New therapy options for myelodysplastic syndrome—all for all or targeted?

Treatment options for patients with myelodysplastic syndrome (MDS) remain limited. However, a new option became available in April 2024 with luspatercept for transfusion-dependent low-risk (LR) patients. There are also positive data for imetelstat, a first-in-class telomerase inhibitor already approved by the Food and Drug administration (FDA), in LR MDS patients refractory to erythropoietin-stimulating agents (ESAs). In contrast, progress is still limited for high-risk (HR) patients, although there are promising data for patients carrying a IDH1 mutation. The latter represents a targeted therapy, while luspatercept can now be used broadly. Nonetheless, at this year’s OeGHO spring meeting, there was a pro and con session where the arguments for and against uncritical versus widespread use of available therapies were discussed.

The phase III, global, open-label, randomized COMMANDS study, investigated luspatercept vs. ESA in the first-line setting among low risk, transfusion-dependent (defined as 2–6 red blood cell units/8 weeks in the last 8 weeks before randomization of the study) MDS patients [1]. Luspatercept is a recombinant fusion protein that binds transforming growth factor β (TGF-β) superfamily ligands to reduce SMAD2 and SMAD3 signaling, resulting in improved late-stage erythropoiesis. The primary endpoint in the COMMANDS study was red blood cell transfusion independence for ≥ 12 weeks with a concomitant mean hemoglobin increase ≥ 1.5 g/dL. In the intention to treat population, there was a significant advantage with regard to treatment with luspatercept vs. ESA therapy (response 60.4% vs. 34.8%). Based on these data and achievement of the primary endpoint, FDA approval was initially granted in August 2023. Just a few days before the ÖGHO Frühjahrstagung, luspatercept has also been approved by the EMA in the first-line setting. This drug is therefore moving from a specific second-line approval (for ring sideroblast-positive patients after failure of ESA) to the first line with a broad approval for all transfusion-dependent low-risk MDS patients, thus, expanding the range of possible treatment options for our patients in the first-line setting.

The subgroup analysis showed that the response was better with luspatercept, regardless of transfusion requirement, serum Epo level, or SF3B1 mutation status. With regard to ring sideroblast status, however, it was shown that patients with a ring sideroblast-positive status benefited in particular. In comparison, the two therapies were similarly effective in patients with ring sideroblast-negative status. A further detailed analysis showed that there was an imbalance in terms of mutational burden within the ring sideroblast-negative subgroup (26% ≥ 4 gene mutations in the luspatercept arm vs. only 3 patients [6.4%] in the ESA arm) [2]. In addition, mutations associated with a poorer prognosis accumulated within this subgroup (e.g., ASXL1, TET2 DNMT3A ZRSR2). The study stratified according to serum EPO (Erythropoietin) level, ring sideroblast status, and transfusion burden, but not according to molecular genetic aberrations. The latter is certainly an increasingly important aspect that would ultimately be mapped by calculating the Molecular International Prognostic Scoring System (IPSS-M) score. A review of the IPSS‑M score showed a clinical benefit in all risk groups with regard to treatment with luspatercept. Ultimately, this shows how heterogeneous the landscape of MDS is and that not every therapy leads to a benefit for every subgroup.

In clinical practice, the question is now being discussed as to whether luspatercept should actually be used as first-line therapy for everyone. Ultimately, this question remains unanswered at present, as there are many factors that should be taken into account in such a decision. In addition to the difference in time of application and patient preferences (3 weekly luspatercept at a clinic vs. weekly ESA at home/in the general practitioner’s office), cost effectiveness is always listed as an argument in the discussions.

Another limitation is the fact that ESA is approved and used before transfusion dependency. This means that many patients have already been treated with ESA before they become transfusion-dependent and are eligible for luspatercept. In the ELEMENT-MDS study (phase III, randomized), luspatercept is currently being investigated in anemic, not yet transfusion-dependent LR-MDS patients vs. ESA [3]. The results of this study will be interesting with regard to the classification of luspatercept in the therapy algorithm for LR-MDS. This study and also real-world data with the use of luspatercept in the first line will show what role luspatercept will play in the treatment of MDS patients.

In parallel, data is accumulating on another TGF‑β inhibitor KER-050 (elritercept), a modified activin receptor type IIA ligand trap, designed to inhibit select TGF‑β superfamily ligands (activins A, B, GDFs 8, 11), which, in contrast to luspatercept, also binds activin A and, thus, has a broader effect on hematopoiesis [4].

Data from the ongoing phase II KER050-MD-201 study have been repeatedly presented at congresses, with promising results in terms of transfusion independence in low-risk patients and patients with a partially high transfusion load (defined as ≥ 4 red blood cell units/8 weeks). The most recent data are from the EHA conference 2024, where the analysis of n = 87 patients (n = 60 ringsideroblast-positive, n = 27 non-ringsideroblast), who received or completed ≥ 24 weeks (median 42 weeks) of therapy showed an overall response rate (ORR) of 56%. In 41% a transfusion-free status ≥ 8 weeks was achieved; the median duration of response has not yet been reached. Among responders, the therapy also appears to have a positive effect on quality of life [5].

Furthermore, in 06/24, the FDA approved imetelstat, a first-in-class investigational telomerase inhibitor in the second line, in patients who have failed ESA therapy. In the randomized phase III IMerge study (n = 118 imetelstat vs. n = 60 placebo; primary endpoint: transfusion-free interval ≥ 8 weeks), a significant benefit was shown in patients with Revised International Prognostic Scoring System (IPSS-R) low-risk or intermediate‑1 risk MDS [6]. Overall, this was achieved in almost 40% of patients receiving imetelstat therapy versus 15% of patients receiving placebo.

Once again, a ring sideroblast-positive status and a low endogenous serum EPO level were predictive of a better response. In particular, patients with a high transfusion burden (≥ 6 units per 8 weeks) also benefited significantly from treatment with imetelstat. It was also shown that this freedom from transfusion was present in 28% of patients over 24 weeks and even lasted for more than 1 year in 18% of patients. The median duration of response was 51.6 weeks versus 13.3 weeks with placebo therapy.

In contrast to luspatercept, which is generally very well tolerated, where the most common side effects are diarrhea and fatigue, grade 3–4 hematological side effects (in particular thrombocytopenia and neutropenia) must be expected with imetelstat therapy [1, 6]. Imetelstat is currently still in the approval process at the EMA.

A future targeted therapy option for IDH1-mutated MDS patients could be therapy with the oral, targeted, small molecule IDH1 inhibitor ivosidenib. Again, this is already a treatment option approved by the FDA in relapsed or refractory myelodysplastic syndromes (MDS) in adults with a susceptible IDH1 mutation. The approval in October 2023 is based on the phase I study with 18 patients, where monotherapy with ivosidenib resulted in an ORR rate of 83.3%, the CR rate was 38.9% [7]. The majority of patients (77.8%) in this study had an IPSS‑R score > 3 and had received prior therapy with a hypomethylating agent. The median overall survival (OS) was 35.7 months and the probabilities of survival at 1, 3, or 5 years were 86.9, 46.3, and 46.3%, respectively, with a higher probability in the case of a lower IPSS‑M risk category.

At the EHA this year, the French working group presented data from the single-arm phase II Idiome study. In this study, ivosidenib was investigated for patients with such a mutation in patients with MDS in three different groups: high-risk patients who failed azacitidine (cohort A, n = 22), high risk as first-line treatment (cohort B, n = 23) and low-risk patients who failed therapy with EPO (cohort C, n = 3). The primary endpoint of the study was overall hematological response within 3 and 6 months (cohorts A and B) and safety for cohort C, respectively [8].

In the refractory cohort A, a response (ORR) was achieved in 64% (n = 14) of patients after three cycles, the median duration of response was 4.8 months. In the treatment-naïve cohort B, the response rates were even higher, at 78% (n = 18) with a CR rate of 61% (n = 11). The addition of azacitidine after three cycles could be added if there was a lack of response. Neither the median OS nor the median duration of response was reached in this cohort with a median follow-up time of 25.2 months [8].

Apart from these promising data, it is still ‘please wait’ in the HR setting. Neither the combination of azacitidine with sabatolimab, a TIM3 antibody, nor magrolimab, a CD47 antibody, could lead to a significant overall survival benefit compared to azacitidine monotherapy [9, 10]. The phase III results of the VERONA trial are eagerly awaited, where the combination of azacitidine and venetoclax vs. azacitidine monotherapy is investigated [11].