Treatment Approaches in Pediatric Acute Myeloid Leukemia: A Narrative Review

1. Introduction

Acute leukemias (AL) are the most common form of malignancies seen in children. Majority (app. 80%) of the diagnosed cases of acute leukemias are acute lymphoblastic leukemia (ALL), whereas small proportion (app. 20%) constitutes the acute myeloid leukemia (AML).1 Though a rare disease of childhood, acute myeloid leukemia (AML) is the second most common hematological malignancy in children after acute lymphoblastic leukemia (ALL).1 The incidence of pediatric AML is about 7 per million cases annually.2 The prognosis of pediatric AML has been found to be poorer with high rate of relapse.1

AML originates from the stem cells precursors with malignant potentials, which typically matures into myeloid series cells (red cells, white cells, and the platelets), resulting in excessive production of malignant myeloid stem cells.3 Environmental predisposing factors accounts for only small percentage, whereas some predisposing symptoms and germline mutations accounts for less than 10% of the pediatric AML.4 Infants less than 2 years of age and the adolescents have the highest incidence among the children. 2 'MLL rearranged leukemia' is found in higher frequency among the infants, whereas the frequency of 'core binding leukemia (CBL)' and mutations associated leukemias increases with age (usually among the adolescents).4

Genetic and molecular studies conducted for better understanding of AML have led to a better grasp of patient risk, categorization and subsequently to a better treatment modalities and protocols.1

Although first time described in 1900, a formal classification of AML was given in 1976 by French American British (FAB) classification, which described 6 subtypes of AML based on morphology and cytochemistry.5 WHO classification given in 2001 incorporated genetic basis for classification and has been continuously revised since then, the latest revision being on 2022.6 The latest 2022 guidelines established by the European LeukemiaNet (ELN) are now considered the standard for leukemia classification.

FAB classification classifies AML into eight subtypes: AFB M0 to M7 as depicted below:

• M0: Undifferentiated AML
• M1: AML with minimal maturation
• M2: AML with maturation
• M3: APL
• M4: Acute myelomonocytic leukemia
• M5: Acute monocytic leukemia
• M6: Acute erythroid leukemia
• M7: Acute megakaryocytic leukemia

As a result of rapid advancement in novel treatment approaches, there has been improved results in management of pediatric AML over past three decades. Currently complete remission rate in treatment of pediatric AML has reached to 90%, whereas the overall 10 year survival has crossed 60%.7 Further advances in supportive care and the allogenic hematopoietic cell transplantation (allo-HCT) has been providing great hope in the management of pediatric AML.

How to rapidly identify the targetable lesions and incorporate the targeted agents, how to assess the response to therapy, how to balance the therapeutic effects of some agents with their potential late effects, and how to decide the use of allo-HCT are some important queries to deal with, by a pediatric oncologist.8

2. Risk Stratification

Treatment for childhood AML should be risk-stratified based on biological factors. This approach helps prevent overtreatment in patients with a good prognosis and enhances outcomes for those with a poor prognosis.3

Despite advancements in treatment, the prognosis for pediatric AML remains variable. This aims to summarize the current understanding of risk stratification in pediatric AML, focusing on cytogenetic, molecular, and clinical factors.

2.1. Cytogenetic Risk Stratification

Cytogenetic analysis remains a cornerstone of risk stratification in pediatric AML. The presence of specific chromosomal abnormalities can significantly impact prognosis:9

• Favorable Risk:
• o t(8;21)
• o inv(16)
• o t(15;17)
• o Core binding factor (CBF) leukemia (e.g., CBF-β/ITD, CBF-β/MYH11)
• o NPM1 mutations without FLT3-ITD
• Intermediate Risk:
• o -Y
• o +8
• o Monosomy 7
• o Other complex karyotypes
• Unfavorable Risk:
• o Complex karyotypes with multiple chromosomal abnormalities
• o FLT3-ITD without NPM1 mutations

2.2. Molecular Risk Stratification

Molecular markers have emerged as important prognostic factors in pediatric AML:

• FLT3-ITD: Internal tandem duplications in the FLT3 gene are associated with a higher risk of relapse and poorer outcomes.10 FLT3-ITD lacks stability for measurable residual disease (MRD) tracking.
• NPM1 Mutations: Mutations in the NPM1 gene are generally associated with a favorable prognosis, especially when combined with a normal karyotype. 11 However, NPM1 mutation is less frequent in pediatric AML as compared to adults.
• CEBPA Mutations: Mutations in the CEBPA gene can have a favorable or intermediate prognosis depending on the type of mutation. 12-13
• IDH1/2 Mutations: Mutations in the IDH1/2 genes are associated with a higher risk of relapse and poorer outcomes. 14

Emerging Molecular Markers: Expanding molecular profiling is improving risk stratification and targeted therapy approaches in AML. For instance, ERG and BAALC gene expression are associated with adverse prognosis in pediatric AML.

2.3. Clinical Risk Stratification

Clinical factors can also influence prognosis in pediatric AML:

• Age: Younger children generally have a better prognosis than older children. 15
• Presenting Features: The presence of symptoms at diagnosis, such as bleeding or sepsis, can be associated with a higher risk of treatment-related mortality. 16
• White Blood Cell Count: A higher white blood cell count at diagnosis can be associated with a higher risk of relapse. 16

2.4. Risk-Based Treatment Approaches

Treatment strategies for pediatric AML are tailored based on risk stratification. Patients with favorable risk factors often receive less intensive treatment regimens, while those with unfavorable risk factors may require more aggressive approaches, including hematopoietic stem cell transplantation (HSCT).17-18

Risk stratification in pediatric AML is a dynamic process that incorporates cytogenetic, molecular, and clinical factors. By accurately assessing risk, healthcare providers can tailor treatment plans to optimize outcomes for children with this disease. Ongoing research is focused on identifying additional prognostic markers and developing novel therapeutic approaches.

3. Standard Chemotherapy Regimens

The standard chemotherapy treatment of AML involves two phases: Induction and the consolidation phase. 19

3.1. Induction Phase

Induction phase is the initial phase of chemotherapy, which is short, and intensive, usually lasting about a week. The main aim of induction therapy is to clear the blast cells in the blood and to reduce the number of blast cells in the bone marrow to normal. Induction therapy depicts the standard of care for all patients with AML, and there are different factors that govern the decisions regarding the selection of induction chemotherapy.19

The current treatment protocols of AML includes the use of high dose of cytarabine and anthracycline in order to achieve complete remission. In this protocol, cytarabine is infused continuously for 7 days, and once daily injection of anthracycline for 3 days. This regimen is referred to as '7+3 regimen'. Cytarabine and anthracycline are used in combination with etoposide or thioguanine.20 The medical research council's 10th AML trial showed comparison between the combinations of DAT (daunorubicin, cytarabine and thioguanine) and ADE (cytarabine, daunorubicin, and etoposide) had no significant differences in complete remission rate. The ten year survival, disease free survival, and the event free survival were found to be similar with those two regimens.20-22 The dose of 90 mg/m²/day of daunorubicin has been associated with improved overall survival.23 This regimen is used for younger patients (age 70 or less), those with ECOG performance status scale <2, and those with denovo AML without poor risk characteristics.19

In patients with poor risk characteristics, secondary AML or therapy related AML, the preferred regimen is FLAG regimen. This regimen combines fludarabine, high dose of ara-C, and G-csf.24

In elderly patients typically of age 70 or older, and deemed fit, most commonly used regimen is Azacitidine or decitabine and venetoclax.25

In cases of suspected APL, treatment with all-trans retinoic acid (ATRA) should be initiated.

How to Assess the Response?

In young and fit individuals who are undergoing induction regimen (7+3, or FLAG), a bone marrow biopsy is usually done after the induction therapy. It is done around the time of peripheral count recovery, particularly when the absolute neutrophil count exceeds 1000/microL and platelet count exceeds 100K/microL with no blasts present. Remission is said to be complete when the marrow shows no morphological evidences of leukemia, aspirates showing less than 5% blast cells, and the differential counts similar to that of the peripheral blood count.19

There is a debate regarding when to perform the post induction bone marrow biopsy. Some suggest to perform it after the full count recovery which is approximately 28 days, whereas some advocate for performing it at 14th day in order to ensure that residual leukemia is not present and bone marrow is chemoablated.19

3.2. Consolidation Therapy

Minimal residual disease may still persist even after the optimum induction therapy such that there is necessity of the consolidation therapy to mitigate the risk of relapse by eliminating the residual disease. Consolidation therapy is started with high dose of cytarabine, also called HiDAC. Those patients who received FLAG regimen during induction ought to receive additional cycles of same regimen during the consolidation phase also. 26

Limitations of Current Chemotherapy: The intensive treatment with standard chemotherapy like cytarabine-based regimen in AML has high treatment-related mortality and long-term toxicities in survivors. This limits the further intensification of therapy.

4. Hematopoietic Stem Cell Transplantation

Hematopoietic stem cell transplantation (HSCT) plays a critical role in the treatment of high-risk pediatric AML. This approach replaces diseased or dysfunctional bone marrow with healthy hematopoietic stem cells (HSCs) from a suitable donor.27 The success of HSCT relies heavily on careful donor selection, human leukocyte antigen (HLA) matching, and appropriate conditioning regimens. 28

4.1. Indications for HSCT in Pediatric AML

HSCT is primarily indicated for children with AML who have high-risk genetic features, fail to achieve remission after induction therapy, or show evidence of minimal residual disease (MRD) following initial treatment.29 The procedure is used both as a rescue therapy after high-dose chemotherapy and total body irradiation (TBI) and for its graft-versus-leukemia (GVL) effect, where donor immune cells help eliminate residual leukemia cells. 30

For pediatric AML, early HLA typing is crucial to identify a suitable donor in case HSCT is required. 31 Patients with an HLA-matched sibling donor (MSD) generally have the best outcomes, while matched unrelated donor (MUD) transplants or haploidentical transplants may be considered when a related donor is unavailable.32

4.2. HLA Typing and Donor Selection

Advancements in next-generation sequencing (NGS) have improved the accuracy and speed of HLA typing, allowing for better donor-recipient matching.33 For MUD transplants, a 9/10 or 10/10 HLA match is preferred to minimize graft-versus-host disease (GVHD) and improve overall survival. 34

Donor selection is based on HLA compatibility, but additional factors, such as donor age, sex, parity, ABO blood type, and cytomegalovirus status can also influence outcomes. 35 Haploidentical transplants, often using a parent as the donor, have become a viable alternative with T-cell depletion techniques reducing the risk of GVHD. 36 In cases of T-cell depleted HSCT, maternal donors may be preferred due to fetal microchimerism, while in non-T-cell depleted settings, paternal donors might offer better outcomes. 37

4.3. Conditioning Regimens

Before transplantation, patients undergo a preparative/conditioning regimen designed to eradicate leukemic cells, suppress the immune system, and create space for donor HSCs to engraft.38-39 The standard myeloablative conditioning (MAC) regimen for pediatric AML includes busulfan and cyclophosphamide, which has shown high efficacy in preventing relapse.40 Recent studies suggest that treosulfan may be a less toxic alternative while maintaining myeloablative effects. 41

TBI-based conditioning is generally avoided in younger children due to its long-term toxicities, including growth impairment, endocrine dysfunction, and secondary malignancies. 42 Reduced-intensity conditioning (RIC) regimens are considered for patients with comorbidities or those requiring a second HSCT after relapse. 43

4.4. HSCT Procedure and Stem Cell Sources

HSCT can be performed using different stem cell sources:

• Bone marrow – Collected from the iliac crest under general anesthesia, typically preferred for pediatric transplants due to a lower incidence of chronic GVHD. 44-45
• Peripheral blood – Mobilized using granulocyte colony-stimulating factor (G-CSF) and collected via apheresis, though it carries a higher GVHD risk. 45-46
• Umbilical cord blood (UCB) – Processed and cryopreserved until transplantation; while it offers rapid availability and lower GVHD risk, engraftment may be slower. 45,47

4.5. Complications and Long-Term Outcomes

Despite advances in HSCT, complications remain a significant challenge.48 Complications of HSCT can primarily be categorized into acute and chronic complications.

i. Acute complications (first 90 days post-transplant) include neutropenia, anemia, mucositis, and infections, with bacterial and fungal sepsis being major concerns. 49 Hepatic veno-occlusive disease (VOD) is a well-known risk, particularly in patients receiving busulfan-based regimens.50 GVHD, a major cause of morbidity and mortality, is more common in MUD or haploidentical transplants and requires immunosuppressive prophylaxis. 51

ii. Chronic complications include endocrine dysfunctions such as growth hormone deficiency and hypothyroidism, pulmonary fibrosis, chronic GVHD, and secondary malignancies. 52 Patients receiving TBI are at higher risk for late effects, including cognitive impairment and cataracts.53-54 Close long-term follow-up is essential to monitor and manage late complications.55

HSCT remains a cornerstone in the treatment of high-risk pediatric AML, offering the potential for long-term remission. Advances in donor selection, HLA typing, conditioning regimens, and post-transplant management have significantly improved outcomes. However, challenges such as GVHD, transplant-related toxicity, and long-term complications persist. Future research focusing on optimizing conditioning regimens, improving GVHD prophylaxis, and enhancing immune reconstitution may further refine HSCT strategies for pediatric AML.

5. Future Directions and Ongoing Research

5.1. Minimal Residual Disease Monitoring in AML

In majority of AML patients, intensive chemotherapy with or without immune-therapeutics leads to a cytomorphological complete remission. However, complete remission does not always necessarily mean the complete cure of the disease. Overt disease recurrence may even occur, if the residual AML cells persist below the limit of cytomorphological detection. Thus, if the leukemia cells are not eradicated completely, or reduced to a level that may self-extinguish or that can be surveilled and removed by the immune system, there is a high chance AML will relapse. 56 Each relapse of AML comes with progressively decreased probability of long-term survival. 57 Thus, complete eradication of all the residual AML cells is required for the complete cure of AML. Therefore, identification of these residual leukemic cells, termed measurable residual disease (MRD), provides a strong prognosis in terms of the clinical outcomes and may have therapeutic implications in the management of AML as well. 57

MRD refers to neoplastic cells that cannot be detected by standard cytomorphological analysis. 58 Complete remission (CR) in AML, as defined by the European LeukemiaNet (ELN) guidelines, is a morphologic leukemia-free assessment of bone marrow (BM) — i.e., <1×10⁹/L) and platelets (>100×10⁹/L) without exogenous growth factor support. 57 In 2017 the European LeukemiaNet (ELN) introduced a new category of MRD-negative complete remission in their 2017 guidelines update. They suggested that patients who test MRD negative have better outcomes than those who test positive after receipt of the same therapy. 56 According to a meta-analysis involving 81 trials and over 11000 patients, there is strong association between MRD negativity and superior disease-free survival. 57

MRD as a Treatment Stratification Tool: MRD is increasingly being used as a treatment stratification tool rather than just a monitoring and disease prognosis tool.

There are various methods of MRD detection which include: karyotyping, FISH, RT-qPCR, multiparameter flow cytometry (MFC), next generation sequencing (NGS). 57 MFC is a powerful technology that is routinely used for diagnosing and MRD monitoring because of its high sensitivity, rapid results, cost-effectiveness, and easy availability. 59 It also has wider applicability (>90% of AML) and rapid turnaround time (TAT). 59 The sensitivity of MFC can reach beyond 0.01% in AML cases with definitive immunophenotypic aberrancies and in identifying cells with immunophenotype of leukemic stem cells (LSC). In addition to tracking the original clones of residual disease, MFC can also identify new clones based on different-from-normal approaches (DfN), which can be missed using molecular techniques based on targeted NGS or qPCR. 59 It also allows easy assessment of hemodilution and thus, the quality of marrow being assessed for MRD detection. The major limitation of this technique is that it is observer-dependent and needs expertise with adequate experience and a standardized approach. 59 Furthermore, the variability in the techniques used such as MFC and molecular methods like FISH, RT-PCR, NGS in MRD monitoring may affect reproducibility. MRD interpretation remains complex, as cut-off values and clinical application still require validation and may impact risk stratification decisions. It also cannot detect mature differentiating myeloid cells with genomic aberrancies like BCR:ABL1 in the chronic phase or mature myeloid cells or monocytes with NPM1 mutations or PML:RARA in case of acute promyelocytic leukemia. 59

Other very powerful technique is NGS which can assess numerous target genes at once. They are widely applicable and can detect mutations in any sequenced portion of a gene. The major disadvantages of this technology are that they have slow turnaround time, are expensive, and require high level of expertise for data interpretation. 57

Other techniques like karyotyping, FISH, RT-qPCR etc. can also be used for MRD detection, but they are not widely applicable and have slow turnaround time than MFC and NGS. 57

5.2. Immunotherapy in AML

The markers like CD33, CD123, CLL1, TIM3, and CD244 are ubiquitously expressed on AML bulk cells and, often, in leukemic stem cells (LSCs), both at the time of diagnosis and the relapse, despite the genetic characteristics and the evolution of leukemic clones. So, they have been considered as ideal targets for AML immunotherapy. 60 Immunotherapy for AML includes monoclonal antibodies which include naked antibodies against AML surface antigens such as CD33 (e.g., lintuzumab) or CD38 (e.g., daratumumab), antibodies conjugated to toxins in various anti-CD33 (gemtuzumab ozogamicin, SGN33A, IMGN779) and anti-CD123 (IMGN632, SL-401, SGN-CD123A) formulations, antibodies conjugated to radioactive particles such as Iodine or Actinium-labeled anti-CD33 or anti-CD45 antibodies, and multiple bispecific antibodies. 60

5.2.1. CD33 Target

The cell surface antigen CD33 is present in more than 80% of patients with AML but is absent from hematopoietic stem cells, rendering it an ideal immunoconjugate target. 61 Gemtuzumab Ozogamicin (GO) is a humanized anti-CD33 monoclonal antibody conjugated with calicheamicin, which was granted approval by the United States (US) Food and Drug Administration (FDA) in 2000. 60 GO is shown to be effective against AML with high CD33 expression according to some pre-clinical studies. But in certain circumstances, GO has been associated with hepatic VOD, which is characterized by jaundice, painful hepatomegaly and/or fluid retention, including ascites and encephalopathy as late manifestations. 60 So, the use of GO was forced to be withdrawn in 2010 owing to this adverse effect, even though it was approved for relapsed AML patients aged over 60 years. 62 However, with the addition of GO to induction chemotherapy, there has been significant survival benefits for patients with favorable cytogenetics, no benefit for patients with high-risk disease, and a trend towards benefit in intermediate-risk patients. It also showed improved survival in younger AML patients. 62

Vadastuximab talirine (SGN-CD33A), a novel anti-CD33 ADC combined with pyrrolobenzodiazepine dimer (PBD), has superior activity in comparison to GO. It promotes significant AML cell death owing to its target over the HMAs. HMAs upregulate CD33 expression. Thus, the drug shows increased DNA incorporation and enhanced cytotoxicity. 62 It has shown results in relapsed/refractory (r/r) adult patients as monotherapy and in combination with a hypomethylating agent. 61

Immune-conjugated targeting antibody, Lintuzumab (anti-CD33 antibody), has been used as the backbone for the development of immune-conjugated antibodies 6060.

5.2.2. Anti-CLL-1 ADC

CLT030 is a humanized monoclonal ADC targeting CLL-1, linked covalently to a highly potent DNA binding payload. It is superior to anti-CD33 ADC by the fact that it does not affect normal hematopoietic cells, thus preventing prolonged myelosuppression. Xenograft models of AML showed robust tumor-growth inhibition with the use of CLT030, thus opening the path for future clinical trials. 62

5.2.3. Targeting Mesothelin

Mesothelin is a cell-surface tumor differentiation antigen expressed on mesothelial cells of serosal lining, associated with malignant transformation, cellular proliferation, and tumor aggressiveness in a variety of solid tumors, including lung, pancreatic, and ovarian origin. Mesothelin was also shown to be expressed in pediatric AML cells. 61 Studies suggest mesothelin as one of the most highly expressed genes in about 30% of childhood AML cases, thus showing a promising future in clinical trials. 61

5.2.4. CD123 Target

CD123 (the interleukin-3 IL3IL3 receptor alpha) was the first antigen that was identified to be specific to LSCs. CD123 is highly expressed on myeloid leukemic blasts in a majority of patients in addition to its presence in HSCs as well as more differentiated myeloid and lymphoid cells. Overall, anti-CD123 immunotherapies have shown limited efficacy and unfavorable safety profiles due to the off-leukemia effects of CD123 targeting. 63 In efforts to specifically kill AML LSCs, a drug was developed named Tagraxofusp, which is an IL3/diphtheria toxin fusion protein that is FDA-approved for the treatment of blastic plasmacytoid dendritic cell neoplasm (BPDCN), a rare malignancy that has high levels of CD123 expression. In addition to normal HSCs, Tagraxofusp has shown potent cytotoxicity against AML LSCs in preclinical studies. 63 According to in-vitro studies, Tagraxofusp induced potent cytotoxic activity against CD123-positive AMLs and myelodysplastic syndrome blast cells. But they were associated with adverse events such as fever, hypoalbuminemia, transaminitis, hypotension and hypocalcemia, and low clinical responses. 60

5.2.5. Checkpoint Inhibitors

Immune checkpoints are key regulators to prevent auto-immune activity that when stimulated dampen the immune response to an immunologic stimulus. Two of the most commonly applied checkpoint inhibitors target PD1-PD1-L and the CTLA4-CD80/86 axis. CTLA-4 is expressed on the surface of activated T cells and binds to CD80/86 on dendritic cells (DCs), leading to deactivation of the T cell. PD-1 is expressed on chronically activated T cells, B cells, DCs, and macrophages. Under physiological circumstances, PD-1 signaling limits the inflammatory immune response to prevent autoimmunity. 64 Clinical trials have shown checkpoint inhibitors to be of significant use in adult patients with AML with good response and survival rate. However, the effectiveness of checkpoint inhibitors in pediatric AML is still of concern. 63

5.2.6. CD47 Target

CD47 is another immune checkpoint antigen which is responsible for AML to evade the immune system. It is expressed on both AML LSCs and normal HSCs. The CD47 antigen is also known as the 'don't eat me' signal that mediates the escape of AML LSCs from phagocytosis by macrophages. Thus, blocking this signal can induce an immune response resulting in the killing of leukemic cells. Magrolimab is an anti-CD47 antibody that induces tumor phagocytosis and eliminates LSCs in AML preclinical models by blocking the CD47 antigen. It has also shown promising results in early-phase clinical trials in adults with newly diagnosed AML when combined with azacitidine via increased CD47 expression. 63

Combining Targeted Agents: Combining targeted agents (e.g., bortezomib, panobinostat, navitoclax) with chemotherapy may enhance treatment efficacy and overcome resistance.

New Therapeutic Classes: With the view of expansion of treatment options for resistant disease, new therapeutic classes are under investigation, such as HSP90 inhibitors (elesclomol) and apoptosis-modulating agents.

Pharmacogenomics and Functional Drug Response Testing: Pharmacogenomics refers to the study of how genetic variation influences drug response in pediatric AML. Variations in drug-metabolizing enzymes can significantly affect chemotherapy toxicity and efficacy. Glutathione S-transferase theta 1 (GSTT1) is a single gene polymorphism that has been studied, the lack of which is associated with increased treatment-related toxicity and poorer overall survival outcomes. Ethnic and population-based genetic differences may contribute to variability in treatment response. Thus, incorporating pharmacogenomic profiling could improve individualized therapy and reduce adverse effects.

It is documented that traditional pharmacogenomics like GSTT1 focuses on static genetic variations, but newer approaches include functional drug response testing. Functional drug response testing such as ex vivo drug sensitivity profiling of patient-derived leukemic cells allows direct measurement of chemotherapy and targeted drug response. This approach helps identify patient-specific therapeutic vulnerabilities even when genomic markers are unclear. Therefore, functional profiling complements genomics by capturing non-genomic mechanisms of drug resistance.

Personalized Treatment Approaches: Drug sensitivity profiling can be used for personalized treatment approaches which identify effective therapies for the individual patient. It also helps to guide treatment in relapsed and refractory AML. This approach allows selection of therapies beyond standard protocols.

6. Discussion

Major Challenges in Clinical Trials: A major challenge in clinical trials of pediatric AML is that AML is a rare disease, limiting patient recruitment for large-scale trials. This necessitates the need for international collaboration and study groups, as global collaboration enhances data sharing and treatment standardization.

Treatment-Related Mortality and Toxicities: The intensive treatment with standard chemotherapy like cytarabine-based regimens in AML has high treatment-related mortality and long-term toxicities in survivors. This limits the further intensification of therapy.

Genetic and Molecular Heterogeneity: AML is highly heterogeneous at the genetic and molecular levels, and this heterogeneity complicates MRD detection, affects treatment response, and limits universal treatment approaches. Therefore, this suggests the need for precision medicine.

Treatment Strategies for Relapsed Disease: Treatment strategies for relapsed pediatric AML include targeted therapies, drug profiling-guided regimens, and bridging to hematopoietic stem cell transplantation (HSCT).

7. Conclusions

Pediatric acute myeloid leukemia (AML), though less frequent than ALL, poses a formidable challenge because of its aggressive course and tendency to relapse. Over the last three decades, there has been remarkable progress that has resulted in better remission rates and overall survival. Much of this improvement has been based on advances in risk stratification, chemotherapy regimens, hematopoietic stem cell transplantation (HSCT), and the development of new therapeutic strategies.

Risk stratification, incorporating cytogenetic, molecular, and clinical factors, is necessary to individualize the intensity of treatment. Patients with favorable risk may receive less intensive therapy, while patients with adverse risk require more intensive measures, typically with HSCT. HSCT remains a significant component of the treatment for high-risk childhood AML, with the potential for long-term remission through the exchange of the involved bone marrow with healthy stem cells. Optimization of donor selection, conditioning regimens, and post-transplant complication control are critical to the optimization of HSCT outcomes.

Despite these advances, challenges remain. Detection and eradication of residual disease (MRD) are critical to preventing relapse. Technologies like multiparameter flow cytometry (MFC) and next-generation sequencing (NGS) are valuable for monitoring MRD, but both are imperfect. The genetic and molecular heterogeneity of AML complicates MRD detection and limits universal treatment approaches, reinforcing the need for precision medicine.

Development of targeted immunotherapies offers very exciting new therapy choices. Targeting CD33, CD123, CLL-1, and CD47 antigens and exploring checkpoint inhibitors has the ability to further improve outcomes in children with AML. Additionally, incorporating pharmacogenomic profiling and functional drug sensitivity testing may enable truly personalized treatment approaches, especially for relapsed and refractory disease. Current studies continue to optimize these strategies, minimize toxicity, and improve the efficacy of these emerging treatments. The long-term vision is to continue to improve long-term survival and quality of life in children with AML.