Should we move past erythropoietin-stimulating agent monotherapy in lower-risk myelodysplastic syndromes?

Myelodysplatic syndromes (MDS) are clonal stem cell disorders characterized by ineffective hematopoiesis with progressive bone marrow failure resulting in cytopenias. Anemia is the most prevalent cytopenia and 25-30% of patients present with transfusion dependence at diagnosis.¹ Erythropoietin-stimulating agents (ESA) are often the first-line therapy in lower-risk disease to reduce symptoms from anemia and eliminate or delay transfusion dependence.^1,2 ESA response rates are substantially lower (<25%) in patients with significant transfusion dependence and endogenous serum erythropoietin (EPO) levels above 200 U/L.³ Most guidelines state that ESA should not be used when EPO levels exceed 500 U/L.⁴ Thus there is a need to improve upon current treatment strategies for this group of patients.

In this issue of Haematologica, Mei et al., report the results of their multicenter, single-arm, prospective trial evaluating recombinant human EPO (rhEPO) plus all-trans retinoic acid (ATRA) and testosterone undecanoate for the treatment of anemia and transfusion dependence in patients with lower-risk MDS.⁵ Biological rationale exists for combining these three agents. ATRA, a standard component of therapy for acute promyelocytic leukemia, promotes hematopoietic cell differentiation and enhances erythroid colony formation.⁶ As monotherapy, ATRA has a suboptimal improvement in anemia for patients with MDS with better results observed in combination with rhEPO.^6,7 Androgens stimulate erythropoiesis by increasing the sensitivity of erythroid progenitor cells to EPO⁸ and when also combined with ATRA, may improve anemia in patients with MDS.⁹ Encouraged by these findings, the authors hypothesized that this triple combination therapy might improve erythroid response rates.

Eligible patients received rhEPO 10,000 IU/day, ATRA 25 mg/ m²/day and oral testosterone undecanoate 80 mg twice daily for 12 weeks with the primary endpoint of hematological improvement-erythroid (HI-E) using 2006 International Working Group criteria.¹⁰ If response was demonstrated at 12 weeks, treatment was continued until progression or unacceptable toxicity. In total, 52 patients with primarily untreated lower-risk disease received treatment and were eligible for analysis with a median age of 65 years and a median hemoglobin (Hb) of 62 g/L (range, 40-91 g/L). At baseline, 56% had a serum EPO level of greater than 500 U/L, 40% harbored mutations in SF3B1 and 52% were transfusion dependent. With a median follow-up of only 20 weeks, 32 (62%) achieved HI-E and notably, 17 of 29 patients (59%) with a baseline serum EPO level of >500 U/L also responded. The post treatment median Hb in responders was 95 g/L (range, 68-152 g/L) compared with 73 g/L in non-responders (range, 46-91 g/L). Of the transfusion-dependent patients, 48% achieved and maintained transfusion independence for ≥12 weeks. Lower response rates were demonstrated in patients with ASXL1 mutations (33% mutated vs. 70% wild-type). Importantly, this regimen was well-tolerated and had a toxicity profile consistent with the individual agents with four reported grade 3 treatment-related adverse events (infection in 3 patients, deep vein thrombosis in 1 patient) although rates of virilization in women were surprisingly absent.

This study is notable for several reasons. First, it combines three previously uncombined agents to target anemia in MDS. Additionally, response rates did not differ between SF3B1-mutated (71%) and wild-type (55%) disease in contrast with 70% (SF3B1-mutated) and 45% (SF3B1-wild-type) achieving transfusion independence with luspatercept in the COMMANDS trial.¹¹ Lastly, high response rates were achieved in patients with very elevated EPO levels (excluded from most trials) and in transfusion-dependent patients. However, transfusion density was not provided, and their definition of transfusion dependence was only a minimum of one packed red blood cell unit per 8 weeks compared with other trials.^11-13

While interesting and certainly promising, this triplet therapy is not yet ready for prime time without comparative data against monotherapy agents like luspatercept and imetelstat in non-Asian patients. For one, the EPO doses used (70,000 IU per week) are much higher than typical starting doses and may play the heavy-hitting roles in these modestly defined responses (8 weeks of transfusion independence). Second, response durability was not provided - an important endpoint in assessing efficacy. Third, fatigue was an adverse event in 38.4%, when amelioration of fatigue is the goal of most erythroid-stimulating approaches. Fourth, external validity remains to be proven since the median baseline Hb levels were more than 10 g/L lower compared with the pivotal luspatercept and imetelstat trials.^11-13 Lastly, the highest Hb levels achieved with HI-E were very modest, possibly without impact on quality of life, given the importance of a Hb threshold of 100 g/L.¹⁴

Table 1.Recent trials for the treatment of anemia in patients with myelodysplastic syndromes.

In summary, the prospective single-arm trial presented by Mei and colleagues builds upon the current paradigm of ESA monotherapy for the treatment of anemia in patients with lower-risk MDS, in particular, for patients with baseline serum EPO levels exceeding 500 U/L. It is worthy of further investigation and comparative analysis with luspatercept, imetelstat (Table 1), hypomethylating agents and others.

Footnotes

• Received June 2, 2025
• Accepted June 17, 2025

Correspondence

L. Mozessohn lee.mozessohn@sunnybrook.ca

Disclosures

LM served on an advisory board for AbbVie. RB has received research funding from BMS and Taiho; has participated on study steering committees for BMS and Takeka; and has received advisory board honoraria from BMS and Taiho.

Contributions

LM wrote the editorial and RB provided critical edits.