Guidelines for allo-HSCT in the treatment of pediatric AML and MDS

Background

In 2007, the American Society for Blood and Marrow Transplantation (ASBMT) published evidence-based guidelines around hematopoietic stem cell transplantation (HSCT) for the treatment of children with acute myeloid leukemia (AML). These have recently been revised by the now American Society for Transplantation and Cellular Therapy (ASTCT). Katherine Tarnock and colleagues conducted a comprehensive literature review to identify best evidence and research from both pediatric and adult studies, which was published in Transplant and Cellular Therapy.¹ Identified literature was reviewed by 14 panel members, graded by consensus, and assigned levels of evidence. The final guidelines were drawn up through further panel and committee review, and the recommendations under each of the key areas are summarized here.

Selecting patients for transplant

With high-risk cytogenetic abnormalities seen in 20─25% of pediatric patients with AML, and high-risk molecular abnormalities well-characterized, the panel made recommendations on which genetic anomalies should be considered an indication for HSCT after first complete remission (CR1). In particular, FLT3 mutations are associated with poor outcomes in children, with a 4-year progression-free survival of only 4%, and an 83% relapse rate relative to FLT3-wild type AML.  The increasing use of next-generation sequencing enables the detection of genetic anomalies associated with high-risk AML, which may not always be detected by karyotyping. Such mutations can be distinct to the pediatric population. A number of high-risk genetic alterations were highlighted by the panel as putative indications for transplant after CR1, including:

• Monosomy5/del(5q), including 5q31, loss of EGR1
  
• Monosomy 7
• High allelic ratio FLT3/internal tandem duplication (ITD)
• Abnormalities of 3q: inv(3)(q21.3q26.2)/t(3;3)(q21.3q26.2)/RPN1-MECOM, t(3;21)(q26.2;q22)/RUNX1-MECOM, t(3;5)(q25;q34)/NPM1-MLF1
• t(6;9)(p22.3;q34.1)/DEK-NUP214
• inv(16)(p13.3q24.3)/CBFA2T3-GLIS2
• 11p15 rearrangement/NUP98-any partner gene (e.g., t(5;11)(q35;p15.5)/NUP98-NSD1, NUP98-KDM5A)
• t(4;11)(q21;q23.3)/KMT2A-AFF1 (MLL-MLLT2)
• t(6;11)(q27;q23.3)/KMT2A-AFDN (MLL-MLLT4)
• t(10;11)(p12.3;q23.3)/KMT2A-MLLT10
• t(10;11)(p12.1;q23.3)/KMT2A-ABI1
• t(11;19)(q23.3;p13.3)/KMT2A-MLLT1 (MLL-ENL)
• Abnormalities of 12p (ETV6): 12p13.2 rearrangement/ETV6-any partner gene, deletions of 12p.13.2/loss of ETV6

The panel also considered other features of high-risk AML. Long-term survival is 15─30% in pediatric patients with AML treated with HSCT who have experienced primary induction failure, defined as failure to achieve <5% morphologic blasts after first or second induction. This is compared to only 5% in patients treated with chemotherapy alone, meaning patients with primary induction failure should be considered for early HSCT at end of induction (EOI). In addition, minimal residual disease (MRD) assessment at EOI allows for sensitive disease detection and the identification of patients at higher risk of relapse and therefore HSCT; 3-year relapse-free survival has been reported at 29% in MRD-positive patients versus 65% in MRD-negative patients.

Recommendations

• Children with favorable risk cytogenetic and molecular lesions (e.g., CBF, NPM1, CEBPA bZip) should not undergo HSCT in CR1, even if they are MRD-positive at first EOI.
• Children with FLT3-ITD mutations or high-risk cytogenetic features should undergo HSCT in CR1.
• Even if not in CR1, children should be considered for allogeneic HSCT following primary induction failure or refractory disease after 2–3 cycles of chemotherapy.
• In the case of relapse, children should be offered HSCT following attempts to obtain complete second remission (CR2).
• At EOI, children who are MRD positive with no other risk-stratifying lesion should be considered for HSCT in CR1.

Conditioning regimen

At present there is no standard HSCT conditioning approach widely accepted for use in AML, although the panel acknowledged that in the United States, myeloablative conditioning (MAC) is preferred to reduced intensity approaches. Both strategies have risks and benefits.

Recommendations

• Reduced-toxicity or reduced-intensity conditioning should be considered in children with inherited bone marrow failure syndromes who develop myelodysplastic syndromes (MDS).
• MAC should be used for patients undergoing first or second HSCT, given a suitable donor and absence of comorbidities that demand a reduced-intensity approach.
• Total body irradiation should not be part of pediatric MAC.

Donor selection

Matched sibling donor HSCT is generally used in pediatric AML, based on existing evidence of lower rates of graft-versus-host disease (GvHD) and improved survival compared to matched or mismatched unrelated donors. However, more recent retrospective studies have not established significant differences between pediatric patients receiving matched sibling donor, matched unrelated donor, or cord blood HSCT. Whilst benefiting from a reduced risk of chronic GvHD, cord blood recipients experience greater prevalence of treatment-related mortality, and delayed platelet, neutrophil engraftment, and immune reconstitution compared to bone marrow and peripheral blood recipients.

Recommendation

• If available, bone marrow should be chosen over unmodified peripheral blood as the donor cell source.

HSCT recommendations in special populations

Special populations include those patients with AML who have unique or specific disease characteristics that influence the behavior of their malignancy. For example, patients with Down Syndrome or treatment-related disease (e.g., secondary to etoposide) have higher risk disease.

Another special AML population are patients with central nervous system (CNS) or cerebrospinal (CSF) involvement (collectively referred to as extramedullary disease [EMD]). EMD is more prevalent in pediatric patients with AML than soft or connective tissue involvement. The panel identified that, while CNS relapse is more common in patients who have been treated for CSF-positive AML, survival outcomes are similar for patients with or without CNS involvement. Clinical trials have not demonstrated treatment benefit with HSCT over chemotherapy for AML with CNS disease; therefore, it is not considered a high-risk characteristic.

The panel also gave consideration to MDS. In most children, MDS are related to inherited genetic predisposition syndromes or inherited bone marrow failure syndromes. Around 50─60% of pediatric patients with MDS can be cured with HSCT (depending on stage), with a 5-year overall survival of 63% for late-stage MDS. Survival outcomes for pediatric patients with treatment-related MDS are poor with chemotherapy alone, with HSCT the only curative treatment and approximately 30% survival.

Recommendations

• HSCT should be considered for patients with Down Syndrome-related AML, with relapsed or highly refractory disease.
• In the case of relapsed or refractory AML with EMD, patients should be considered for HSCT, with radiation to disease sites persisting after completion of conditioning therapy.
• In patients with therapy-induced AML, intensive chemotherapy should be considered to reduce tumor burden prior to HSCT.
• HSCT should not be considered for patients with acute promyelocytic leukemia (APML) in CR1.
• For relapsed APML, autologous HSCT should be considered in CR2 with PML-RARα polymerase chain reaction negativity.

Preventing relapse after HSCT

In children, relapse rates after HSCT are between 25% and 60%. A number of strategies are under development with the aim of reducing the risk of relapse, such as:

• Prophylactic maintenance chemotherapy
• Withdrawal of immunosuppression
• Donor lymphocyte infusions
• Targeted small molecule inhibitors
• Adoptive immunotherapy (e.g., antigen specific T-cell receptor or chimeric antigen receptor T cells)

Recommendations

• AML treatment should be given to maximize the depth of any remission induced before patients receive donor lymphocyte infusions.
• Patients positive for FLT3-ITD mutations should receive FLT3 inhibitors after HSCT.

Conclusion

Considering the treatment of pediatric AML, HSCT remains a core treatment in certain populations. Further refinement of knowledge around who will benefit the most, and who should undergo HSCT early, is needed. Consideration of high-risk factors informs choice of HSCT conditioning regimen and helps to reduce the potential for treatment-related mortality. While the main aim of HSCT is to eliminate disease, late effects, GvHD, and treatment-related comorbidities must be minimized given the potential for a long life-expectancy in children in whom it is curative.