Prime salvage treatment for pediatric acute myeloid leukemia

In this issue of Haematologica, Fan et al. report relatively favorable outcomes of children with refractory or relapsed acute myeloid leukemia (R/R AML) using priming with decitabine followed by the combination of low-dose subcutaneous cytarabine, low-dose idarubicin and granulocyte colony-stimulating factor (G-CSF; DP-IAG).¹ In 101 children, the majority of whom had refractory disease, 73% achieved a complete remission (CR) with or without hematologic count recovery, and 3-year probability of overall survival (pOS) was 61%. This does indeed compare favorably with the largest and only randomized study reported so far in pediatric R/R AML, in which the best arm (fludarabine and cytarabine [FLA] plus G-CSF and liposomal daunorubicin) resulted in a CR rate of 69% and a 4-year pOS of 40%.² There are no prospective, large studies that report equally good outcomes overall as those shown by Fan et al., as reviewed recently.³ However, it is too early to conclude that this regimen using priming with decitabine, chemosensitization with G-CSF, and low-dose idarubicin and cytarabine is the new standard-of-care in pediatric R/R AML; that would require a larger, randomized study. Also, this is knowing that 30 out of the 101 patients received targeted therapy (sorafenib or dasatinib) in addition to DP-IAG.

By and large, most pediatric AML experts would agree that FLA plus or minus an anthracycline or gemtuzumab ozogamicin is the standard-of-care for salvage or reinduction chemotherapy in case of refractory or relapsed disease, respectively.^4,5 However, since neither remission rates nor survival rates with that approach are satisfactory, we do need clinical trials exploring other regimens. Randomized trials are possible through international collaboration, and using measurable residual disease as early surrogate endpoint, thereby reducing the patient numbers needed to prove that a new regimen is better.⁶ Certainly, the DP-IAG regimen should be one of the regimens to be investigated. Epigenetic priming resulting in lower genome-wide methylation is an interesting concept, that was also tested in the US multicenter AML16 trial. At the 2025 ASH meeting, data were presented on 200 children with newly diagnosed AML, showing the safety of priming with either decitabine or azacytidine, and suggesting better outcomes than previous trials, albeit in historical comparisons.⁷ Interestingly, decitabine appeared to be superior to azacytidine.

Another approach that is promising and potentially effective in the majority of patients is the combination of chemotherapy with a bcl2-inhibitor as chemosensitizer by enabling drug-induced apoptosis.⁵ Especially venetoclax is frequently used in children and well tolerated, while the majority of children with AML have leukemic cells expressing bcl2. Therefore, we also need a larger, randomized study on any chemotherapy regimen plus or minus venetoclax. In fact, such a trial is ongoing in R/R AML, the ICC-101/ APAL2020D trial (clinicaltrials.gov identifier NCT05183035). In addition to the novel, broadly applicable regimens mentioned above, more and more targeted treatment options have become available.⁵ Obvious examples already being or soon to be studied in pediatric AML include flt3-inhibitors such as sorafenib, gilteritinib and quizartinib, and menin-inhibitors such as revumenib and bleximenib. Flt3-inhibitors have already demonstrated improved survival in randomized trials in adults with AML when combined with chemotherapy and as continuation treatment post-transplant. The Children’s Oncology Group (COG) reported on the safety and efficacy of sorafenib in children with AML and an FLT3-ITD, and results show improved outcome as compared with patients not receiving sorafenib.⁸ The COG is currently investigating gilteritinib (clinicaltrials.gov identifier NCT04293562), while the NOPHO-DB-SHIP consortium is enrolling children on the Quizartinib trial (EU CT 2023-505000-27-01), linked to the Master of CHIP-AML22 (EU CT 2023-504999-25-00).

There is a large number of other promising new agents for children with R/R AML, such as antibody-drug conjugates, bispecific antibodies, IDH1/2 inhibitors, inhibitors of MDM2 and MDMX proteins, and inhibitors of nuclear export, such as selinexor. Finally, there is the development of cellular therapies, already successful in acute lymphoblastic leukemia.⁵

It is very likely that all these innovations in the treatment of R/R AML in children and adolescents will result in higher survival rates. Apart from survival, short- and long-term quality of life will also become an increasingly important endpoint. Hopefully, more effective and less toxic novel treatments will mean that less use will be made of the old-fashioned, toxic chemotherapy, ideally, even making it possible to omit allogeneic stem cell transplantation (alloSCT), at least in some patients. After all, alloSCT is associated with significant morbidity and late effects, and even mortality; perhaps very sensitive techniques measuring the level of molecular residual disease pre- and post-SCT will be helpful in this context.⁹ Meanwhile, alloSCT is standard-of-care for R/R AML, even if that could mean a second transplant, and outcomes in that setting are improving.¹⁰ Ultimately, we should avoid refractory and relapsed disease, and, therefore, it is important to quickly move novel agents and novel treatment modalities through to clinical trials in newly diagnosed AML. In trying to further improve outcomes in pediatric AML, we should not forget that the vast majority of children with AML are diagnosed in low- and middle-income countries. Efforts to develop less toxic and affordable regimens that can be applied in resource-limited settings is, therefore, very important as well.

Footnotes

• Received April 9, 2026
• Accepted June 5, 2026

Correspondence

G.J.L. Kaspers gjl.kaspers@amsterdamumc.nl

Disclosures

No conflicts of interest to disclose.