Genetic Insights and Novel Trial Designs Contribute to an Era of Individualized MDS Management

Rami Komrokji, MD

Myelodysplastic syndrome (MDS) treatment selection is increasingly being guided by genetic disease features, which are also informing the study of novel agents. These novel agents may soon move the field beyond symptom management toward personalized disease modification, which could improve long-term treatment outcomes, according to Rami Komrokji, MD.

“Patients with deletion 5q [del(5q)] preferentially receive lenalidomide [Revlimid],” Komrokji said in a summary of the lower-risk MDS treatment paradigm during an interview with OncLive®. “Ring sideroblast [RS]–positive patients preferentially receive luspatercept-aamt [Reblozyl] up front. For the rest, you have a choice between erythropoiesis-stimulating agents [ESAs] and luspatercept as frontline therapy. Then we have imetelstat [Rytelo] for patients after ESA or luspatercept progression, particularly in those who are highly transfusion dependent.”

Luspatercept was FDA approved in 2023 for the treatment of anemia in patients with very low– to intermediate-risk, transfusion-dependent MDS based on data from the phase 3 COMMANDS trial (NCT03682536), in which the rate of red blood cell (RBC) transfusion independence for at least 12 weeks combined with a mean hemoglobin level increase of at least 1.5 g/dL within the first 24 weeks was 58.5% among patients who received luspatercept (n = 86) vs 31.2% among those who received epoetin alfa (n = 48).1

Furthermore, imetelstat was FDA approved in 2024 for the treatment of anemia in patients with low- to intermediate-1–risk transfusion-dependent MDS who have lost response to, not responded to, or are not eligible for ESAs.2 This regulatory decision was backed by findings from the phase 3 IMerge trial (NCT02598661), in which patients in the imetelstat arm (n = 118) achieved an 8-week RBC transfusion independence rate of 39.8% (95% CI, 30.9%-49.3%) vs 15% (95% CI, 7.1%-26.6%) in the placebo arm (n = 59; P <.001).

In the interview, Komrokji discussed the evolution of treatments for patients with lower-risk MDS that reflect novel classification systems; how major clinical trials like COMMANDS and IMerge have introduced agents that have improved transfusion independence rates in patients with lower-risk disease; ongoing trials that are exploring novel targets and therapies, including oral hypomethylating agents (HMAs) and NLRP3 inhibitors; and future treatment sequencing considerations for incorporating immunotherapy into the MDS armamentarium. 

Komrokji is vice chair of the Malignant Hematology Department and a senior member of the Malignant Hematology and Experimental Therapeutics Program at Moffitt Cancer Center in Tampa, Florida; as well as a professor in medicine & oncologic sciences in the College of Medicine at the University of South Florida.

OncLive: How does the current lower-risk MDS landscape reflect recent updates to molecular testing practices and classification systems?

Komrokji: We've made a lot of progress in better understanding the disease. Now we even talk about a precursor state before MDS. We have patients with clonal hematopoiesis of indeterminate potential [CHIP] and clonal cytopenia of undetermined significance [CCUS]. Patients with CCUS will have low blood counts. They will have somatic mutations or cytogenetic abnormalities, but not enough dysplasia to call it MDS. [In 2024], we published [research looking at data] from Moffitt and the National MDS Study [NCT02775383] showing that high-risk CCUS outcomes and depth of cytopenia are similar to [those of] lower-risk MDS. We are broadening [the population of patients that] we might consider lower risk in the future.

The classification has moved into a molecular definition of certain entities. Classification is a reflection of biology. [We have tried to identify] a group of patients who have a certain biology to explain the underlying ineffective hematopoiesis. In the newer [MDS] classification systems, whether the World Health Organization or the International Consensus Classification, we have entities such as del(5q), which we've known about for many years; SF3B1 mutations; p53 mutations; and patients with low and high blast counts. Regarding what we classify as lower risk, it's del(5q), SF3B1 mutations, and low blast counts.

Historically, classification [has been] more pathological. However, [the updated classification systems] emphasize important clinical aspects. For patients with del(5q), we have a unique treatment, lenalidomide that we've used for many years. For [patients with] SF3B1 mutations, now luspatercept has emerged as a preferred treatment, even up front. Classification is affecting our choice of therapy.

We've also refined what we consider lower risk by refining the risk stratification models. We have now the International Prognostic Scoring System—Molecular criteria where we integrate the clinical variables of cytopenias, blasts, and cytogenetics, and we also add the molecular mutations. [From there], we divide patients into 6 groups. The first 3 are considered lower risk. [This system] upstages a subset of patients, but even in the most [updated] risk stratification criteria, approximately 58% of patients [with MDS] are considered lower risk.

[We also need to consider] the natural history of patients with lower-risk MDS. What happens to those patients? [Approximately] 30% of patients, including all-comers, will progress to higher-risk MDS or acute myeloid leukemia [AML], and [approximately] 70% of patients will remain lower risk. The main issue in [the lower-risk] population is the complications from cytopenia. In many of those patients the cause of death, unfortunately, is still attributed to MDS, whether it's a direct result of those cytopenias or an interplay with comorbidity. Cytopenias are not trivial. Historically, the goal of treatment in those patients was to make them transfusion independent, mainly as a palliative [method] to manage symptomatic anemia. However, the field is moving more into [investigating whether we] could alter the natural history [of the disease] if we restore effective hematopoiesis in those patients.

Another [research topic] that's coming up is the causes of mortality among lower-risk patients with MDS. A subset of patients [with MDS] and AML will progress, and that's important, but most patients will not progress. However, what's emerging is that patients [with MDS] are at higher risk of cardiac events. Cardiac mortality is a common cause of mortality.

Historically, we used to think of [this issue] as simplistic: patients become anemic, and they have cardiac stress. However, there is evolving literature on the importance of CHIP, somatic mutations, inflammatory milieu, and accelerated atherosclerosis. The story may not be just cytopenias; the clonal hematopoiesis and inflammatory milieu that's created in MDS could be contributing to [mortality].

What clinical trials have helped define the lower-risk MDS treatment armamentarium?

The past few years have been exciting in MDS. Originally, [research] focused on higher-risk MDS, but breakthroughs and drug approvals have predominantly happened in lower-risk MDS. The 2 major trials that translated to FDA approvals [for patients with lower-risk MDS] investigated luspatercept and imetelstat.

Luspatercept is a fusion trap protein that neutralizes TGF-β ligands and regulates terminal erythroid differentiation. The first approval [of this agent] was based on data from the phase 3 MEDALIST study [NCT02631070] which enrolled patients with RS-positive, or what we now call SF3B1-mutant, MDS after ESA progression. These patients were transfusion dependent and were randomly assigned to receive luspatercept or placebo. [This trial] demonstrated a [38%] transfusion independence rate [of at least 8 weeks] with luspatercept and led to the approval [of the agent] in that setting.

That was followed by a study called COMMANDS, [which moved] luspatercept to the upfront setting. This was the first time we had a randomized clinical trial comparing ESAs with a newer agent [in MDS]. [COMMANDS enrolled] patients with lower-risk, transfusion-dependent MDS—both RS positive and RS negative—and randomly assigned them to receive luspatercept or ESAs. The primary end point was a bit more robust than historical [ones]. [The trial investigated the rate of] patients becoming transfusion independent for [at least] 12 weeks and having an objective [mean] hemoglobin level increase of [at least] 1.5 g/dL [within the first 24 weeks]. Luspatercept outperformed ESAs, particularly in the RS-positive [population, for which] the study was highly enriched.

There were only 48 and 49 RS-negative patients [in the luspatercept and ESA arms], respectively. In the RS-positive [population], we saw a doubling of the response rate and duration of response [with luspatercept vs ESAs]. In the small subset of RS-negative patients, the responses seemed to be similar, but probably more durable. That led to the FDA expanding the label for luspatercept [beyond] post-ESA progression as upfront [therapy]. The National Comprehensive Cancer Network [NCCN] guidelines have been updated to reflect that; in the RS-positive population, it's the preferred choice, and in the RS-negative population, it is an option.

The other pivotal trial has been the IMerge trial. This investigated imetelstat, a telomerase inhibitor. Telomerase activity is increased in MDS cells, as well as in other cancers. The IMerge study enrolled patients after ESA progression and randomly assigned them to receive imetelstat IV infusion once a month or placebo. [This trial] showed that [39.8%] of patients became transfusion independent [for at least 8 weeks], with a median [duration of transfusion independence] of [51.6 weeks]. Hemoglobin levels increased by [a median of] 3.6 g/dL [with imetelstat], so there was robust activity among patients after ESA progression. What's unique in this study was that patients were included if they were heavily transfusion dependent. Historically, we had no agent that was active in that subset of patients with erythropoietin [EPO] levels above 500 mU/mL.

The downside is that imetelstat has a myelosuppressive effect. [It is associated with] grade 3 or 4 thrombocytopenia and neutropenia, which happens in the first cycle or 2 and requires dose holding or adjustment. However, the good news is that [in IMerge], the rates of clinical manifestations, such as febrile neutropenia, were less than 1%, and the rate of bleeding was low.

[The IMerge data] led to the FDA approval of imetelstat in patients post-ESA progression, and the NCCN guidelines also incorporate it [for the treatment of patients with] high transfusion burden. Activity was seen both in RS-positive and RS-negative populations. The NCCN guidelines even position it as a potential option in patients up front if their EPO levels are above 500 mU/mL, they have a high transfusion burden, or even if ESAs and luspatercept don't perform well.

What emerging targets and ongoing research may set investigational MDS agents on the approval path?

We have oral HMAs now. Decitabine plus cedazuridine [Inqovi] has been FDA approved as an oral [combination]. The label is for intermediate- and higher-risk MDS. The oral formulation of azacitidine is being developed as well. Whether [the benefit of oral HMA formulations is] just ease of use for patients or different preferential activity is yet to be seen. There are ongoing studies with oral HMAs. We usually reserve HMAs for patients who have concomitant neutropenia or thrombocytopenia that limit the use of other agents, or for [patients who have progressed on] several lines of therapy.

Elritercept [KER-050] is similar to luspatercept. [Elritercept] is an activin receptor type IIA fusion trap protein, [whereas] luspatercept is an activin receptor type IIB [fusion trap protein]. Elritercept has been tested in phase 1 and 2 trials and showed promising activity both in RS-positive and RS-negative [patients], even [those with] higher transfusion burden. At least preclinically, it [has also shown the potential for] more trilineage responses. It's [under investigation] in the phase 3 RENEW trial [NCT06499285], where it's going to be used after ESA progression in transfusion-dependent patients who are RS positive or RS negative.

It's exciting that we [may] have 2 agents in the same class. We'll see if [one] retains activity after [progression on the other one]. However, elritercept is the next drug in a phase 3 trial that, if positive, will hopefully lead to FDA approval.

We are also evaluating moving from the point of treating patients only when they become transfusion dependent to treating patients with symptomatic anemia. In practice, most of the time we don't wait for patients to become transfusion dependent [before treating them]. The ELEMENT-MDS trial [NCT05949684] is a randomized phase 3 trial that is similar to COMMANDS. In the ELEMENT-MDS study, patients with lower-risk MDS who are not transfusion dependent, are randomly assigned to receive luspatercept or ESAs. The primary end point is transfusion [dependence rate during any continuous 16-week interval]. This is an exciting trial because it's evaluating changing the dogma of waiting for patients to be transfusion dependent [to see whether there is a] benefit with earlier treatment. It will shape the landscape of when to treat and [show] whether luspatercept is a better option than ESAs in that setting.

There are several other promising drugs that are in earlier-phase studies. A lot of efforts are targeting the inflammatory milieu in the bone marrow, because that seems to be an important aspect [of MDS]. There is the NLRP inflammasome complex. There are NLRP inhibitors in clinical trials. There are the IRAK1 and IRAK4 pathways, and there are IRAK4 and IRAK1 dual inhibitors in clinical trials as well.

There are also drugs trying to target the hepcidin and hemojuvelin pathways. DISC-0974 was tested in myelofibrosis and showed promising anemia response, and it's also been [evaluated] in MDS. Those are several promising pathways being explored actively in clinical trials that are not yet in randomized phase 3 trials.

You coauthored a paper on the management of MDS, which was based on the 2025 inaugural Bridging the Gaps: Leukemia, Lymphoma, and Multiple Myeloma Meeting. What controversies and unanswered questions about the treatment of patients with MDS were discussed at the meeting?

There are many controversies [and questions] that need more answers and more clinical trials. One is how we treat patients with high-risk CCUS, because although the drug approvals [we've had for MDS are] not for patients with high-risk CCUS, in practice, we see those patients who are cytopenic and need treatment, and we have [increasingly] convincing evidence that those patients probably have lower-risk MDS.

Another controversy is when to start therapy. Do we wait, as we've done historically, for patients to become transfusion dependent [to align with current drug approval indications], or is there an advantage of earlier treatment for those patients?

Also, how do we sequence therapies? Do we move luspatercept up front in all-comers? Is there still some role for ESAs? What would the responses be if we used luspatercept first then ESAs, or the other way around? Could there be an advantage of combining those therapies? We've had early results showing that sometimes after luspatercept progression, adding an ESA can salvage patients. There is no 100% consensus though. People are moving luspatercept up front, but there's a bit of controversy [about that] in the RS-negative population. People are moving to start treatment earlier, but most of those data are based on responder/nonresponder analyses, so [that answer is] not conclusive.

Regarding higher-risk MDS, [there is also] a lot of controversy [about whether patients] need to have a response before transplant. At what last cutoff can you take patients to transplant? We presented data from the VALIDATE consortium showing that HMA use before transplant did not affect the transplant outcome. The consensus was that if patients did not have a response to HMAs, that lack of response should not preclude them from receiving allogeneic stem cell transplant [allo-SCT].

In the higher-risk setting, the challenge had been how to improve the current standard, because for almost 2 decades we have used HMAs. Those [agents] have led to complete response rates of approximately 15% to 20% and median survival durations of approximately 1 and a half years. Several trials adding agents to HMAs have unfortunately failed to show an advantage, and part of it is probably not just the agents, but the designs of the trials and the heterogeneity of the populations that were included in the trials.

Another part of the discussion is that we have to have more homogeneous patient groups. For example, many of those trials were enriched with patients with TP53-mutant disease, which is a subset of MDS and AML that unfortunately does not do well [with current treatments], and that may have affected the trial outcomes. Some of those studies were not powered enough to detect a stepwise improvement in survival, etc.

Another challenging area is biallelic TP53-mutant MDS, where the 2 copies of p53 are not functional. That's the worst subset of MDS. This group of patients with MDS and AML does poorly [with treatment]. Unfortunately, most trials have not shown improvement in that subset. Even in the AML literature, where venetoclax [Venclexta] had been an advancement and [patients receive] azacitidine [Vidaza] plus venetoclax, there was no survival advantage in patients with p53 mutations.

The major controversy in the [MDS] field is: Do you transplant those patients? Patients with TP53-mutated disease who are going to transplant have an approximately 80% chance of recurrence, so the cure rates [with transplant in this population] are only approximately 20% compared with other subtypes, where [the rates are approximately] 40% to 50%. The transplant [supporters] will say that a 20% cure rate is better than 0%, because in those patients, unfortunately, survival durations are less than 1 year. In younger patients, that's still the go-to [treatment]. The people against transplant will [ask whether an] 80% chance of recurrence is ethically worth putting the patient through a transplant.

I think [this decision is] not that simple; it's complex and depends on many variables. Patients [can make] individual decisions, but [transplant is] a huge area. There are institutions that will immediately transplant every patient with p53 mutations and institutions that will not transplant patients with p53 mutations. [This is] an area of unmet need that needs more data and more consensus.

Another area that has been discussed is what to do at the time of progression on standard therapies, particularly HMAs, in higher-risk patients not going to transplant. What's emerging as an important point is to reassess the disease and look for targeted abnormalities. We always have to advocate to repeat the evaluation of the disease and check the molecular profile. In approximately one-third of patients a target may be identified.

IDH1/IDH2 mutations are not the most common in MDS [and are present in approximately] 5% to 10% of patients. We see the emergence of FLT3 mutations in approximately 5% [of patients and] NPM1 [mutations can also occur]. Collectively, approximately 20% to 30% of patients are [eligible for] targeted therapy.

Now we have an IDH1 inhibitor, ivosidenib [Tibsovo], approved by the FDA for relapsed/refractory [IDH1-mutant] MDS, so that could be an option. Historically, the duration of survival for patients post-HMA progression has been approximately 4 to 6 months. In the subset of patients [with IDH1-mutant disease] who receive an IDH1 inhibitor, [the survival rate increases to approximately] 36%. Those are effective treatments in the subset of patients [with IDH mutations].

How do we move those agents [to earlier lines of therapy] in those small subsets? There have been trials by a French group trying to move IDH1 and IDH2 inhibitors more upfront and more in the lower-risk population. There are even trials investigating them in CCUS. We should shy away from lumping patients with targetable mutations [into 1] group [and thinking that] HMAs are always the standard-of-care therapy for those patients.

What lessons have been learned from previous research investigating immunotherapy in MDS?

Unfortunately, [the phase 1 study (NCT02117219) evaluating durvalumab [Imfinzi] alone or in combination with tremelimumab [Imjudo] and azacitidine in patients with MDS was] terminated, and I don't think we are moving forward [with that research]. We published the results [in Annals of Hematology in March 2025]. A challenging area in MDS has been taking advantage of the immune therapies that have been breakthroughs in solid tumors, lymphomas, and myelomas. We've done numerous studies. Other groups were involved with other agents like nivolumab [Opdivo] and ipilimumab [Yervoy]. Recently, there were big, randomized phase 3 trials investigating sabatolimab [MBG453] and magrolimab. All of those are immune therapies that unfortunately failed to show an advantage [in patients with MDS].

The main issue with those trials is that the immune system in patients with MDS is exhausted; there's clear evidence that these patients' T cells are exhausted. [Immunotherapies are] trying to harness an immune system that's not functional, and recruiting T cells that are not functional is not going to be effective. We know for sure that immune therapy is effective in MDS, because allogeneic transplant is the main treatment. Allo-SCT is an immune therapy, as it replaces the immune system for the patient. We are starting to move into considering off-the-shelf treatments and donor T cells. If we are going to move to CAR T-cell therapies, [we want to consider] off-the-shelf [T-cell therapy] approaches or maybe try to start positioning those patients [for CAR T-cell therapy] after transplant to prevent relapse when the immune system is still intact.

There are other trials investigating new agents. There is a pathway called Clever-1 that seems to be important that relies on macrophages. There are promising early data in leukemia and MDS, and [Clever-1 inhibitors are] going into trials now. However, immune therapies cannot rely on the patient's immune system [in MDS], because it's not intact, and immune deregulation is part of the disease. Therefore, we have to [use] off-the-shelf or allogeneic donor T cells or move [immunotherapies to] after transplant where we've replaced the patient's immune system.

References

1. US FDA approves Bristol Myers Squibb's Reblozyl (luspatercept-aamt) as first-line treatment of anemia in adults with lower-risk myelodysplastic syndromes who may require transfusions. News release. Bristol Myers Squibb. August 28, 2023. Accessed May 13, 2025. https://news.bms.com/news/details/2023/U.S.-FDA-Approves-Bristol-Myers-Squibbs-Reblozyl-luspatercept-aamt-as-First-Line-Treatment-of-Anemia-in-Adults-with-Lower-Risk-Myelodysplastic-Syndromes-MDS-Who-May-Require-Transfusions/default.aspx
2. FDA approves imetelstat for low- to intermediate-1 risk myelodysplastic syndromes with transfusion-dependent anemia. FDA. June 6, 2024. Accessed May 13, 2025. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-imetelstat-low-intermediate-1-risk-myelodysplastic-syndromes-transfusion-dependent