts were of myeloid phenotype in 41 (56.9%), lymphoid in 29 (40.3%) and mixed in 2 (2.8%) pts. The bcr-abl protein was p210 in 59 (81.9%), p190 in 5 (6.9%) and p230 in 1 (1.4%) pt. Four pts had prior allogeneic stem cell transplant (alloSCT) for CML-CP. Of the pts who had an ABL kinase domain mutation tested, 42.9% (18/42) pts were found to have an ABL kinase domain mutation. The most common TKD mutation was T315I (n=6). Next generation sequencing was performed in 32 pts. The most common mutations were ASXL1 (n=10), RUNX1 (n=4), BCOR (n=2), BCORL1 (n=2), DNMT3A (n=2) and 1 each in CSF3R, KMT2A, EZH2, NF1 and TP53. Of the 72 pts, 68 underwent CML-BP directed therapy. While only 9 (13.2%) pts were treated with TKI alone (4 dasatinib, 2 imatinib, 2 ponatinib and 1 asciminib), 54 (79.4%) pts were treated with a combination of chemotherapy with TKI (24 dasatinib, 13 ponatinib, 9 imatinib, 6 nilotinib, 1 bosutinib and 1 asciminib), and 5 (7.4%) pts were treated with chemotherapy alone. The response rate after CML-BP directed therapy was 72.1% (n=49). Best response was return to CML-CP in 17 pts, CHR with PCyR in 1 pt, CCyR in 11 pts, MMR in 5 pts, MR4 in 4 pts, MR4.5 in 5 pts and MR5 in 6 pts. Rate of response with third and second generation TKI was significantly better than first generation TKI (80% and 77.1% vs 36.3%; p=0.02287). 30 (44.1%) pts proceeded to alloSCT. Of the 49 pts that responded, 16 (32.6%) relapsed. The median time to relapse was 5 months (range 1-46 months). At a median follow up of 57 months, the median OS for all pts was 18 months. There was no difference in OS based on blast phenotype (14 months for myeloid, 21 for lymphoid and 7.5 for mixed; p=0.1185), initial diagnosis (38 months CP-CML, 13 for CML-AP and 16 for de-novo BP pts; p=0.6380), or TKI (13 months for first generation, 20 for second and not reached for third generation TKI; p=0.1729). The OS of pts that were treated with a combination of TKI and chemotherapy was significantly longer compared to pts treated with TKI or chemo alone (28 vs 13 or 4.5 months; p=0.0003). Of the 32 pts that responded, OS for pts that underwent alloSCT was significantly longer compared to pts that did not (not reached vs 14 months; p= 0.0022). Conclusion: CML-BP continues to have a dismal prognosis. Treatment is heterogenous involving various combinations of TKI and chemotherapy. In our analysis, treatment with second or third generation TKI led to better responses. In addition, treatment with chemotherapy in combination with TKI and alloSCT led to improved OS. Updated results from further multi-institution analysis will be presented at the meeting.

Background. The inclusion of gene mutations and chromosomal abnormalities in the 2022 WHO and ICC Classifications of MDS has enhanced diagnostic precision and is expected to improve clinical decision-making process. Although these two systems share similarities, clinically relevant discrepancies still exist and potentially cause inconsistency in their adoption in a clinical setting. In this study on behalf of the International Consortium for MDS (icMDS), we adopted a data-driven approach to provide a harmonization roadmap between the 2022 WHO and ICC classification for MDS. A modified Delphi Process consensus approach is currently ongoing among icMDS experts to finalize a harmonized MDS classification scheme. Methods. We analyzed retrospective international cohorts of patients with a diagnosis of MDS (n=7017) and AML (n=1002) according to WHO 2016 criteria. Hierarchical Dirichlet Processes were applied to define clusters capturing broad dependencies among all gene mutations and cytogenetic abnormalities. To investigate the features of importance and their impact on the clustering process, we employed the SHapley Additive exPlanations approach (SHAP). This allowed to define harmonized labels for each clinical entity. The clinical relevance of the unsupervised clustering was assessed through the analysis of phenotypic features and outcomes among each group. ( Blood 2022;140: 9828-9830) Results. Patients' characteristics are summarized in Table 1. We identified 9 clusters, defined by specific genomic features. The cluster of highest hierarchical importance was characterized by biallelic inactivation of TP53 (biTP53). According to SHAP analysis, bi TP53 was defined as 2 or more TP53 mutations, or 1 mutation with copy number loss or cnLOH. Most patients assigned to bi TP53 cluster had TP53 VAF&gt;10% (77.9%) and complex karyotype (70.1%). Assignment to bi TP53 cluster was irrespective of blast count. Patients with monoallelic TP53 mutation segregated into other clusters. Hierarchically, the second cluster included patients with del(5q). SHAP analysis highlighted 5q deletion alone, or with one other chromosomal abnormality other than -7, and absence of bi TP53, as the most relevant features. Most of these patients had blast counts &lt;5% (88.1%). The third distinct cluster included patients with SF3B1 mutations (in the absence of concurrent del(7q), abn3q26.2, complex karyotype or RUNX1 mutation). Most patients with MDS and SF3B1 mutation had &lt;5% blasts (94.2%). Common co-mutated variants in the SF3B1 cluster included mutant DNMT3A (25.2%) and TET2 (38.3%). Morphologically defined MDS cases (i.e., not meeting criteria for bi TP53, del(5q) or SF3B1) were preferentially assigned to the following additional clusters: SF3B1 and concurrent higher-risk mutations (e.g., RUNX1 and ASXL1); SRSF2 and concomitant TET2 mutations; U2AF1 mutations with del(20q), del(7q) or -7; SRSF2 with TET2 mutations and co-mutational patterns including RUNX1 and ASXL1; and AML-like genomic signatures. Our analyses suggest that morphologically defined MDS is characterized by a large heterogeneity in terms of mutation profiles, not entirely captured by the presence of unilineage versus multilineage dysplasia, percentage of bone marrow blasts, and presence of hypocellularity and fibrosis. To better investigate the continuum between high risk MDS (i.e., patients with ≥10% blasts) and AML, an exploratory comparison was made using a cohort of AML (defined according to WHO 2016) patients analyzed using the same statistical methods. Only a partial overlap in genetic signatures was observed between MDS with ≥10% blasts and AML. However, similarities were observed between the AML-like MDS clusters (characterized by mutant NPM1, bZIP CEBPA, and Core Binding Factor abnormalities) and AML clusters defined by the same genetic signature, thus supporting the classification of these entities as AML, irrespective of blast count. Conclusion. Our study demonstrated the utility of a data-driven approach based on advanced statistical methods to generate a harmonized classification for MDS. Table 2 shows a provisional, hierarchical classification algorithm. Further refinement of entity labels and classification criteria is the subject of the ongoing modified Delphi Process consensus approach among icMDS experts.