Citations

Abstract Excessively restrictive inclusion and exclusion criteria in clinical trials are one of many barriers to clinical trial enrollment for patients with myelodysplastic syndromes/neoplasms (MDSs). Many organizations are developing efforts to increase clinical trial eligibility; yet, several recent publications focused on patients with MDS suggest that many patients with this disease may be excluded from clinical trials unnecessarily. Clinical trial eligibility should reflect the phase of the study and risks of the agent being studied. Phase 3 trials should be less restrictive than early-phase trials to represent the real-world population as closely as possible. We hypothesize that many clinical trials, particularly phase 3 trials, have unnecessarily restrictive eligibility criteria. This study aims to evaluate the most common eligibility criteria according to phase of trial and to determine whether criteria correspond with drug safety signals. We identified MDS clinical trials registered on ClinicalTrials.gov from 1 January 2000 to 1 September 2023 and analyzed the eligibility criteria of 191 therapeutic MDS trials. We found that categorical inclusion and exclusion criteria are remarkably similar in representation across trial phases. Additionally, only 13% of trials are concordant with drug safety signals, suggesting that the eligibility criteria are often arbitrary. On behalf of the icMDS (International Consortium for Myelodysplastic Syndromes), an association of international MDS experts, we provide a position statement on restrictive eligibility criteria for MDS clinical trials that should be avoided with the aim of removing barriers to clinical trial enrollment.

Abstract Myelodysplastic syndromes/neoplasms (MDSs) are heterogeneous stem cell malignancies characterized by poor prognosis and no curative therapies outside of allogeneic hematopoietic stem cell transplantation. Despite some recent approvals by the US Food and Drug Administration, (eg, luspatercept, ivosidenib, decitabine/cedazuridine, and imetelstat), there has been little progress in the development of truly transformative therapies for the treatment of patients with MDS. Challenges to advancing drug development in MDS are multifold but may be grouped into specific categories, including criteria for risk stratification and eligibility, response definitions, time-to-event end points, transfusion end points, functional assessments, and biomarker development. Strategies to address these challenges and optimize future clinical trial design for patients with MDS are presented here.

Introduction The International Prognostic Scoring System (IPSS) and its revised version (IPSS-R) were developed using datasets from patients (pts) with untreated MDS. The cytogenetic risk groups used in IPSS-R were expanded to 5 groups from the 3 used in IPSS. Among the new groups introduced in IPSS-R is the “very complex” karyotype (VCK) where ≥4 karyotypic abnormalities are seen compared to “complex karyotype (CK)” with only 3 abnormalities. However, the prognostic impact of VCK vs CK, as well as individual cytogenetic abnormalities and TP53 mutation (TP53mut) status within the CK and VCK subgroups in pts with MDS treated with hypomethylating agents (HMAs) has not been well studied. Hence, we analyzed these questions in a large cohort of HMA-treated MDS pts in the icMDS VALIDATE database. Methods The icMDS VALIDATE database contains MDS pts treated with frontline HMA monotherapy or HMA-based combinations from 22 specialized international centers. Pts with unknown HMA type, date of HMA initiation, missing age, missing sex, missing cytogenetic data, bone marrow or peripheral blood blasts ≥20%, or missing survival status were excluded. We performed time-to-event analyses using the Kaplan-Meier method and compared groups by the log-rank test. We measured overall survival (OS) from time of HMA initiation. VALIDATE protocol was exempted, acknowledged, or approved by all centers' IRBs. VALIDATE is supported by an independent grant from Abbvie. Results A total of 1940 pts were included. Median age was 70 (range: 18-99) years and 66% were men. According to IPSS-R cytogenetic risk groups, 1.9% (37) were very good, 43.4% (841) good, 15.5% (301) intermediate, 11.6% (226) poor, and 27.6% (535) very poor. By HMA type, 71% of pts were treated with azacitidine, 14% with decitabine, 6% with HMA and venetoclax, and 9% with another HMA combination. Approximately 28% of pts underwent allogeneic hematopoietic stem cell transplant (alloHSCT). The IPSS-R cytogenetic risk groups had different OS (p<0.001)with median OS (mOS; months [mo], 95% CI) as follows: very good (43.4, 34.7-72.5), good (41.7, 36.4-46.8), intermediate (24.5, 23.1-32.4), poor (24.2, 19.2-28.5), and very poor (13.6, 12.5-14.5). The difference remained significant (p=0.001) after censoring at time of alloHSCT with mOS as follows: very good (40.4, 23.2-64.3), good (24.3, 22.2-26.3), intermediate (19.3, 15.1-22.2), poor (12.2, 10.3-14.8), and very poor (9.1, 8.2-19.7). Interestingly, the difference between mOS for intermediate and poor risk cytogenetics became more apparent after censoring at time of alloHSCT. Next, we compared the outcomes of pts with CK and VCK to non-complex karyotype (NCK) MDS. The mOS for pts with VCK MDS (13.6 mo, 12.5-14.5) was worse compared to pts with CK (19.9 mo, 19.9-16.8; p=0.004) and NCK (36.1 mo, 32.7-37.8; p<0.001) MDS. Among the 1458 pts with available molecular data, more pts with VCK (75%, 330) MDS had a concurrent TP53mut compared to pts with CK (40%, 21; p<0.001) and NCK (7%, 64 p<0.001) MDS. Considering only patients with TP53mut MDS, mOS for pts with VCK (12.2, 11.1-13.9) was worse compared with NCK (24.2, 16-1-53.1; p<0.001) but not with CK (19.9, 9.5-NR; p=0.9); the difference between CK and NCK was not significant (p=0.16). Among pts with TP53 wildtype (TP53WT) MDS, VCK (15.0, 13.2-18.2) had a similar mOS to TP53WT CK (18.8, 12.6-41.1; p=0.19) while both groups had worse mOS compared to TP53WT NCK (33.4, 30.1-37.0) (p=0.01). In a survival analysis of 616 pts with MDS with ≥3 cytogenetic abnormalities, -17/del(17p) (12.8 mo; p=0.009), inv(3)/t(3q)/del(3q) (10.9 mo; p=0.006), and del(7q) (12.1mo; p=0.024), though not -7 (12.6mo; p=0.23), were associated with lower mOS compared to the pts with MDS without these abnormalities (14.3mo). Conclusions Using one of the largest databases of HMA-treated MDS pts, we show that IPSS-R cytogenetic risk categories associate with significantly different survivals among HMA-treated MDS pts, even after censoring at time of alloHSCT. Importantly, VCK MDS was associated with worse mOS compared to CK and NCK MDS but not in pts with a concurrent TP53mut. Among pts with MDS with ≥3 chromosomal abnormalities, having -17/del(17p), inv(3)/t(3q)/del(3q), or del(7q) was associated with worse survival.

Background. Study of myeloid neoplasms (MN) has been rapidly transformed by genome characterization. Gene mutations have been reported to influence disease phenotype and risk of disease progression, and the evaluation of mutational status can provide valuable insights to improve the decision-making process. In this international study, we evaluated the clinical impact of mutational screening in patients with CMML, which is characterized by significant clinical and genomic heterogeneity, including a substantial proportion of patients with poor outcomes and unmet needs. Currently, only limited molecular information has been incorporated into CMML classification and prognostication. Methods. We retrospectively collected 3182 CMML patients. We used MOSAIC, an AI-based framework for multimodal analysis in rare cancers (PMID: 38875514) to develop innovative, molecular-based tools for classification and prognostication. An independent validation of the results on 516 prospectively collected patients was performed. Results. Bayesian Networs and Hierarchical Dirichlet Processes were used to identify genomic associations and define a CMML molecular classification. We identified 11 clusters with distinct clinical and genomic features, including splicing machinery, signal transduction and tyrosine kinase pathways aberrations, and high-risk molecular signatures (i.e., AML-like and TP53 mutations). Up to 15% of patients showed clear genomic overlap with other MN. Importantly, extensive multi-color flow cytometry on T lymphocytes, NK and myeloid cells (BD FACSymphony A5 Cell Analyzer) revealed specific immunologic and inflammation signatures associated with molecular subgroups. We then integrated molecular and clinical information to build an international CMML Prognostic Scoring System (iCPSS). We assessed the performance of different statistical and AI-based methods; L1-penalized Cox Model resulted as the best performing method. Selected features included hematological parameters (WBC, Hb, PLT and marrow blasts), cytogenetic abnormalities, and mutations in 10 genes (ASXL1, DNMT3A, EZH2, NRAS, RUNX1, SETBP1, STAG2, TET2, TP53, U2AF1). By Bayesian thresholds optimization, we identified 5 risk classes (i.e., very-low, low, intermediate, high and very-high) with median overall survival (OS) ranging from 99 to 9 months (P<.001), and median leukemia-free survival (LFS) ranging from 210 to 18 months (P<.001). iCPSS provided better patient discrimination across all clinical endpoints compared to currently available prognostic tools, with a Concordance Index (CI) for OS of 0.75 versus 0.62-0.64, respectively. Notably, compared to available scoring systems, up to 40% of patients were reassigned to higher or lower risk classes by the iCPSS. External validation on an independent prospective cohort confirmed that the performance of iCPSS is superior to existing prognostic models (CI for OS 0.71 vs 0.54-0.61, respectively) To demonstrate the clinical utility of iCPSS, we focused on 753 patients who underwent allogeneic stem cell transplantation (HSCT). In multivariable analysis, iCPSS stratified the probability of OS post-HSCT (P<.001) and identified groups of patients with different probabilities of disease relapse, ranging from 9% to 62% (P<.001). We employed a clinical- and genomic-based decision support system (PMID: 38723212) to determine the optimal timing of HSCT. Preliminary analyses indicated that higher-risk patients according to iCPSS (i.e., intermediate, high, and very high) benefit from an immediate HSCT strategy. Lower-risk patients (i.e., very low and low) had life expectancy maximized with delayed HSCT. Modeling decision analysis using iCPSS compared to conventional scores, resulted in a change in transplantation policy for a significant proportion of patients (up to 25%). Conclusion. Molecular information significantly improves the classification of CMML patients, providing a basis for refining diagnostic boundaries with other MN and for a more rational inclusion of patients in clinical trials. The iCPSS demonstrated superior performance over currently available scoring systems and improved HSCT decision making process.

Introduction: Response assessment is used to evaluate effectiveness of hypomethylating agent (HMA)-based therapy in myelodysplastic syndromes/neoplasms (MDS) in clinical practice and trials. Recently, the clinical response criteria (IWG 2023) and the clinical-molecular risk assessment (IPSS-M) in MDS were published. However, it remains unknown whether the IWG 2023 response criteria have an independent impact on overall survival (OS) when adjusting for baseline IPSS-M risk. Additionally, the association of IWG 2023 response criteria with OS in patients (pts) undergoing allogeneic stem cell transplantation (allo-HCT) versus those who do not is unknown. Lastly, it is unclear if optimal time-to-transplant impacts OS in MDS pts who achieve a IWG 2023 defined remission with HMA therapy. We examined these questions in a large population of MDS pts within the multi-institutional international consortium for MDS (icMDS) VALIDATE database. Methods: High-risk MDS pts with IPSS ≥1.5 or IPSS-R >3.5 treated with HMA-based therapy were included in the study. Baseline IPSS-M risk was stratified into very low/low/moderate low (VL+L+ML), moderate-high/high (MH+H), and very high (VH). Response to treatment was evaluated using the IWG 2023 criteria and grouped into: composite complete response (cCR) and no cCR. cCR included: complete response (CR), CR equivalent (CRequ), CR bilineage (CRbi), CR unilineage (CRuni), and CR with partial hematologic recovery (CRh). We analyzed the differential impact of achieving cCR on HMA-based therapy prior to transplant on OS in pts who either did or did not undergo allo-HCT, while adjusting for their baseline IPSS-M risk in Cox regression analysis. Optimal time-to-transplant was determined based on age and IPSS-M category as described by Tentori et al. (JCO 2024). Pts proceeding to allo-HCT were stratified on whether they underwent transplant within the optimal time-to-transplant timeframe after MDS diagnosis or not. VALIDATE protocol was exempted, acknowledged, or approved by all centers' IRBs. VALIDATE is supported by an independent grant from AbbVie. Results: We analyzed 762 MDS pts. The median age at MDS diagnosis was 68 years (range: 19-95 years) and 64% of pts were male. Distribution of the baseline IPSS-M risk was: VL+L+ML in 110 (14%) pts, MH+H in 347 (46%) pts, VH in 305 (40%) pts. Pts received azacitidine monotherapy, decitabine monotherapy, and HMA-based combination therapy (e.g. HMA + venetoclax) in 52%, 21%, and 28% of cases, respectively. A total of 282/762 (37%) pts achieved cCR as their best response to HMA-based therapy. The median OS for all pts was 18.7 months (95% CI: 16.6-21.3 months). Pts who did not undergo allo-HCT (n=436, 57%) had an improved median OS if they achieved cCR when compared to pts in the no cCR group in all IPSS-M risk groups (VL+L+ML: 23.5 months vs. 10.7 months; MH+H: 20.8 months vs. 11.6 months; VH: 13.8 months vs. 7.5 months). Cox analysis demonstrated an independent effect of cCR on OS in these pts when adjusted for baseline IPSS-M risk (cCR vs. no cCR: HR 0.60, 95% CI 0.48-0.76, p<0.001). In contrast, in pts who subsequently received an allo-HCT (n=326, 43%), achieving cCR with HMA prior to allo-HCT did not provide an additional OS benefit, when adjusted for baseline IPSS-M risk (cCR vs. no cCR: HR 1.00, 95% CI 0.73-1.38, p=0.986). The median time-to-transplant was 6.7 months and 34% of transplanted MDS pts received their allo-HCT during the optimal timeframe. Among pts who achieved cCR with HMA-based therapy prior to allo-HCT, optimal time-to-transplant was associated with significantly improved OS (optimal vs. non-optimal timing: HR 0.55; p=0.033). Conclusion: Baseline IPSS-M risk and allo-HCT had an independent impact on OS, while IWG 2023 defined cCR had an independent effect on OS only in non-transplanted MDS pts. While this retrospective analysis occurred at large academic centers with specific expertise in the management of MDS, our data may have an impact on clinical practice: (i) failure to achieve cCR on HMA therapy prior to transplant should not impact the decision to proceed to allo-HCT, (ii) upfront allo-HCT without prior HMA therapy might also be an option if the clinical situation and availability of donors allow, and (iii) allo-HCT should ideally be performed during the timeframe that is considered optimal as defined by pt age and baseline IPSS-M.

Abstract Clonal cytopenia of undetermined significance (CCUS) represents a distinct disease entity characterized by myeloid-related somatic mutations with a variant allele fraction of ≥2% in individuals with unexplained cytopenia(s) but without a myeloid neoplasm (MN). Notably, CCUS carries a risk of progressing to MN, particularly in cases featuring high-risk mutations. Understanding CCUS requires dedicated studies to elucidate its risk factors and natural history. Our analysis of 357 patients with CCUS investigated the interplay between clonality, cytopenia, and prognosis. Multivariate analysis identified 3 key adverse prognostic factors: the presence of splicing mutation(s) (score = 2 points), platelet count of <100 × 109/L (score = 2.5), and ≥2 mutations (score = 3). Variable scores were based on the coefficients from the Cox proportional hazards model. This led to the development of the clonal cytopenia risk score (CCRS), which stratified patients into low- (score of <2.5 points), intermediate- (score of 2.5 to <5), and high-risk (score of ≥5) groups. The CCRS effectively predicted 2-year cumulative incidence of MN for low- (6.4%), intermediate- (14.1%), and high-risk (37.2%) groups, respectively, by the Gray test (P < .0001). We further validated the CCRS by applying it to an independent CCUS cohort of 104 patients, demonstrating a c-index of 0.64 (P = .005) in stratifying the cumulative incidence of MN. Our study underscores the importance of integrating clinical and molecular data to assess the risk of CCUS progression, making the CCRS a valuable tool that is practical and easily calculable. These findings are clinically relevant, shaping the management strategies for CCUS and informing future clinical trial designs.

Background and Significance: The rare nature and diagnostic difficulty associated with myelodysplastic/myeloproliferative neoplasms (MDS/MPN) has limited the inclusion of patients (pts) with these diseases, other than chronic myelomonocytic leukemia (CMML), in relevant clinical trials. This has resulted in the complete lack of approved agents for the treatment of pts with non-CMML MDS/MPN. Moreover, ASTX727 (oral decitabine-cedazuridine) was approved for CMML based on a study of only 16 pts with CMML. A larger study of CMML and non-CMML MDS/MPN is needed. MDS/MPNs appear to be BCL2-dependent (Bogenberger et al., 2015; Jilg et al., 2016), but data for venetoclax combination therapy are derived from retrospective analyses of relapsed/refractory disease (Cortes et al. 2019; Hammond et al., 2020) or single-arm studies not dedicated to MDS/MPN (NCT04655755, Venugopal et al. 2021). Venetoclax exposures are not affected by coadministration of ASTX727 and predict responses similar to historical data (Mannis et al. 2021). Based on these promising early data and the critical need, we proposed a randomized study of ASTX727 alone or in combination with venetoclax for pts with MDS/MPN and >5% blasts. Study Design and Methods: This is a National Cancer Institute-sponsored, multi-center, international, randomized phase 2 study (NCT05600894, NCI 10538) designed to evaluate the efficacy of ASTX727 alone or in combination with venetoclax in pts with newly diagnosed MDS/MPN with >5% marrow blasts. Key eligibility criteria include allowing hydroxyurea cytoreduction to a WBC <10,000/mm 3 for proliferative disease and excluding pts with need for emergent disease-directed therapy as well as prior therapy for myeloid malignancy other than hydroxyurea or thrombopoietic/erythropoietin stimulating agents. Pts will be randomized 1:1 to receive either ASTX727 alone (Days 1-5) or in combination with venetoclax, initially administered with a 3-day dosing ramp-up and given for 14 days per 28-day cycle (Figure 1). Pts are stratified by disease subtype, specifically CMML and non-CMML MDS/MPN. The primary endpoint is the rate of complete remission (adapted from the MDS/MPN-specific response criteria developed by Savona et al. 2015). Secondary endpoints include the overall response rate, overall survival, progression-free survival, allogeneic hematopoietic stem cell transplantation rate, clonality at time of hematologic remission, number of red cell and platelet transfusions required and toxicity of ASTX727 versus ASTX727 + venetoclax. Pts randomized to the ASTX727 monotherapy arm and without response after 4 cycles of therapy, or with progressive disease after an initial response, cross over to the ASTX727 + venetoclax arm. A futility analysis is planned for each cohort (first 18 and 13 pts in each treatment arm for the CMML and non-CMML cohorts, respectively). We assume that 85% and 90% of pts in the CMML and non-CMML MDS/MPN cohorts, respectively, randomized to the ASTX727 monotherapy arm will not achieve the primary endpoint and will cross over to the ASTX727 + venetoclax combination arm. The estimated accrual time of a total of 70 pts with CMML and 50 pts with non-CMML MDS/MPN will require about 2.9 and 2.1 years, respectively. Planned integrated studies include the use of an apoptosis multiplex immunoassay to evaluate how the addition of a BCL-2 inhibitor will influence the ability of ASTX727 to induce apoptosis and if particular thresholds of apoptosis are predictive of clinical response. Planned exploratory studies include the use of: 1) RNA-seq to evaluate if differences in baseline transcriptional programs appear to be predictive of response to ASTX727 and ASTX727+venetoclax, 2) next generation (Oncomine Comprehensive Assay (OCAv3)/whole exome sequencing to uncover biomarkers predictive of response/outcome, validate risk modeling to stratify these pts and detect/monitor measurable residual disease (MRD) on serial marrow samples that can predict depth of response with translation in improvement in event-based endpoints, and 3) circulating tumor DNA as a non-invasive means (i.e., without need for a marrow evaluation) of prospective MRD detection for prognostic use. This ongoing, randomized study may inform frontline therapeutic options, provide prospective data for venetoclax combination after HMA failure, and generate striking correlative science for pts with “high risk” MDS/MPN.

Introduction: MDS are primarily diseases of the elderly, with median age at diagnosis of 72 years. Focused data on clinical and molecular characteristics, as well as outcomes, in younger pts under age 60 years are limited. We conducted this multicenter retrospective study to describe pt characteristics, cytogenetic and molecular profiles, treatment modalities, and outcomes in younger pts with MDS, including adolescent and young adults (AYA). Methods: The study included pts diagnosed with MDS between 2002-2023 at 7 academic centers. We compared baseline characteristics, treatment, and outcomes of AYA (18-39 years) and younger adult (YA) pts (40-59 years). Responses were assessed using the modified IWG 2006 response criteria which included complete remission (CR), marrow CR (mCR), and hematologic improvement (HI). Wilcoxon rank sum test and Fisher's exact (and Chi-square) test were used to compare pt and disease characteristics for continuous and categorical variables, respectively. Overall survival (OS) was estimated using the Kaplan-Meier method and compared using log-rank test. Results: A total of 173 pts were identified (AYA, N=50; YA, N=123). Median age at diagnosis was 50 years (range, 18-59), with 51% females. Baseline characteristics are shown in Table 1. Therapy-related MDS (t-MDS) was observed in 40 pts (23%) with the remaining classified as de novo MDS (N=133; 77%). Inherited bone marrow failure syndromes were documented in 15 pts (9%), with Shwachman-Diamond Syndrome (N=4) and Fanconi anemia (N=4) being the most frequent. Using the WHO 2016 classification, the most common subtypes of MDS in both the AYA and YA cohorts were MDS with multilineage dysplasia (38% vs 19%), MDS with excess blasts-2 (MDS-EB) (24% vs 29%), and MDS-EB-1 (16% vs 20%). Similarly, using the WHO 2022 classification, the most common subtypes were MDS with low blasts (50% vs 21%), MDS with increased blasts-2 (MDS-IB-2) (24% vs 24%), and MDS-IB-1 (16% vs 18%). Stratifying pts using the IPSS-R cytogenetic (CG) risk group, the most common CG risk category identified was good CG (34% vs 52%, p=0.012). Poor/very poor CG were present in 34% and 37% of the AYA and YA groups, respectively. STAG2 and NRAS mutations occurred more frequently in the AYA pts compared to the YA group, while the YA cohort was enriched for the following mutations: TP53, DNMT3A, SF3B1, SRSF2, and TET2 (Figure 1). In the AYA population, 18 pts (38%) had very low/low, 10 (21%) intermediate, and 20 pts (42%) had high/very high risk MDS using IPSS-R. In the YA population, 32 pts (26%) had very low/low, 32 (26%) intermediate, and 57 pts (47%) had high/very high risk MDS using IPSS-R. Twenty-seven percent of pts with very low to intermediate-risk MDS using IPSS-R were upstaged to moderate high to very high-risk MDS using IPSS-M. The majority of pts (AYA: 75%; YA: 63%, p=0.5) were transfusion independent at baseline. Single-agent hypomethylating agents were the most frequently used frontline therapy, used in 17 AYA pts (34%) and 58 YA pts (48%), with more AYA pts receiving upfront allogeneic stem cell transplantation (Allo-SCT) compared to YA pts (28% vs 5%). After frontline therapy (Table 1), an additional 14 AYA (28%) and 52 YA (42%) pts received Allo-SCT. In pts with IPSS-R intermediate to very high-risk, response rates (HI+CR+mCR+mCR with HI) to frontline therapies were 46% and 54.5% in the AYA and YA groups, respectively. After median follow-up of 24 months [range, 1-218], median OS was numerically longer in the AYA versus YA group (155 vs 53 months, p=0.2). AML transformation occurred in 24% and 20% of the pts in the AYA and YA groups, respectively. Median OS was significantly longer in de novo compared to t-MDS pts (56 vs 27 months, p=0.03). Median OS among Allo-SCT compared to non-SCT pts were not reached vs 155 months ( p=0.9) in the AYA group and 53 vs 54 months ( p=0.4) in the YA group. IPSS-R (very low+low vs intermediate vs high+very high) was able to distinguish differences in OS in the YA (133 vs 53 vs 17 months, p=0.0002) and AYA (155 vs not reached vs 18 months, p=0.006) pts, with the exception of the intermediate-risk group in the AYA cohort. Conclusions: In this study, d e novoMDS was seen in the majority of younger pts with MDS (both AYA and YA). Mutations in STAG2 and NRAS were more common in the AYA pts, while YA pts had MDS more enriched for TP53, DNMT3A, TET2 and splicing mutations. This study is still ongoing with a plan to compare the abovementioned groups to older pts (≥60 years) with MDS.

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>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 <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 <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.