MDS Foundation EHA Summaries 2026

Research plays a big role in helping people with myelodysplastic syndrome (MDS) live better lives. At the 2026 European Hematology Association (EHA) Congress in Stockholm, Sweden, scientists and doctors from around the world met from June 11–14 to share what they’ve been learning. Almost 150 MDS-focused research studies were presented.

The MDS Foundation is grateful for the patients, caregivers, and researchers who make this work possible. You can help move science forward by taking part in clinical research. Some studies — including online surveys and registries you can do from home — are listed on the Foundation’s website.

Below is a summary of some of the MDS‑related research shared at EHA 2026.

Low Risk MDS

Abstract: EHA-1789 Bafna V, et al: OFIRNOFLAST, A FIRST-IN-CLASS NEK7 INHIBITOR, INDUCES ROBUST TRANSFUSION INDEPENDENCE AND MULTILINEAGE HEMATOLOGICAL IMPROVEMENT IN LOWER-RISK MYELODYSPLASTIC SYNDROME(LR-MDS): PHASE 2 RESULTS

What the Study Was About: Scientists tested ofirnoflast, a new oral drug that blocks NEK7, a key protein that helps activate the NLRP3 inflammasome, which is overactive in MDS and harms blood‑forming cells. The goal was to see if shutting this pathway down could restore healthy blood production in people with lower‑risk, ESA‑refractory MDS.
Who Was in the Study: 37 patients were treated for up to 32 weeks. About half had a high transfusion burden, and others had low or no transfusion needs. Patients represented a mix of MDS subtypes, including del(5q), SF3B1, and MDS/MPN, with a median IPSS‑R score of 3.5.
Safety Findings: Ofirnoflast had an excellent safety profile. Only 27% had treatment‑related side effects, and just two were grade 3 or higher. No treatment‑related serious events occurred, and only one patient stopped treatment because of a side effect.
Main Efficacy Results: Among 30 evaluable patients, 67% had overall hematologic improvement. This included 63% improving red cells, 60% improving platelets, and 50% improving neutrophils. In transfusion‑dependent patients, 55% became transfusion‑independent for at least 8 weeks, and 39% for at least 16 weeks, with a median TI duration of 28.5 weeks.
Blood Count and Remission Responses: Patients who responded had a median hemoglobin rise of 4.2 g/dL, starting from a low baseline of 7.5 g/dL. Remarkably, 30% achieved a complete remission, and another 3% had a marrow CR. Responses occurred across all MDS subtypes and mutation backgrounds.
Bottom Line: Ofirnoflast produced strong, multi‑lineage blood improvements and high transfusion‑independence rates with minimal toxicity. These results provide the first clinical proof that blocking NEK7 can calm the NLRP3 inflammasome and restore healthy blood production, making ofirnoflast a promising new therapy for lower‑risk MDS patients who have few options.

Abstract EHA-3474 Della Porta MG, et al: LUSPATERCEPT (LUSPA) INITIATED AT THE MAXIMUM-APPROVED DOSE IN PATIENTS (PTS) WITH LOWER-RISK MYELODYSPLASTIC SYNDROMES (LR-MDS) WHO REQUIRE TRANSFUSIONS: PRIMARY ANALYSIS FROM MAXILUS

What the Study Was About: Researchers tested luspatercept (LUSPA) at the highest approved dose (1.75 mg/kg) in people with lower‑risk MDS who needed blood transfusions and had either never used ESAs or had stopped responding to them.
Who Was in the Study: 106 patients took part (53 ESA‑naive and 53 ESA‑relapsed/refractory/intolerant). The median age was 76, and most had needed transfusions recently. About half had SF3B1 mutations, and many had low erythropoietin levels, which often predicts better response.
Main Findings in ESA‑Naive Patients: About 81% reached the main goal—8+ weeks without transfusions plus a hemoglobin rise—often within 1 day. 83% stayed transfusion‑free for at least 8 weeks, and 76% for at least 12 weeks. Even patients without ring sideroblasts or with low EPO levels showed strong responses.
Main Findings in ESA‑Relapsed/Refractory Patients: About 62% reached the main goal, usually within 2 days. 68% stayed transfusion‑free for at least 8 weeks, and 51% for at least 12 weeks. Responses were still meaningful even in tougher‑to‑treat groups.
Side Effects and Safety: Most patients had side effects, but they were similar to what is already known for LUSPA. Serious side effects happened in 54–66%, but only 7–11% were treatment‑related. A small number (3.8%) progressed to higher‑risk MDS, and none developed AML.
Bottom Line: Starting LUSPA at the maximum dose led to high and lasting transfusion independence in both ESA‑naive and ESA‑experienced patients. The treatment was well tolerated, and the results support its strong safety and effectiveness in lower‑risk MDS.

Abstract: EHA-4111 Park S, at al: EARLY VERSUS DELAYED ESA IN LOWER-RISK NON -TRANSFUSION-DEPENDENT MDS: FINAL PHASE III EPO-PRETAR RESULTS ON TRANSFUSION DEPENDENCE AND ERYTHROID RESPONSE

What the Study Was About: Doctors studied whether starting EPO alfa early (when hemoglobin was 9–10.5 g/dL) helps people with lower‑risk MDS avoid needing blood transfusions longer, compared with waiting to start EPO until hemoglobin dropped below 9 g/dL.
Who Was in the Study: 84 patients took part (43 early‑start, 41 delayed‑start). In the early group, almost everyone (95%) began EPO right away; in the delayed group, only 80% eventually needed EPO, starting much later (about 16 months after joining the study).
Main Findings on Transfusions: Starting EPO early did NOT delay the time until patients became transfusion‑dependent. About 49% of early‑start patients and 46% of delayed‑start patients eventually needed regular transfusions, with no meaningful difference between groups.
Erythroid Response (Hemoglobin Improvement): Early EPO clearly helped patients’ blood counts rise faster. At 12 weeks, 85% of early‑start patients had a strong hemoglobin response, compared with 55% in the delayed group. Responses also lasted longer with early EPO (about 33 vs 25 months, though not statistically significant).
Survival and Disease Progression: Overall survival and progression‑free survival were similar in both groups. About 4 patients in each arm developed AML, showing no difference in disease worsening.
Safety and Quality of Life: Treatment was well tolerated in both groups, with no new safety concerns. Quality‑of‑life scores stayed similar, whether EPO was started early or delayed.

Abstract: EHA-6964  Komrokji RS, et al: REAL-WORLD OUTCOMES OF IMETELSTAT: INTERROGATING SAFETY, EFFICACY AND PREDICTORS OF RESPONSE IN HEAVILY PRETREATED LOWER-RISK MDS PATIENTS

What the Study Was About: Doctors looked at how well imetelstat, a telomerase‑blocking medicine recently approved for lower‑risk MDS, works in real‑world patients. In the IMerge trial, about 40% of patients became transfusion‑independent, so this study checked if similar results happen outside clinical trials.
Who Was in the Study: 40 lower‑risk MDS patients were treated at one cancer center. Most (36/40) needed transfusions at the start. Patients had many prior treatments—88% had luspatercept, 70% had ESAs, 38% had HMAs, and 43% had lenalidomide. The median hemoglobin before treatment was 7.4 g/dL, and most had SF3B1 mutations (80%).
Main Response Results: The rate of ≥8‑week transfusion independence was 37.5% (15/40), similar to clinical trial results. Responses lasted a median of 185 days (~6 months), and nearly halfof responders were still benefiting at the time of analysis. Responders received more cycles (10 vs. 5).
Predictors of Response: Patients with an ASXL1 mutation responded more often (67% vs. 29%). A higher baseline platelet count (≥192) was the strongest predictor of benefit. Prior HMA or lenalidomide use seemed to reduce response rates, but not significantly.
Safety Findings: Side effects were mostly low blood counts, especially after the first two cycles—35% had grade 3/4 neutropenia and 28% had grade 3/4 thrombocytopenia after cycle 1. Importantly, no cases of febrile neutropenia occurred. Some patients needed dose reductions or supportive care (platelet transfusions or G‑CSF).
Bottom Line: In real‑world practice, imetelstat showed meaningful transfusion independence, even in heavily pretreated patients, and had a manageable safety profile. New potential predictors of response—platelet count and ASXL1 mutation—may help identify patients most likely to benefit.

Abstract: EHA-2044 Campelo MD, et al: Title: UPDATED PHASE 2 RESULTS OF ELRITERCEPT IN PATIENTS (PTS) WITH TRANSFUSION-DEPENDENT (TD) LOWER-RISK MYELODYSPLASTIC NEOPLASMS (LR-MDS)

What the Study Was About: Doctors tested elritercept, a new treatment that targets TGF‑β signaling, to help people with lower‑risk MDS who depend on regular blood transfusions. The goal was to see if it could help patients become transfusion‑independent and improve their blood counts.
Who Was in the Study: 96 patients received the recommended dose (3.75 mg/kg every 4 weeks). Many had been on treatment for over a year (median 13.8 months). Patients included both ring sideroblast–positive and non‑RS groups, and both low and high transfusion burden groups.
Safety Findings: Elritercept was well tolerated. About 46% had treatment‑related side effects, but only 5% were grade 3 or higher. The most common issues were nausea (7%) and diarrhea (6%), with no major safety concerns.
Transfusion Independence Results: By week 24, 38.5% of patients became transfusion‑independent for at least 8 weeks, including 29% of high‑burden patients and 44% of RS‑positive patients. By week 48, 27% achieved 24‑week TI, and responses happened very quickly (median 0.4 weeks). Patients who achieved TI often stayed independent for a long time—median 111 weeks.
Hemoglobin and Blood Count Improvements: HI‑E responses occurred in 49% of patients by week 24, with a median time to response of 3.1 weeks. Responses were seen across subgroups, including non‑RS patients and those with EPO <500 U/L. About 32% had sustained neutrophil increases and 33% had platelet improvements.
Bottom Line: Elritercept produced rapid, durable transfusion independence and strong erythroid responses in lower‑risk MDS, even in harder‑to‑treat groups like high‑burden or non‑RSpatients. The drug was well tolerated, supporting its potential as a meaningful new therapy for transfusion‑dependent LR‑MDS.

Abstract: EHA-3488 Santini V, et al: UPDATED OVERALL SURVIVAL AND LONG-TERM TRANSFUSION INDEPENDENCE IN ESA-NAIVE PATIENTS WITH LOWER-RISK MDS: RESULTS FROM THE PHASE 3 COMMANDS TRIAL

What the Study Was About: This study looked at long‑term results from the phase 3 COMMANDS trial, comparing luspatercept (LUSPA) with epoetin alfa (EA) in ESA‑naive, transfusion‑dependent lower‑risk MDS. The goal was to see whether LUSPA improves overall survival, transfusion independence, and hemoglobin levels over more than 3 years of follow‑up.
Who Was in the Study: Adults with LR‑MDS who had never received an ESA and were transfusion‑dependent were randomized 1:1 to LUSPA or EA. Patients were stratified by transfusion burden, ring sideroblast status, and serum EPO level.
Overall Survival Results: After >3 years of follow‑up, median OS was not reached with LUSPA but was 46 months with EA (HR 0.78). This favorable trend for LUSPA was seen across subgroups, though not statistically significant.
Transfusion Independence & Hemoglobin Responses:
TI ≥12 weeks: 76% with LUSPA vs 56% with EA
TI lasting ≥2.5 years: 25% vs 11%
Median TI duration: 184 weeks vs 95 weeks
Hb rise ≥1.5 g/dL: 80% vs 59%
Duration of Hb response: 73 weeks vs 48 weeks LUSPA produced longer, deeper, and more durable erythroid responses.
Disease Progression & Safety: Fewer LUSPA patients progressed to higher‑risk MDS (3.3% vs 7.3%), and AML progression was similar (5% vs 6%). No new safety issues emerged. About 18% of LUSPA patients were still on treatment at cutoff.
Bottom Line: With more than 3 years of follow‑up, luspatercept continues to outperform epoetin alfa, offering longer survival trends, much higher and more durable transfusion independence, and stronger hemoglobin improvements, with no new safety concerns. These results reinforce LUSPA as the preferred first‑line therapy for ESA‑naive, transfusion‑dependentLR‑MDS.

Abstract: EHA-4146 Guo W, et al:  WHICH PATIENTS WITH MDS-H AND MDS-LB BENEFIT FROM ALLOGENEIC HEMATOPOIETIC STEM CELL TRANSPLANTATION? EVIDENCE FROM A LARGE SINGLE-CENTER RETROSPECTIVE COHORT OF 636 PATIENTS

What the Study Was About: Doctors studied which patients with hypoplastic MDS (MDS‑h) or MDS with low blasts (MDS‑LB) actually benefit from allogeneic stem cell transplant (allo‑HSCT). These subtypes usually have low blasts and no defining cytogenetic abnormalities, so it’s unclear who truly needs transplant.
Who Was in the Study: The team analyzed 636 patients (146 MDS‑h, 490 MDS‑LB) treated between 2016–2024. The median age was 52, and 42% were transfusion‑dependent (TD) at diagnosis. Over half (52%) were IPSS‑M higher‑risk. A total of 112 patients eventually underwent transplant.
Impact of Transfusion Dependence: TD was a major negative prognostic factor, regardless of IPSS‑M risk.
5‑year OS in TD: 39% (higher‑risk) and 49% (lower‑risk)
5‑year OS in TI: 65% (higher‑risk) and 81% (lower‑risk) This shows that transfusion status adds important risk information beyond IPSS‑M.
Who Benefited From Transplant:
In TD + IPSS‑M higher‑risk patients, transplant significantly improved survival (5‑year OS 86% with HSCT vs 47% without, P=0.004).
In TD + lower‑risk patients, no benefit was seen.
In TI patients, transplant helped only those with higher‑risk IPSS‑M, while lower‑risk TI patients had excess early mortality after HSCT.
Progression to Transfusion Dependence: Among patients who were TI at diagnosis, 46% eventually became TD. The strongest predictors of becoming TD were IPSS‑M higher‑risk (HR 2.41) and MDS‑LB subtype (HR 2.20).
Bottom Line: Transplant decisions in MDS‑h and MDS‑LB should consider both IPSS‑M molecular risk and transfusion status.
Clear benefit: TD + IPSS‑M higher‑risk
Possible benefit: TI + IPSS‑M higher‑risk
No benefit / potential harm: Lower‑risk patients, especially if TI This supports a risk‑adapted, personalized transplant strategy.

Abstract: EHA-2510 Zeiden AM, et al: ROXADUSTAT IMPROVES TRANSFUSION INDEPENDENCE IN LR-MDS PATIENTS WITH ANEMIA AND HIGH TRANSFUSION BURDEN AND IN RING SIDEROBLAST POSITIVE AND NEGATIVE DISEASE: POST-HOC ANALYSIS OF MATTERHORN STUDY

What the Study Was About: Researchers looked at which lower‑risk MDS patients with anemia benefit from roxadustat, an oral drug that boosts red blood cell production. This post‑hocanalysis focused on patients with high transfusion burden (HTB) and those with or without ring sideroblasts (RS).
Who Was in the Study: The analysis included 140 patients from the MATTERHORN Phase 3 trial. For this post‑hoc study, 37 patients met criteria for high transfusion burden (22 on roxadustat, 15 on placebo). Patients had very low, low, or intermediate‑risk MDS, with or without prior ESA exposure.
Main Transfusion Independence Results (HTB Group):
8‑week TI (first 28 weeks): 36% with roxadustat vs 7% with placebo (p=0.041)
8‑week TI (first 52 weeks): 46% vs 13% (p=0.043)
Longer TI durations also favored roxadustat. These results show meaningful and durable reductions in transfusion needs in a group that is typically very hard to treat.
Effect in Ring Sideroblast–Positive and –Negative Patients: In the full analysis set with RS data, both RS+ and RS– patients showed improvements in transfusion independence and hemoglobin levels, suggesting roxadustat works regardless of RS status.
Safety Findings: Side effects were mostly low‑grade and manageable. No new safety concerns were identified, and the drug was generally well tolerated.
Bottom Line: Roxadustat produced clinically meaningful and durable transfusion independence, especially in high‑burden LR‑MDS patients, with benefits seen in both RS+ and RS– disease. As an oral option, it addresses a major unmet need, and a new global Phase 3 trial is planned for 2026.

Abstract: EHA-2040 Liu Y, et al: HETROMBOPAG FOR THROMBOCYTOPENIA IN LOWER-RISK MYELODYSPLASTIC SYNDROMES: A PROSPECTIVE SINGLE-CENTER STUDY

What the Study Was About: Doctors tested hetrombopag, an oral thrombopoietin‑receptor agonist, to see whether it can safely raise platelet counts in lower‑risk MDS patients with thrombocytopenia, a group with few treatment options and high bleeding risk.
Who Was in the Study: 56 adults with LR‑MDS and low platelets were followed for a median of 61 weeks. All had <5% blasts, IPSS low or Int‑1 risk, and platelets either <30×10⁹/L or 30–100×10⁹/L with bleeding. Hetrombopag was started at 5 mg/day and increased up to 15 mg/day based on response.
Platelet Response Results:
Overall platelet response: 59% (33/56)
Responses within 24 weeks: 54%
Complete responses: 10 patients
Median time to response: 6 weeks
Duration of response: Not reached Patients with platelets <30×10⁹/L or IPSS‑M >0 responded less often.
Transfusion & Bleeding Improvements:
Among 21 platelet‑transfusion‑dependent patients, 57% became transfusion‑independent.
Among 30 RBC‑transfusion‑dependent patients, 30% became transfusion‑independent.
Bleeding events decreased during treatment.
Erythroid improvement occurred in 50% of anemic patients.
Safety Findings: Side effects occurred in 15 patients, mostly infections. No blood clots (arterial or venous) were seen. Serious grade 3–4 events (e.g., pulmonary infection, angina, heart failure, sepsis) were not considered drug‑related. Only 2 patients showed disease progression (one to IB2, one to AML).
Bottom Line: Hetrombopag produced meaningful platelet increases, reduced bleeding and transfusion needs, and showed no signal of increased leukemic progression. The drug appears safe and effective for thrombocytopenic LR‑MDS and deserves further study.

High Risk MDS

Abstract EHA-1801 Platzbecker U, et al: CPX-351 VERSUS STANDARD OF CARE AS BRIDGING TO ALLOHCT IN PATIENTS WITH HIGHER-RISK MYELODYSPLASTIC NEOPLASMS OR OLIGOBLASTIC ACUTE MYELOID LEUKEMIA: RESULTS OF THE RANDOMIZED PHASE 2 PALOMA STUDY

What the Study Was About: Doctors compared CPX‑351 with standard treatments (azacitidine or regular chemotherapy) to see which worked better for people with high‑risk MDS or early AML who were planning to get a stem cell transplant within 6 months.
Who Was in the Study: 150 patients joined the trial. They were randomly placed into either the CPX‑351 group (75 people) or the standard‑care group (75 people total: 52 got chemo, 23 got azacitidine).
Main Findings: Patients who received CPX‑351 stayed free of major problems longer. Their event‑free survival (EFS) was not reached, meaning many were still doing well at follow‑up. The standard‑care group had a median EFS of 11.1 months. Overall survival looked similar so far (CPX‑351 not reached vs. 35.4 months), but this part was not statistically significant yet.
Transplant Results: A very high number of patients in both groups were able to move forward with their stem cell transplant—57 people (85%) in the CPX‑351 group and 64 people (86%) in the standard‑care group.
Side Effects: The most common serious side effects were low platelets (65 CPX‑351 vs. 55 standard care), low white blood cells (55 vs. 46), and fever with infection (44 vs. 35). These side effects are expected with these types of treatments.
Bottom Line: CPX‑351 met the main goal of the study by giving patients longer time before their disease got worse compared with standard treatments, while still allowing most patients to reach transplant.

Abstract: EHA-4187 Streuer A, et al: FUNCTIONAL TP53 PHENOTYPE SCORE REFINES PROGNOSTIC STRATIFICATION IN MONOALLELIC TP53-MUTATED MYELODYSPLASTIC NEOPLASMS BEYOND ALLELIC STATE

What the Study Was About: Researchers looked at whether a TP53 functional phenotype score (PS)—a number showing how damaging a TP53 mutation is—can better predict outcomes in monoallelic TP53‑mutated MDS, a group known to have mixed and unpredictable survival.
Who Was in the Study: The team analyzed 4,505 patients with MDS from two large cohorts. Among them, 271 had monoallelic TP53 mutations and 499 had multi‑hit TP53, which is known to be very high‑risk. Every TP53 variant was scored using a lab‑based system that measures how much the mutation disrupts TP53 function.
Main Findings on Survival: Survival dropped stepwise across groups: TP53‑wild‑type (42 months) → TP53‑monoallelic (23 months) → TP53‑multi‑hit (9 months). Within the monoallelic group, patients with high PS (≥7) had much worse survival (14 months) than those with low PS (<7) (39 months).
AML Progression and Risk Features: High‑PS patients had a higher risk of transforming to AML and were more likely to have complex karyotypes, a known high‑risk feature. Importantly, PS identified high‑risk patients even inside “low‑risk” clinical categories.
Role of VAF (Mutation Burden): A higher VAF (≥22%) also predicted worse survival, but PS and VAF were not correlated, meaning they measure different biological risks. When combined, they sharply separated outcomes—from near wild‑type survival (≈50 months) to multi‑hit‑like survival (≈9 months).
Bottom Line: Monoallelic TP53‑mutated MDS is not one single disease. Using functional TP53 scores plus VAF gives a much clearer picture of risk than allelic state alone and may help identify patients who need earlier aggressive treatment or transplant evaluation.

Abstract: EHA-4114 Zhou X, et al: A NEW ORAL DOSAGE FORM OF DECITABINE/CEDAZURIDINE DEMONSTRATED CONSISTENT EFFICACY AND FAVORABLE SAFETY PROFILE IN MYELODYSPLASTIC SYNDROMES PATIENTS IN CHINA: FIRST RESULTS FROM A PIVOTAL TRIAL

What the Study Was About: Doctors tested a new oral version of decitabine/cedazuridine (DEC‑C) in people with higher‑risk MDS or CMML in China. This oral form is designed to give the same drug exposure as the IV version but in a more convenient pill.
Who Was in the Study: 73 patients were treated across 15 centers. Most (96%) had MDS, and the median age was 65 years. Part 1 of the study compared oral DEC‑C with IV decitabine in a crossover design to confirm equivalent drug levels.
Main Efficacy Results: After a median of 5 treatment cycles, the overall response rate was 54.8% using IWG 2006 criteria. This included 22% complete responses, plus additional marrow CRand hematologic improvement responses. Using newer IWG 2023 criteria, the ORR was 56%, again showing strong activity similar to the FDA‑approved ASTX727.
Survival Outcomes: After about 9 months of follow‑up, the median overall survival was 19.1 months, and the 12‑month survival rate was 76%. Leukemia‑free survival was also 19.1 months, with a 71% 12‑month LFS rate.
Safety Findings: The oral DEC‑C had a similar safety profile to IV decitabine. About 7% of patients had treatment‑emergent events that led to death, and 4% were considered treatment‑related (pneumonia, sepsis, septic shock). No unexpected toxicities were seen.
Bottom Line: The new oral DEC‑C showed strong, consistent efficacy and a favorable safety profile, supporting its use as the first oral hypomethylating agent option for MDS patients in China, comparable to the established IV formulation.

Abstract: EHA-2146 Nie ZY, et al. BORTEZOMIB COMBINED WITH AZACITIDINE AND VENETOCLAX IN HIGH-RISK MYELODYSPLASTIC NEOPLASMS: A SINGLE-CENTER RETROSPECTIVE STUDY

What the Study Was About: Doctors tested a three‑drug combination—bortezomib + azacitidine + venetoclax (called VAB)—to see whether it could improve outcomes for people with high‑risk MDS, a group that usually responds poorly to current treatments.
Who Was in the Study: 20 patients with newly diagnosed high‑risk MDS were treated. The median age was 51 which is younger than typical MDS populations. About 45% had EB1, 45% had EB2, and 25% had complex karyotypes. The median number of mutations was 3, and 40% later went to stem cell transplant.
Early Response Results: The overall composite remission rate (CRc) was 75%, meaning 3 out of 4 patients had a meaningful response. After just one cycle, 60% had already reached CRc. The high‑risk group responded better than the very‑high‑risk group (100% vs. 55%).
Survival Outcomes: After a median follow‑up of 17.4 months, only 4 patients had died and 2 progressed to AML. The median overall survival had not yet been reached, suggesting encouraging durability. The 1‑year and 2‑year survival rates were both 78.6%.
Safety Findings: The most common side effects were grade ≥3 low blood counts, which were expected and manageable. No unexpected toxicities were reported, and the regimen was considered tolerable.
Bottom Line: The VAB triplet (bortezomib + azacitidine + venetoclax) produced high remission rates, rapid responses, and promising survival in newly diagnosed high‑risk MDS. This early experience suggests the regimen may offer a new, more effective first‑line option for this difficult‑to‑treat group.

Abstract: EHA-2788 Bataller A, et al: HYPOMETHYLATING AGENTS WITH OR WITHOUT VENETOCLAX IN TREATMENT-NAÏVE MYELODYSPLASTIC SYNDROMES WITH EXCESS BLASTS: RESULTS FROM A LARGE RETROSPECTIVE COHORT

What the Study Was About: Doctors compared outcomes in treatment‑naïve MDS with excess blasts (MDS‑EB) treated with hypomethylating agents (HMAs) alone versus HMA + venetoclax(HMAVen). The goal was to identify which subgroups actually benefit from adding venetoclax.
Who Was in the Study: The study included 358 patients (245 HMA, 113 HMAVen) treated from 2010–2025. Median age was ~70. Blast levels were higher in the HMAVen group (12% vs 8%). High‑risk and very‑high‑risk IPSS‑R categories were similar between groups. Mutations were balanced, including TP53 (~33%), ASXL1 (23–32%), and DNMT3A (~15%).
Response Rates: Adding venetoclax significantly improved overall response rates:
IWG 2006 ORR: 62% (HMA) vs 89% (HMAVen)
IWG 2023 ORR: 51% vs 77% Complete remission rates were similar (~35%). More HMAVen patients proceeded to transplant (43% vs 26%).
Overall Survival: Despite higher response rates, median OS was similar:
19 months (HMA) vs 18 months (HMAVen) Two‑year OS was also comparable (41% vs 46%). No OS differences were seen across HMA types.
Subgroups That Benefited From Venetoclax: In unadjusted analyses, HMAVen improved OS in:
ASXL1‑mutated patients (HR 0.45)
IPSS‑R very high‑risk (HR 0.65)
IPSS‑M very high‑risk (HR 0.65)
BM blasts ≥10% (HR 0.69) After adjusting for transplant, these benefits weakened, suggesting venetoclax mainly helps by increasing the chance of reaching transplant. Markov modeling confirmed HMAVen doubled the likelihood of transitioning to HSCT in these subgroups.
Bottom Line: Venetoclax does not improve overall survival for all MDS‑EB patients, but specific high‑risk groups—ASXL1‑mutated, very‑high‑risk IPSS‑R/IPSS‑M, or blasts ≥10%—appear to benefit. The main advantage is higher response rates and greater access to transplant, which likely drives survival improvements.

Diagnostic and Prediction Studies

Abstract: EHA-2407 Della Porta MG, et al: CYTO-CMML: A MUTATION-INDEPENDENT CYTOGENETIC RISK CLASSIFIER FOR CHRONIC MYELOMONOCYTIC LEUKEMIA

What the Study Was About: Researchers wanted to improve how doctors predict outcomes in chronic myelomonocytic leukemia (CMML) by creating a new cytogenetic (chromosome-based) risk system that does not depend on gene mutations. This helps guide decisions about stem cell transplant, the only curative option.
Who Was in the Study: The team analyzed over 2,100 patients with CMML from two large groups (971 in the discovery set and 1,175 in the validation set). Patients were older (median age ~72–73), and follow‑up lasted about 4–5 years.
Main Findings on Survival: Some chromosome changes (like –Y, 5q–, 7q–) had survival similar to normal karyotype (~38 months). But others—such as complex karyotype, –7, +21, i(17q), inv(3), t(1;3)—had very poor survival (~10 months). All other abnormalities fell in between (~22 months).
Risk Groups and Blast Transformation: Only the high‑risk cytogenetic group had a clearly higher chance of turning into AML, with a 1‑year blast transformation rate of 33%, compared with 8–9% in the other groups. This pattern was consistent in both study cohorts.
Independent Predictors: Even after adjusting for other factors (mutations, blood counts, blast percentage), high‑risk cytogenetics remained a strong predictor of both shorter survival and higher AML risk. The intermediate group was less reliable, so researchers combined low + intermediate into a single “standard‑risk” category.
Bottom Line: The new Cyto‑CMML system is a simple, mutation‑independent way to classify CMML patients into standard‑risk or high‑risk groups. It improves prediction of survival and AML progression and may help doctors decide who needs earlier transplant or more aggressive treatment.

Abstract: EHA-1673 Palenzuela LG, et al: VALIDATION OF MOLECULAR PROGNOSTIC MODELS FOR CHRONIC MYELOMONOCYTIC LEUKEMIA IN A LARGE MULTICENTER COHORT

What the Study Was About: Doctors tested how well molecular risk models (which use gene mutations) predict survival and AML transformation in chronic myelomonocytic leukemia (CMML). These models include GFM, MMM, CPSSmol, and BLASTmol, but they haven’t been validated in large real‑world groups until now.
Who Was in the Study: The study included 977 CMML patients from multiple centers. The median age was 71, median overall survival was 29 months, and about 21% eventually developed AML. Most patients received supportive care, while those who had a stem cell transplant lived the longest (about 52 months).
Clinical and Genetic Risk Factors: Worse survival was linked to high white blood cells, high monocytes, >5% bone marrow blasts, and circulating immature cells. High‑risk cytogenetics shortened survival (15 vs. 39 months). Mutations in ASXL1, RUNX1, and TP53 were also associated with poorer outcomes.
How the Models Performed: Most molecular models separated patients into clear risk groups. MMM performed best, showing a strong stepwise drop in survival (from 106 months in low‑risk to 20 months in high‑risk). BLASTmol also clearly identified high‑risk patients. GFM worked but had overlap in the middle groups. CPSSmol performed the weakest and did not reach statistical significance.
AML Transformation Results: AML progression patterns matched the survival results. MMM again showed the clearest separation, with high‑risk patients having a 41% chance of AML at 3 years. BLASTmol also identified high‑risk transformation, while GFM and CPSSmol did not significantly separate groups.
Bottom Line: Molecular CMML‑specific models—especially MMM—are better than older clinical scoring systems at predicting survival and AML risk. This supports using molecular profiling to guide treatment decisions, including when to consider transplant.

Abstract: EHA-3713  Stompór P, et al: Title: OUTCOMES AND CHARACTERISTICS OF LOWER-RISK MDS PATIENTS WITH ELEVATED SERUM ERYTHROPOIETIN LEVELS (≥500 IU/L): A MULTICENTER STUDY BY THE POLISH ADULT LEUKEMIA GROUP (PALG)

What the Study Was About: Doctors studied people with lower‑risk MDS who had very high erythropoietin levels (≥500 IU/L) to understand whether this group has different symptoms, lab results, or survival compared with patients whose EPO levels are lower.
Who Was in the Study: The analysis included 435 lower‑risk MDS patients from 50 centers in Poland. Of these, 56 patients (13%) had sEPO ≥500 IU/L, while the rest had lower levels and served as the comparison group.
Clinical Features of High‑EPO Patients: Patients with sEPO ≥500 IU/L were younger (about 69 vs. 72 years) but had much more severe anemia (hemoglobin 7.65 vs. 9.1 g/dL). They also had higher ferritin levels, suggesting more iron overload.
Disease Burden and Risk Features: High‑EPO patients were far more likely to be transfusion‑dependent (77% vs. 44%) and had more bone marrow fibrosis (13% vs. 7%). They also had higher rates of tobacco exposure, which may worsen inflammation and anemia.
Survival Outcomes: Patients with sEPO ≥500 IU/L had significantly shorter survival. Their median overall survival was 33 months, compared with 48 months in the lower‑EPO group, showing that high EPO is a marker of more aggressive disease biology.
Bottom Line: An sEPO level ≥500 IU/L identifies a distinct high‑risk subgroup within lower‑risk MDS—characterized by severe anemia, high transfusion needs, fibrosis, and worse survival. These patients may need more personalized and proactive management.

Abstract: EHA-2079 Saini K, et al: HEAVY METAL ABNORMALITIES IN UNEXPLAINED CYTOPENIAS AND LOW‑BLAST MYELODYSPLASTIC NEOPLASMS: A PROSPECTIVE INTERVENTIONAL COHORT STUDY FROM NORTH INDIA

What the Study Was About: Doctors looked at whether heavy metal problems—like low copper or high levels of toxic metals—might explain unexplained low blood counts or low‑blast MDS, and whether fixing these problems could improve blood counts.
Who Was in the Study: 120 patients from North India were followed. All had unexplained cytopenias or low‑blast MDS, and common causes (like B12 deficiency or infections) were ruled out. The median age was 46, and about 54% had low‑blast MDS.
How Common Heavy Metal Issues Were: Heavy metal abnormalities were found in two‑thirds of patients (67%), and one‑third had more than one abnormality. Copper deficiency was the most common (~65%), while excess zinc, lead, and arsenic were the main toxic exposures.
Clinical Clues: Some patients had ring sideroblasts (10.6%), which were linked to high mercury and zinc levels. These findings often mimicked low‑risk MDS, making diagnosis tricky without metal testing.
Treatment Responses: Among 50 patients with follow‑up, 70% improved after targeted treatment (copper replacement or chelation). Responders had an average hemoglobin rise of ~3 g/dL, and 62% became transfusion‑independent. Starting treatment earlier (about 12 months after detection vs. 29 months) made a big difference.
Bottom Line: Heavy metal problems—especially copper deficiency with toxic metal excess—are common and treatable causes of unexplained cytopenias and low‑blast MDS. Early testing and correction can lead to major blood count improvements, so heavy metal screening should be part of the standard work‑up in high‑exposure regions.

Abstract: EHA-2607 Wang YH, et al: A FOUR-GENE STEMNESS SCORE PREDICTS HMA RESISTANCE AND IDENTIFIES RATIONAL 5-AZACITIDINE COMBINATION STRATEGIES IN MYELODYSPLASTIC SYNDROMES

What the Study Was About: Researchers wanted to understand why some people with myelodysplastic syndromes (MDS) do not respond well to hypomethylating agents (HMAs) like azacitidine, and whether a stemness‑based gene score could predict resistance and guide better drug combinations.
Who Was in the Study: The main group included 101 MDS patients treated with 5‑azacitidine, all with RNA‑sequencing available at diagnosis. Several independent validation cohorts (VC1–VC5) were used to confirm the findings across different hospitals and datasets.
Key Discovery — The HMA‑LSC4 Score: Researchers identified four stemness‑related genes (LAPTM4B, SOCS2, ZBTB46, MMRN1) that together formed a weighted “HMA‑LSC4” score. A high score was linked to worse survival, more stem‑cell–like biology, and mutations such as ASXL1, SRSF2, and TP53.
Predicting HMA Response: Patients with a low HMA‑LSC4 score responded to azacitidine much more often (77%) than those with a high score (38%). The score also predicted overall survival and AML‑free survival, even after adjusting for IPSS‑R and high‑risk mutations. These results were validated across multiple cohorts.
Identifying Better Drug Combinations: Using computational drug‑response modeling, the team found that patients with high HMA‑LSC4 scores were more sensitive to IAP inhibitors and MCL‑1 inhibitors. Lab testing showed strong synergy when these drugs were combined with azacitidine (synergy scores >30), suggesting a rational combination strategy for HMA‑resistantdisease.
Bottom Line: The new HMA‑LSC4 stemness score reliably predicts who will resist azacitidine and identifies promising combination therapies (IAP or MCL‑1 inhibitors) that may overcome resistance. This approach could help personalize treatment for patients with higher‑risk MDS.

Abstract: EHA-1654 Tseng TC, et al: CLINICAL, GENETIC, AND TRANSCRIPTOMIC CHARACTERIZATION OF MYELODYSPLASTIC SYNDROME/NEOPLASMS WITH AUTOIMMUNE DISEASES

What the Study Was About: Doctors studied whether people with myelodysplastic syndromes (MDS) who also have autoimmune diseases (AID) represent a distinct subtype with different biology and outcomes, and whether the timing of the autoimmune disease (before vs. after MDS) affects prognosis.
Who Was in the Study: The team reviewed 1,196 MDS patients over more than 40 years. 137 patients (11.5%) had an autoimmune disease such as lupus, rheumatoid arthritis, Sjögren’s, autoimmune hemolytic anemia, or thyroiditis. Genetic and RNA‑sequencing data were available for a subset of patients.
Clinical Features of MDS With Autoimmunity: Compared with patients without AID, those with AID had higher white blood cells and platelets, lower bone marrow blasts, and lower IPSS‑R risk, but also lower lymphocyte‑to‑monocyte ratios—suggesting a different immune environment.
Molecular and Transcriptomic Findings: Patients with AID showed upregulation of inflammatory and innate‑immune genes, including FCGR1A, SOCS3, NLRP3, IFI30, IL10, and IL13RA1. Pathway analysis showed activation of interferon‑α, IL‑6/JAK/STAT3, inflammasome, pyroptosis, TLR signaling, and DNA repair, with suppressed heme metabolism—indicating a strong pro‑inflammatory state.
Impact on Survival: After adjusting for age, sex, IPSS‑R, HMA use, and transplant, AID was independently linked to worse survival (OS aHR 1.36; LFS aHR 1.40). The effect depended on timing:
AID before or at MDS diagnosis → significantly worse OS and LFS
AID developing after MDS → no negative impact Patients with early AID also had more high‑risk mutations (RUNX1, TP53, EZH2, IDH2, SRSF2, U2AF1, DNMT3A, NRAS, FLT3‑ITD, KMT2A‑PTD).
Bottom Line: Autoimmune disease marks a distinct, more inflammatory MDS subtype, especially when AID occurs before or at the time of MDS diagnosis. These patients have higher‑risk mutations and worse survival, suggesting that immune dysregulation contributes to disease biology and may require different management strategies.

Abstract: EHA-3320 Marcinek A, et al: LONGITUDINAL IPSS-M RISK EVALUATION IN 205 MDS PATIENTS: IS REASSESSMENT BEYOND INITIAL DIAGNOSIS NECESSARY?

What the Study Was About: Researchers wanted to learn whether the IPSS‑M risk score, which uses mutations plus blood counts and cytogenetics, should be rechecked over time in people with MDS—especially since the score was designed for diagnosis and many patients never had full molecular testing at the start.
Who Was in the Study: The team analyzed 205 patients, each with paired bone marrow samples—one at diagnosis and one later (median 2.1 years apart). All samples had full morphology, cytogenetics, and molecular testing. Patients who had undergone transplant were excluded.
How Often IPSS‑M Changed: About half of all patients changed risk category over time. Most changes were upstaging (moving to a higher‑risk group).
Very low/low: 46% up, 3% down
Moderate low/moderate high: 48% up, 24% down
High/very high: 31% up, 19% down
Clinical Changes Linked to Upstaging: Patients who moved to a higher‑risk IPSS‑M category showed clear signs of disease worsening:
Bone marrow blasts rose (3.5% → 5%)
Hemoglobin dropped (10.5 → 8.6 g/dL)
Platelets fell (124 → 73 ×10⁹/L) These changes were not seen in patients whose IPSS‑M stayed stable.
Genetic Evolution: Upstaged patients often gained new cytogenetic abnormalities (+8, –7, complex karyotype) and new high‑risk mutations, including TP53, CBL, EZH2, NRAS, RUNX1. This shows that clonal evolution is a major driver of IPSS‑M worsening.
Bottom Line: About half of MDS patients shift IPSS‑M category over time, meaning reassessment can be important, especially when blood counts worsen. If molecular testing wasn’t done at diagnosis, doing IPSS‑M later is reasonable. More research is needed to decide whether routine serial IPSS‑M monitoring should guide treatment decisions.

Abstract: EHA-2606 Bayar A, et al: GENOMIC LANDSCAPE AND SURVIVAL OUTCOMES IN MYELODYSPLASTIC SYNDROMES WITH FIBROSIS: A REAL-WORLD ANALYSIS FOR REFINING THE HIGH-RISK PHENOTYPE IN THE IPSS-M ERA

What the Study Was About: Researchers examined whether bone marrow fibrosis (MF)—scarring inside the marrow—defines a more aggressive form of MDS, even in the modern IPSS‑M era, where molecular genetics already guide risk assessment.
Who Was in the Study: The team analyzed 130 adults with primary MDS diagnosed between 2015–2025. About 38% had moderate‑to‑severe fibrosis (MF2/3). Fibrosis was graded using standard European criteria and grouped as MF0/1 vs. MF2/3.
Clinical and Genetic Features of Fibrotic MDS: Patients with MF2/3 were more often male (74% vs. 44%), had higher marrow cellularity, and were much more likely to have excess blasts (39% vs. 10%). In the subset with sequencing, signal‑transduction mutations (e.g., SH2B3, SETBP1) were enriched in MF2/3 (37% vs. 7%).
Impact on Survival: Fibrosis had a major negative effect on outcomes. Median overall survival was 19 months in MF2/3 vs. 72 months in MF0/1. Importantly, fibrosis remained an independent predictor of poor survival even after adjusting for IPSS‑R (HR 2.14) and IPSS‑M (HR 3.37).
Added Prognostic Value: Adding fibrosis to IPSS‑R and IPSS‑M improved model accuracy (C‑index increases from 0.734→0.785 for IPSS‑R and 0.767→0.810 for IPSS‑M). The adverse effect of fibrosis persisted even in <5% blast patients and in lower‑risk scoring groups, showing that fibrosis captures risk not explained by genetics alone.
Bottom Line: Moderate‑to‑severe marrow fibrosis identifies a biologically aggressive MDS subtype with significantly worse survival, independent of molecular risk. Incorporating fibrosis into routine risk scoring could refine prognosis and support earlier treatment escalation.

Abstract: EHA-6714 Pilo F, et al: IPSS-M OUTPERFORMS IPSS-R IN PREDICTING TIME TO TRANSFUSION DEPENDENCY IN PATIENTS WITH LOW RISK MYELODYSPLASTIC SYNDROMES

What the Study Was About: Researchers wanted to see whether the IPSS‑M molecular scoring system is better than the older IPSS‑R at predicting how soon lower‑risk MDS patients become transfusion‑dependent, an important milestone that affects quality of life and treatment decisions.
Who Was in the Study: The team analyzed 70 patients with low or intermediate‑1 IPSS‑R MDS who had full NGS molecular profiling at diagnosis. Patients with high IPSS‑R or other causes of anemia were excluded. At diagnosis, 32% were already transfusion‑dependent.
Incident Transfusion Dependence: Among the 47 patients who were transfusion‑independent at diagnosis, 6 patients (13%) became transfusion‑dependent during follow‑up, with a median time of 13.9 months.
IPSS‑M Strongly Predicted Time to TD:
IPSS‑M: HR 3.67, p=0.008 → significant predictor of faster progression to TD
IPSS‑R: HR 2.31, p=0.08 → not statistically significant When both scores were included in the same model, IPSS‑M remained predictive, while IPSS‑R lost significance. Patients who were low/intermediate‑1 by IPSS‑R but reclassified as high‑risk by IPSS‑M had the shortest time to TD.
Role of SF3B1 Mutations:
SF3B1 single mutations showed a trend toward faster TD (HR 4.47).
After adjusting for IPSS‑M, the association became significant (HR 6.68, p=0.037).
No TD events occurred in SF3B1 co‑mutated cases, suggesting different biology.
Bottom Line: IPSS‑M outperforms IPSS‑R in predicting which lower‑risk MDS patients will become transfusion‑dependent. Molecular risk reclassification—especially identifying patients who appear “low‑risk” clinically but are high‑risk by IPSS‑M—provides important early warning for when to introduce therapies targeting ineffective erythropoiesis. SF3B1 single mutations may add additionalpredictive value.

The MDS Foundation provides trusted information and support for people affected by bone marrow failure diseases. To read the full abstracts from the 2026 EHA Congress visit: https://library.ehaweb.org/eha/2026/eha-2026/4225769/faculty.presenters.abstract.book.2026.html?f=menu%3D5%2Abrowseby%3D8%2Asortby%3D2%2Amedia%3D6%2Alabel%3D29748

Below is a summary of some of the MDS‑related research shared at EHA 2026.

Low Risk MDS

What the Study Was About: Scientists tested ofirnoflast, a new oral drug that blocks NEK7, a key protein that helps activate the NLRP3 inflammasome, which is overactive in MDS and harms blood‑forming cells. The goal was to see if shutting this pathway down could restore healthy blood production in people with lower‑risk, ESA‑refractory MDS.
Who Was in the Study: 37 patients were treated for up to 32 weeks. About half had a high transfusion burden, and others had low or no transfusion needs. Patients represented a mix of MDS subtypes, including del(5q), SF3B1, and MDS/MPN, with a median IPSS‑R score of 3.5.
Safety Findings: Ofirnoflast had an excellent safety profile. Only 27% had treatment‑related side effects, and just two were grade 3 or higher. No treatment‑related serious events occurred, and only one patient stopped treatment because of a side effect.
Main Efficacy Results: Among 30 evaluable patients, 67% had overall hematologic improvement. This included 63% improving red cells, 60% improving platelets, and 50% improving neutrophils. In transfusion‑dependent patients, 55% became transfusion‑independent for at least 8 weeks, and 39% for at least 16 weeks, with a median TI duration of 28.5 weeks.
Blood Count and Remission Responses: Patients who responded had a median hemoglobin rise of 4.2 g/dL, starting from a low baseline of 7.5 g/dL. Remarkably, 30% achieved a complete remission, and another 3% had a marrow CR. Responses occurred across all MDS subtypes and mutation backgrounds.
Bottom Line: Ofirnoflast produced strong, multi‑lineage blood improvements and high transfusion‑independence rates with minimal toxicity. These results provide the first clinical proof that blocking NEK7 can calm the NLRP3 inflammasome and restore healthy blood production, making ofirnoflast a promising new therapy for lower‑risk MDS patients who have few options.

What the Study Was About: Researchers tested luspatercept (LUSPA) at the highest approved dose (1.75 mg/kg) in people with lower‑risk MDS who needed blood transfusions and had either never used ESAs or had stopped responding to them.
Who Was in the Study: 106 patients took part (53 ESA‑naive and 53 ESA‑relapsed/refractory/intolerant). The median age was 76, and most had needed transfusions recently. About half had SF3B1 mutations, and many had low erythropoietin levels, which often predicts better response.
Main Findings in ESA‑Naive Patients: About 81% reached the main goal—8+ weeks without transfusions plus a hemoglobin rise—often within 1 day. 83% stayed transfusion‑free for at least 8 weeks, and 76% for at least 12 weeks. Even patients without ring sideroblasts or with low EPO levels showed strong responses.
Main Findings in ESA‑Relapsed/Refractory Patients: About 62% reached the main goal, usually within 2 days. 68% stayed transfusion‑free for at least 8 weeks, and 51% for at least 12 weeks. Responses were still meaningful even in tougher‑to‑treat groups.
Side Effects and Safety: Most patients had side effects, but they were similar to what is already known for LUSPA. Serious side effects happened in 54–66%, but only 7–11% were treatment‑related. A small number (3.8%) progressed to higher‑risk MDS, and none developed AML.
Bottom Line: Starting LUSPA at the maximum dose led to high and lasting transfusion independence in both ESA‑naive and ESA‑experienced patients. The treatment was well tolerated, and the results support its strong safety and effectiveness in lower‑risk MDS.

Abstract: EHA-4111 Park S, at al: EARLY VERSUS DELAYED ESA IN LOWER-RISK NON -TRANSFUSION-DEPENDENT MDS: FINAL PHASE III EPO-PRETAR RESULTS ON TRANSFUSION DEPENDENCE AND ERYTHROID RESPONSE

What the Study Was About: Doctors studied whether starting EPO alfa early (when hemoglobin was 9–10.5 g/dL) helps people with lower‑risk MDS avoid needing blood transfusions longer, compared with waiting to start EPO until hemoglobin dropped below 9 g/dL.
Who Was in the Study: 84 patients took part (43 early‑start, 41 delayed‑start). In the early group, almost everyone (95%) began EPO right away; in the delayed group, only 80% eventually needed EPO, starting much later (about 16 months after joining the study).
Main Findings on Transfusions: Starting EPO early did NOT delay the time until patients became transfusion‑dependent. About 49% of early‑start patients and 46% of delayed‑start patients eventually needed regular transfusions, with no meaningful difference between groups.
Erythroid Response (Hemoglobin Improvement): Early EPO clearly helped patients’ blood counts rise faster. At 12 weeks, 85% of early‑start patients had a strong hemoglobin response, compared with 55% in the delayed group. Responses also lasted longer with early EPO (about 33 vs 25 months, though not statistically significant).
Survival and Disease Progression: Overall survival and progression‑free survival were similar in both groups. About 4 patients in each arm developed AML, showing no difference in disease worsening.
Safety and Quality of Life: Treatment was well tolerated in both groups, with no new safety concerns. Quality‑of‑life scores stayed similar, whether EPO was started early or delayed.

Abstract: EHA-6964  Komrokji RS, et al: REAL-WORLD OUTCOMES OF IMETELSTAT: INTERROGATING SAFETY, EFFICACY AND PREDICTORS OF RESPONSE IN HEAVILY PRETREATED LOWER-RISK MDS PATIENTS

What the Study Was About: Doctors looked at how well imetelstat, a telomerase‑blocking medicine recently approved for lower‑risk MDS, works in real‑world patients. In the IMerge trial, about 40% of patients became transfusion‑independent, so this study checked if similar results happen outside clinical trials.
Who Was in the Study: 40 lower‑risk MDS patients were treated at one cancer center. Most (36/40) needed transfusions at the start. Patients had many prior treatments—88% had luspatercept, 70% had ESAs, 38% had HMAs, and 43% had lenalidomide. The median hemoglobin before treatment was 7.4 g/dL, and most had SF3B1 mutations (80%).
Main Response Results: The rate of ≥8‑week transfusion independence was 37.5% (15/40), similar to clinical trial results. Responses lasted a median of 185 days (~6 months), and nearly halfof responders were still benefiting at the time of analysis. Responders received more cycles (10 vs. 5).
Predictors of Response: Patients with an ASXL1 mutation responded more often (67% vs. 29%). A higher baseline platelet count (≥192) was the strongest predictor of benefit. Prior HMA or lenalidomide use seemed to reduce response rates, but not significantly.
Safety Findings: Side effects were mostly low blood counts, especially after the first two cycles—35% had grade 3/4 neutropenia and 28% had grade 3/4 thrombocytopenia after cycle 1. Importantly, no cases of febrile neutropenia occurred. Some patients needed dose reductions or supportive care (platelet transfusions or G‑CSF).
Bottom Line: In real‑world practice, imetelstat showed meaningful transfusion independence, even in heavily pretreated patients, and had a manageable safety profile. New potential predictors of response—platelet count and ASXL1 mutation—may help identify patients most likely to benefit.

Abstract: EHA-2044 Campelo MD, et al: Title: UPDATED PHASE 2 RESULTS OF ELRITERCEPT IN PATIENTS (PTS) WITH TRANSFUSION-DEPENDENT (TD) LOWER-RISK MYELODYSPLASTIC NEOPLASMS (LR-MDS)

What the Study Was About: Doctors tested elritercept, a new treatment that targets TGF‑β signaling, to help people with lower‑risk MDS who depend on regular blood transfusions. The goal was to see if it could help patients become transfusion‑independent and improve their blood counts.
Who Was in the Study: 96 patients received the recommended dose (3.75 mg/kg every 4 weeks). Many had been on treatment for over a year (median 13.8 months). Patients included both ring sideroblast–positive and non‑RS groups, and both low and high transfusion burden groups.
Safety Findings: Elritercept was well tolerated. About 46% had treatment‑related side effects, but only 5% were grade 3 or higher. The most common issues were nausea (7%) and diarrhea (6%), with no major safety concerns.
Transfusion Independence Results: By week 24, 38.5% of patients became transfusion‑independent for at least 8 weeks, including 29% of high‑burden patients and 44% of RS‑positive patients. By week 48, 27% achieved 24‑week TI, and responses happened very quickly (median 0.4 weeks). Patients who achieved TI often stayed independent for a long time—median 111 weeks.
Hemoglobin and Blood Count Improvements: HI‑E responses occurred in 49% of patients by week 24, with a median time to response of 3.1 weeks. Responses were seen across subgroups, including non‑RS patients and those with EPO <500 U/L. About 32% had sustained neutrophil increases and 33% had platelet improvements.
Bottom Line: Elritercept produced rapid, durable transfusion independence and strong erythroid responses in lower‑risk MDS, even in harder‑to‑treat groups like high‑burden or non‑RSpatients. The drug was well tolerated, supporting its potential as a meaningful new therapy for transfusion‑dependent LR‑MDS.

Abstract: EHA-3488 Santini V, et al: UPDATED OVERALL SURVIVAL AND LONG-TERM TRANSFUSION INDEPENDENCE IN ESA-NAIVE PATIENTS WITH LOWER-RISK MDS: RESULTS FROM THE PHASE 3 COMMANDS TRIAL

What the Study Was About: This study looked at long‑term results from the phase 3 COMMANDS trial, comparing luspatercept (LUSPA) with epoetin alfa (EA) in ESA‑naive, transfusion‑dependent lower‑risk MDS. The goal was to see whether LUSPA improves overall survival, transfusion independence, and hemoglobin levels over more than 3 years of follow‑up.
Who Was in the Study: Adults with LR‑MDS who had never received an ESA and were transfusion‑dependent were randomized 1:1 to LUSPA or EA. Patients were stratified by transfusion burden, ring sideroblast status, and serum EPO level.
Overall Survival Results: After >3 years of follow‑up, median OS was not reached with LUSPA but was 46 months with EA (HR 0.78). This favorable trend for LUSPA was seen across subgroups, though not statistically significant.
Transfusion Independence & Hemoglobin Responses:
TI ≥12 weeks: 76% with LUSPA vs 56% with EA
TI lasting ≥2.5 years: 25% vs 11%
Median TI duration: 184 weeks vs 95 weeks
Hb rise ≥1.5 g/dL: 80% vs 59%
Duration of Hb response: 73 weeks vs 48 weeks LUSPA produced longer, deeper, and more durable erythroid responses.
Disease Progression & Safety: Fewer LUSPA patients progressed to higher‑risk MDS (3.3% vs 7.3%), and AML progression was similar (5% vs 6%). No new safety issues emerged. About 18% of LUSPA patients were still on treatment at cutoff.
Bottom Line: With more than 3 years of follow‑up, luspatercept continues to outperform epoetin alfa, offering longer survival trends, much higher and more durable transfusion independence, and stronger hemoglobin improvements, with no new safety concerns. These results reinforce LUSPA as the preferred first‑line therapy for ESA‑naive, transfusion‑dependentLR‑MDS.

What the Study Was About: Doctors studied which patients with hypoplastic MDS (MDS‑h) or MDS with low blasts (MDS‑LB) actually benefit from allogeneic stem cell transplant (allo‑HSCT). These subtypes usually have low blasts and no defining cytogenetic abnormalities, so it’s unclear who truly needs transplant.
Who Was in the Study: The team analyzed 636 patients (146 MDS‑h, 490 MDS‑LB) treated between 2016–2024. The median age was 52, and 42% were transfusion‑dependent (TD) at diagnosis. Over half (52%) were IPSS‑M higher‑risk. A total of 112 patients eventually underwent transplant.
Impact of Transfusion Dependence: TD was a major negative prognostic factor, regardless of IPSS‑M risk.
5‑year OS in TD: 39% (higher‑risk) and 49% (lower‑risk)
5‑year OS in TI: 65% (higher‑risk) and 81% (lower‑risk) This shows that transfusion status adds important risk information beyond IPSS‑M.
Who Benefited From Transplant:
In TD + IPSS‑M higher‑risk patients, transplant significantly improved survival (5‑year OS 86% with HSCT vs 47% without, P=0.004).
In TD + lower‑risk patients, no benefit was seen.
In TI patients, transplant helped only those with higher‑risk IPSS‑M, while lower‑risk TI patients had excess early mortality after HSCT.
Progression to Transfusion Dependence: Among patients who were TI at diagnosis, 46% eventually became TD. The strongest predictors of becoming TD were IPSS‑M higher‑risk (HR 2.41) and MDS‑LB subtype (HR 2.20).
Bottom Line: Transplant decisions in MDS‑h and MDS‑LB should consider both IPSS‑M molecular risk and transfusion status.
Clear benefit: TD + IPSS‑M higher‑risk
Possible benefit: TI + IPSS‑M higher‑risk
No benefit / potential harm: Lower‑risk patients, especially if TI This supports a risk‑adapted, personalized transplant strategy.

What the Study Was About: Researchers looked at which lower‑risk MDS patients with anemia benefit from roxadustat, an oral drug that boosts red blood cell production. This post‑hocanalysis focused on patients with high transfusion burden (HTB) and those with or without ring sideroblasts (RS).
Who Was in the Study: The analysis included 140 patients from the MATTERHORN Phase 3 trial. For this post‑hoc study, 37 patients met criteria for high transfusion burden (22 on roxadustat, 15 on placebo). Patients had very low, low, or intermediate‑risk MDS, with or without prior ESA exposure.
Main Transfusion Independence Results (HTB Group):
8‑week TI (first 28 weeks): 36% with roxadustat vs 7% with placebo (p=0.041)
8‑week TI (first 52 weeks): 46% vs 13% (p=0.043)
Longer TI durations also favored roxadustat. These results show meaningful and durable reductions in transfusion needs in a group that is typically very hard to treat.
Effect in Ring Sideroblast–Positive and –Negative Patients: In the full analysis set with RS data, both RS+ and RS– patients showed improvements in transfusion independence and hemoglobin levels, suggesting roxadustat works regardless of RS status.
Safety Findings: Side effects were mostly low‑grade and manageable. No new safety concerns were identified, and the drug was generally well tolerated.
Bottom Line: Roxadustat produced clinically meaningful and durable transfusion independence, especially in high‑burden LR‑MDS patients, with benefits seen in both RS+ and RS– disease. As an oral option, it addresses a major unmet need, and a new global Phase 3 trial is planned for 2026.

Abstract: EHA-2040 Liu Y, et al: HETROMBOPAG FOR THROMBOCYTOPENIA IN LOWER-RISK MYELODYSPLASTIC SYNDROMES: A PROSPECTIVE SINGLE-CENTER STUDY

What the Study Was About: Doctors tested hetrombopag, an oral thrombopoietin‑receptor agonist, to see whether it can safely raise platelet counts in lower‑risk MDS patients with thrombocytopenia, a group with few treatment options and high bleeding risk.
Who Was in the Study: 56 adults with LR‑MDS and low platelets were followed for a median of 61 weeks. All had <5% blasts, IPSS low or Int‑1 risk, and platelets either <30×10⁹/L or 30–100×10⁹/L with bleeding. Hetrombopag was started at 5 mg/day and increased up to 15 mg/day based on response.
Platelet Response Results:
Overall platelet response: 59% (33/56)
Responses within 24 weeks: 54%
Complete responses: 10 patients
Median time to response: 6 weeks
Duration of response: Not reached Patients with platelets <30×10⁹/L or IPSS‑M >0 responded less often.
Transfusion & Bleeding Improvements:
Among 21 platelet‑transfusion‑dependent patients, 57% became transfusion‑independent.
Among 30 RBC‑transfusion‑dependent patients, 30% became transfusion‑independent.
Bleeding events decreased during treatment.
Erythroid improvement occurred in 50% of anemic patients.
Safety Findings: Side effects occurred in 15 patients, mostly infections. No blood clots (arterial or venous) were seen. Serious grade 3–4 events (e.g., pulmonary infection, angina, heart failure, sepsis) were not considered drug‑related. Only 2 patients showed disease progression (one to IB2, one to AML).
Bottom Line: Hetrombopag produced meaningful platelet increases, reduced bleeding and transfusion needs, and showed no signal of increased leukemic progression. The drug appears safe and effective for thrombocytopenic LR‑MDS and deserves further study.

High Risk MDS

What the Study Was About: Doctors compared CPX‑351 with standard treatments (azacitidine or regular chemotherapy) to see which worked better for people with high‑risk MDS or early AML who were planning to get a stem cell transplant within 6 months.
Who Was in the Study: 150 patients joined the trial. They were randomly placed into either the CPX‑351 group (75 people) or the standard‑care group (75 people total: 52 got chemo, 23 got azacitidine).
Main Findings: Patients who received CPX‑351 stayed free of major problems longer. Their event‑free survival (EFS) was not reached, meaning many were still doing well at follow‑up. The standard‑care group had a median EFS of 11.1 months. Overall survival looked similar so far (CPX‑351 not reached vs. 35.4 months), but this part was not statistically significant yet.
Transplant Results: A very high number of patients in both groups were able to move forward with their stem cell transplant—57 people (85%) in the CPX‑351 group and 64 people (86%) in the standard‑care group.
Side Effects: The most common serious side effects were low platelets (65 CPX‑351 vs. 55 standard care), low white blood cells (55 vs. 46), and fever with infection (44 vs. 35). These side effects are expected with these types of treatments.
Bottom Line: CPX‑351 met the main goal of the study by giving patients longer time before their disease got worse compared with standard treatments, while still allowing most patients to reach transplant.

Abstract: EHA-4187 Streuer A, et al: FUNCTIONAL TP53 PHENOTYPE SCORE REFINES PROGNOSTIC STRATIFICATION IN MONOALLELIC TP53-MUTATED MYELODYSPLASTIC NEOPLASMS BEYOND ALLELIC STATE

What the Study Was About: Researchers looked at whether a TP53 functional phenotype score (PS)—a number showing how damaging a TP53 mutation is—can better predict outcomes in monoallelic TP53‑mutated MDS, a group known to have mixed and unpredictable survival.
Who Was in the Study: The team analyzed 4,505 patients with MDS from two large cohorts. Among them, 271 had monoallelic TP53 mutations and 499 had multi‑hit TP53, which is known to be very high‑risk. Every TP53 variant was scored using a lab‑based system that measures how much the mutation disrupts TP53 function.
Main Findings on Survival: Survival dropped stepwise across groups: TP53‑wild‑type (42 months) → TP53‑monoallelic (23 months) → TP53‑multi‑hit (9 months). Within the monoallelic group, patients with high PS (≥7) had much worse survival (14 months) than those with low PS (<7) (39 months).
AML Progression and Risk Features: High‑PS patients had a higher risk of transforming to AML and were more likely to have complex karyotypes, a known high‑risk feature. Importantly, PS identified high‑risk patients even inside “low‑risk” clinical categories.
Role of VAF (Mutation Burden): A higher VAF (≥22%) also predicted worse survival, but PS and VAF were not correlated, meaning they measure different biological risks. When combined, they sharply separated outcomes—from near wild‑type survival (≈50 months) to multi‑hit‑like survival (≈9 months).
Bottom Line: Monoallelic TP53‑mutated MDS is not one single disease. Using functional TP53 scores plus VAF gives a much clearer picture of risk than allelic state alone and may help identify patients who need earlier aggressive treatment or transplant evaluation.

What the Study Was About: Doctors tested a new oral version of decitabine/cedazuridine (DEC‑C) in people with higher‑risk MDS or CMML in China. This oral form is designed to give the same drug exposure as the IV version but in a more convenient pill.
Who Was in the Study: 73 patients were treated across 15 centers. Most (96%) had MDS, and the median age was 65 years. Part 1 of the study compared oral DEC‑C with IV decitabine in a crossover design to confirm equivalent drug levels.
Main Efficacy Results: After a median of 5 treatment cycles, the overall response rate was 54.8% using IWG 2006 criteria. This included 22% complete responses, plus additional marrow CRand hematologic improvement responses. Using newer IWG 2023 criteria, the ORR was 56%, again showing strong activity similar to the FDA‑approved ASTX727.
Survival Outcomes: After about 9 months of follow‑up, the median overall survival was 19.1 months, and the 12‑month survival rate was 76%. Leukemia‑free survival was also 19.1 months, with a 71% 12‑month LFS rate.
Safety Findings: The oral DEC‑C had a similar safety profile to IV decitabine. About 7% of patients had treatment‑emergent events that led to death, and 4% were considered treatment‑related (pneumonia, sepsis, septic shock). No unexpected toxicities were seen.
Bottom Line: The new oral DEC‑C showed strong, consistent efficacy and a favorable safety profile, supporting its use as the first oral hypomethylating agent option for MDS patients in China, comparable to the established IV formulation.

Abstract: EHA-2146 Nie ZY, et al. BORTEZOMIB COMBINED WITH AZACITIDINE AND VENETOCLAX IN HIGH-RISK MYELODYSPLASTIC NEOPLASMS: A SINGLE-CENTER RETROSPECTIVE STUDY

What the Study Was About: Doctors tested a three‑drug combination—bortezomib + azacitidine + venetoclax (called VAB)—to see whether it could improve outcomes for people with high‑risk MDS, a group that usually responds poorly to current treatments.
Who Was in the Study: 20 patients with newly diagnosed high‑risk MDS were treated. The median age was 51 which is younger than typical MDS populations. About 45% had EB1, 45% had EB2, and 25% had complex karyotypes. The median number of mutations was 3, and 40% later went to stem cell transplant.
Early Response Results: The overall composite remission rate (CRc) was 75%, meaning 3 out of 4 patients had a meaningful response. After just one cycle, 60% had already reached CRc. The high‑risk group responded better than the very‑high‑risk group (100% vs. 55%).
Survival Outcomes: After a median follow‑up of 17.4 months, only 4 patients had died and 2 progressed to AML. The median overall survival had not yet been reached, suggesting encouraging durability. The 1‑year and 2‑year survival rates were both 78.6%.
Safety Findings: The most common side effects were grade ≥3 low blood counts, which were expected and manageable. No unexpected toxicities were reported, and the regimen was considered tolerable.
Bottom Line: The VAB triplet (bortezomib + azacitidine + venetoclax) produced high remission rates, rapid responses, and promising survival in newly diagnosed high‑risk MDS. This early experience suggests the regimen may offer a new, more effective first‑line option for this difficult‑to‑treat group.

What the Study Was About: Doctors compared outcomes in treatment‑naïve MDS with excess blasts (MDS‑EB) treated with hypomethylating agents (HMAs) alone versus HMA + venetoclax(HMAVen). The goal was to identify which subgroups actually benefit from adding venetoclax.
Who Was in the Study: The study included 358 patients (245 HMA, 113 HMAVen) treated from 2010–2025. Median age was ~70. Blast levels were higher in the HMAVen group (12% vs 8%). High‑risk and very‑high‑risk IPSS‑R categories were similar between groups. Mutations were balanced, including TP53 (~33%), ASXL1 (23–32%), and DNMT3A (~15%).
Response Rates: Adding venetoclax significantly improved overall response rates:
IWG 2006 ORR: 62% (HMA) vs 89% (HMAVen)
IWG 2023 ORR: 51% vs 77% Complete remission rates were similar (~35%). More HMAVen patients proceeded to transplant (43% vs 26%).
Overall Survival: Despite higher response rates, median OS was similar:
19 months (HMA) vs 18 months (HMAVen) Two‑year OS was also comparable (41% vs 46%). No OS differences were seen across HMA types.
Subgroups That Benefited From Venetoclax: In unadjusted analyses, HMAVen improved OS in:
ASXL1‑mutated patients (HR 0.45)
IPSS‑R very high‑risk (HR 0.65)
IPSS‑M very high‑risk (HR 0.65)
BM blasts ≥10% (HR 0.69) After adjusting for transplant, these benefits weakened, suggesting venetoclax mainly helps by increasing the chance of reaching transplant. Markov modeling confirmed HMAVen doubled the likelihood of transitioning to HSCT in these subgroups.
Bottom Line: Venetoclax does not improve overall survival for all MDS‑EB patients, but specific high‑risk groups—ASXL1‑mutated, very‑high‑risk IPSS‑R/IPSS‑M, or blasts ≥10%—appear to benefit. The main advantage is higher response rates and greater access to transplant, which likely drives survival improvements.

Diagnostic and Prediction Studies

Abstract: EHA-2407 Della Porta MG, et al: CYTO-CMML: A MUTATION-INDEPENDENT CYTOGENETIC RISK CLASSIFIER FOR CHRONIC MYELOMONOCYTIC LEUKEMIA

What the Study Was About: Researchers wanted to improve how doctors predict outcomes in chronic myelomonocytic leukemia (CMML) by creating a new cytogenetic (chromosome-based) risk system that does not depend on gene mutations. This helps guide decisions about stem cell transplant, the only curative option.
Who Was in the Study: The team analyzed over 2,100 patients with CMML from two large groups (971 in the discovery set and 1,175 in the validation set). Patients were older (median age ~72–73), and follow‑up lasted about 4–5 years.
Main Findings on Survival: Some chromosome changes (like –Y, 5q–, 7q–) had survival similar to normal karyotype (~38 months). But others—such as complex karyotype, –7, +21, i(17q), inv(3), t(1;3)—had very poor survival (~10 months). All other abnormalities fell in between (~22 months).
Risk Groups and Blast Transformation: Only the high‑risk cytogenetic group had a clearly higher chance of turning into AML, with a 1‑year blast transformation rate of 33%, compared with 8–9% in the other groups. This pattern was consistent in both study cohorts.
Independent Predictors: Even after adjusting for other factors (mutations, blood counts, blast percentage), high‑risk cytogenetics remained a strong predictor of both shorter survival and higher AML risk. The intermediate group was less reliable, so researchers combined low + intermediate into a single “standard‑risk” category.
Bottom Line: The new Cyto‑CMML system is a simple, mutation‑independent way to classify CMML patients into standard‑risk or high‑risk groups. It improves prediction of survival and AML progression and may help doctors decide who needs earlier transplant or more aggressive treatment.

Abstract: EHA-1673 Palenzuela LG, et al: VALIDATION OF MOLECULAR PROGNOSTIC MODELS FOR CHRONIC MYELOMONOCYTIC LEUKEMIA IN A LARGE MULTICENTER COHORT

What the Study Was About: Doctors tested how well molecular risk models (which use gene mutations) predict survival and AML transformation in chronic myelomonocytic leukemia (CMML). These models include GFM, MMM, CPSSmol, and BLASTmol, but they haven’t been validated in large real‑world groups until now.
Who Was in the Study: The study included 977 CMML patients from multiple centers. The median age was 71, median overall survival was 29 months, and about 21% eventually developed AML. Most patients received supportive care, while those who had a stem cell transplant lived the longest (about 52 months).
Clinical and Genetic Risk Factors: Worse survival was linked to high white blood cells, high monocytes, >5% bone marrow blasts, and circulating immature cells. High‑risk cytogenetics shortened survival (15 vs. 39 months). Mutations in ASXL1, RUNX1, and TP53 were also associated with poorer outcomes.
How the Models Performed: Most molecular models separated patients into clear risk groups. MMM performed best, showing a strong stepwise drop in survival (from 106 months in low‑risk to 20 months in high‑risk). BLASTmol also clearly identified high‑risk patients. GFM worked but had overlap in the middle groups. CPSSmol performed the weakest and did not reach statistical significance.
AML Transformation Results: AML progression patterns matched the survival results. MMM again showed the clearest separation, with high‑risk patients having a 41% chance of AML at 3 years. BLASTmol also identified high‑risk transformation, while GFM and CPSSmol did not significantly separate groups.
Bottom Line: Molecular CMML‑specific models—especially MMM—are better than older clinical scoring systems at predicting survival and AML risk. This supports using molecular profiling to guide treatment decisions, including when to consider transplant.

What the Study Was About: Doctors studied people with lower‑risk MDS who had very high erythropoietin levels (≥500 IU/L) to understand whether this group has different symptoms, lab results, or survival compared with patients whose EPO levels are lower.
Who Was in the Study: The analysis included 435 lower‑risk MDS patients from 50 centers in Poland. Of these, 56 patients (13%) had sEPO ≥500 IU/L, while the rest had lower levels and served as the comparison group.
Clinical Features of High‑EPO Patients: Patients with sEPO ≥500 IU/L were younger (about 69 vs. 72 years) but had much more severe anemia (hemoglobin 7.65 vs. 9.1 g/dL). They also had higher ferritin levels, suggesting more iron overload.
Disease Burden and Risk Features: High‑EPO patients were far more likely to be transfusion‑dependent (77% vs. 44%) and had more bone marrow fibrosis (13% vs. 7%). They also had higher rates of tobacco exposure, which may worsen inflammation and anemia.
Survival Outcomes: Patients with sEPO ≥500 IU/L had significantly shorter survival. Their median overall survival was 33 months, compared with 48 months in the lower‑EPO group, showing that high EPO is a marker of more aggressive disease biology.
Bottom Line: An sEPO level ≥500 IU/L identifies a distinct high‑risk subgroup within lower‑risk MDS—characterized by severe anemia, high transfusion needs, fibrosis, and worse survival. These patients may need more personalized and proactive management.

Abstract: EHA-2079 Saini K, et al: HEAVY METAL ABNORMALITIES IN UNEXPLAINED CYTOPENIAS AND LOW‑BLAST MYELODYSPLASTIC NEOPLASMS: A PROSPECTIVE INTERVENTIONAL COHORT STUDY FROM NORTH INDIA

What the Study Was About: Doctors looked at whether heavy metal problems—like low copper or high levels of toxic metals—might explain unexplained low blood counts or low‑blast MDS, and whether fixing these problems could improve blood counts.
Who Was in the Study: 120 patients from North India were followed. All had unexplained cytopenias or low‑blast MDS, and common causes (like B12 deficiency or infections) were ruled out. The median age was 46, and about 54% had low‑blast MDS.
How Common Heavy Metal Issues Were: Heavy metal abnormalities were found in two‑thirds of patients (67%), and one‑third had more than one abnormality. Copper deficiency was the most common (~65%), while excess zinc, lead, and arsenic were the main toxic exposures.
Clinical Clues: Some patients had ring sideroblasts (10.6%), which were linked to high mercury and zinc levels. These findings often mimicked low‑risk MDS, making diagnosis tricky without metal testing.
Treatment Responses: Among 50 patients with follow‑up, 70% improved after targeted treatment (copper replacement or chelation). Responders had an average hemoglobin rise of ~3 g/dL, and 62% became transfusion‑independent. Starting treatment earlier (about 12 months after detection vs. 29 months) made a big difference.
Bottom Line: Heavy metal problems—especially copper deficiency with toxic metal excess—are common and treatable causes of unexplained cytopenias and low‑blast MDS. Early testing and correction can lead to major blood count improvements, so heavy metal screening should be part of the standard work‑up in high‑exposure regions.

Abstract: EHA-2607 Wang YH, et al: A FOUR-GENE STEMNESS SCORE PREDICTS HMA RESISTANCE AND IDENTIFIES RATIONAL 5-AZACITIDINE COMBINATION STRATEGIES IN MYELODYSPLASTIC SYNDROMES

What the Study Was About: Researchers wanted to understand why some people with myelodysplastic syndromes (MDS) do not respond well to hypomethylating agents (HMAs) like azacitidine, and whether a stemness‑based gene score could predict resistance and guide better drug combinations.
Who Was in the Study: The main group included 101 MDS patients treated with 5‑azacitidine, all with RNA‑sequencing available at diagnosis. Several independent validation cohorts (VC1–VC5) were used to confirm the findings across different hospitals and datasets.
Key Discovery — The HMA‑LSC4 Score: Researchers identified four stemness‑related genes (LAPTM4B, SOCS2, ZBTB46, MMRN1) that together formed a weighted “HMA‑LSC4” score. A high score was linked to worse survival, more stem‑cell–like biology, and mutations such as ASXL1, SRSF2, and TP53.
Predicting HMA Response: Patients with a low HMA‑LSC4 score responded to azacitidine much more often (77%) than those with a high score (38%). The score also predicted overall survival and AML‑free survival, even after adjusting for IPSS‑R and high‑risk mutations. These results were validated across multiple cohorts.
Identifying Better Drug Combinations: Using computational drug‑response modeling, the team found that patients with high HMA‑LSC4 scores were more sensitive to IAP inhibitors and MCL‑1 inhibitors. Lab testing showed strong synergy when these drugs were combined with azacitidine (synergy scores >30), suggesting a rational combination strategy for HMA‑resistantdisease.
Bottom Line: The new HMA‑LSC4 stemness score reliably predicts who will resist azacitidine and identifies promising combination therapies (IAP or MCL‑1 inhibitors) that may overcome resistance. This approach could help personalize treatment for patients with higher‑risk MDS.

Abstract: EHA-1654 Tseng TC, et al: CLINICAL, GENETIC, AND TRANSCRIPTOMIC CHARACTERIZATION OF MYELODYSPLASTIC SYNDROME/NEOPLASMS WITH AUTOIMMUNE DISEASES

What the Study Was About: Doctors studied whether people with myelodysplastic syndromes (MDS) who also have autoimmune diseases (AID) represent a distinct subtype with different biology and outcomes, and whether the timing of the autoimmune disease (before vs. after MDS) affects prognosis.
Who Was in the Study: The team reviewed 1,196 MDS patients over more than 40 years. 137 patients (11.5%) had an autoimmune disease such as lupus, rheumatoid arthritis, Sjögren’s, autoimmune hemolytic anemia, or thyroiditis. Genetic and RNA‑sequencing data were available for a subset of patients.
Clinical Features of MDS With Autoimmunity: Compared with patients without AID, those with AID had higher white blood cells and platelets, lower bone marrow blasts, and lower IPSS‑R risk, but also lower lymphocyte‑to‑monocyte ratios—suggesting a different immune environment.
Molecular and Transcriptomic Findings: Patients with AID showed upregulation of inflammatory and innate‑immune genes, including FCGR1A, SOCS3, NLRP3, IFI30, IL10, and IL13RA1. Pathway analysis showed activation of interferon‑α, IL‑6/JAK/STAT3, inflammasome, pyroptosis, TLR signaling, and DNA repair, with suppressed heme metabolism—indicating a strong pro‑inflammatory state.
Impact on Survival: After adjusting for age, sex, IPSS‑R, HMA use, and transplant, AID was independently linked to worse survival (OS aHR 1.36; LFS aHR 1.40). The effect depended on timing:
AID before or at MDS diagnosis → significantly worse OS and LFS
AID developing after MDS → no negative impact Patients with early AID also had more high‑risk mutations (RUNX1, TP53, EZH2, IDH2, SRSF2, U2AF1, DNMT3A, NRAS, FLT3‑ITD, KMT2A‑PTD).
Bottom Line: Autoimmune disease marks a distinct, more inflammatory MDS subtype, especially when AID occurs before or at the time of MDS diagnosis. These patients have higher‑risk mutations and worse survival, suggesting that immune dysregulation contributes to disease biology and may require different management strategies.

Abstract: EHA-3320 Marcinek A, et al: LONGITUDINAL IPSS-M RISK EVALUATION IN 205 MDS PATIENTS: IS REASSESSMENT BEYOND INITIAL DIAGNOSIS NECESSARY?

What the Study Was About: Researchers wanted to learn whether the IPSS‑M risk score, which uses mutations plus blood counts and cytogenetics, should be rechecked over time in people with MDS—especially since the score was designed for diagnosis and many patients never had full molecular testing at the start.
Who Was in the Study: The team analyzed 205 patients, each with paired bone marrow samples—one at diagnosis and one later (median 2.1 years apart). All samples had full morphology, cytogenetics, and molecular testing. Patients who had undergone transplant were excluded.
How Often IPSS‑M Changed: About half of all patients changed risk category over time. Most changes were upstaging (moving to a higher‑risk group).
Very low/low: 46% up, 3% down
Moderate low/moderate high: 48% up, 24% down
High/very high: 31% up, 19% down
Clinical Changes Linked to Upstaging: Patients who moved to a higher‑risk IPSS‑M category showed clear signs of disease worsening:
Bone marrow blasts rose (3.5% → 5%)
Hemoglobin dropped (10.5 → 8.6 g/dL)
Platelets fell (124 → 73 ×10⁹/L) These changes were not seen in patients whose IPSS‑M stayed stable.
Genetic Evolution: Upstaged patients often gained new cytogenetic abnormalities (+8, –7, complex karyotype) and new high‑risk mutations, including TP53, CBL, EZH2, NRAS, RUNX1. This shows that clonal evolution is a major driver of IPSS‑M worsening.
Bottom Line: About half of MDS patients shift IPSS‑M category over time, meaning reassessment can be important, especially when blood counts worsen. If molecular testing wasn’t done at diagnosis, doing IPSS‑M later is reasonable. More research is needed to decide whether routine serial IPSS‑M monitoring should guide treatment decisions.

What the Study Was About: Researchers examined whether bone marrow fibrosis (MF)—scarring inside the marrow—defines a more aggressive form of MDS, even in the modern IPSS‑M era, where molecular genetics already guide risk assessment.
Who Was in the Study: The team analyzed 130 adults with primary MDS diagnosed between 2015–2025. About 38% had moderate‑to‑severe fibrosis (MF2/3). Fibrosis was graded using standard European criteria and grouped as MF0/1 vs. MF2/3.
Clinical and Genetic Features of Fibrotic MDS: Patients with MF2/3 were more often male (74% vs. 44%), had higher marrow cellularity, and were much more likely to have excess blasts (39% vs. 10%). In the subset with sequencing, signal‑transduction mutations (e.g., SH2B3, SETBP1) were enriched in MF2/3 (37% vs. 7%).
Impact on Survival: Fibrosis had a major negative effect on outcomes. Median overall survival was 19 months in MF2/3 vs. 72 months in MF0/1. Importantly, fibrosis remained an independent predictor of poor survival even after adjusting for IPSS‑R (HR 2.14) and IPSS‑M (HR 3.37).
Added Prognostic Value: Adding fibrosis to IPSS‑R and IPSS‑M improved model accuracy (C‑index increases from 0.734→0.785 for IPSS‑R and 0.767→0.810 for IPSS‑M). The adverse effect of fibrosis persisted even in <5% blast patients and in lower‑risk scoring groups, showing that fibrosis captures risk not explained by genetics alone.
Bottom Line: Moderate‑to‑severe marrow fibrosis identifies a biologically aggressive MDS subtype with significantly worse survival, independent of molecular risk. Incorporating fibrosis into routine risk scoring could refine prognosis and support earlier treatment escalation.

Abstract: EHA-6714 Pilo F, et al: IPSS-M OUTPERFORMS IPSS-R IN PREDICTING TIME TO TRANSFUSION DEPENDENCY IN PATIENTS WITH LOW RISK MYELODYSPLASTIC SYNDROMES

What the Study Was About: Researchers wanted to see whether the IPSS‑M molecular scoring system is better than the older IPSS‑R at predicting how soon lower‑risk MDS patients become transfusion‑dependent, an important milestone that affects quality of life and treatment decisions.
Who Was in the Study: The team analyzed 70 patients with low or intermediate‑1 IPSS‑R MDS who had full NGS molecular profiling at diagnosis. Patients with high IPSS‑R or other causes of anemia were excluded. At diagnosis, 32% were already transfusion‑dependent.
Incident Transfusion Dependence: Among the 47 patients who were transfusion‑independent at diagnosis, 6 patients (13%) became transfusion‑dependent during follow‑up, with a median time of 13.9 months.
IPSS‑M Strongly Predicted Time to TD:
IPSS‑M: HR 3.67, p=0.008 → significant predictor of faster progression to TD
IPSS‑R: HR 2.31, p=0.08 → not statistically significant When both scores were included in the same model, IPSS‑M remained predictive, while IPSS‑R lost significance. Patients who were low/intermediate‑1 by IPSS‑R but reclassified as high‑risk by IPSS‑M had the shortest time to TD.
Role of SF3B1 Mutations:
SF3B1 single mutations showed a trend toward faster TD (HR 4.47).
After adjusting for IPSS‑M, the association became significant (HR 6.68, p=0.037).
No TD events occurred in SF3B1 co‑mutated cases, suggesting different biology.
Bottom Line: IPSS‑M outperforms IPSS‑R in predicting which lower‑risk MDS patients will become transfusion‑dependent. Molecular risk reclassification—especially identifying patients who appear “low‑risk” clinically but are high‑risk by IPSS‑M—provides important early warning for when to introduce therapies targeting ineffective erythropoiesis. SF3B1 single mutations may add additionalpredictive value.

The MDS Foundation provides trusted information and support for people affected by bone marrow failure diseases. To read the full abstracts from the 2026 EHA Congress visit: https://library.ehaweb.org/eha/2026/eha-2026/4225769/faculty.presenters.abstract.book.2026.html?f=menu%3D5%2Abrowseby%3D8%2Asortby%3D2%2Amedia%3D6%2Alabel%3D29748

Join the Fight Against MDS

Low Risk MDS

High Risk MDS

Diagnostic and Prediction Studies

Get in touch

MDS Foundation EHA Summaries 2026

Low Risk MDS

High Risk MDS

Diagnostic and Prediction Studies