Allogeneic Hematopoietic Stem Cell Transplantation for Acute Myeloid Leukemia and Myelodysplastic Syndrome in Adults

Hematopoietic Stem Cell Transplantation for Acute Myeloid Leukemia in First Complete Remission

Adult AML patients can be stratified into good-risk, intermediate-risk, and poor-risk groups based on the disease characteristics including numeric and/or structural chromosomal abnormalities and the presence of specific mutations that can also direct choice of curative post-remission therapy, especially for patients less than 60 years of age (see Chapter 60 ). For instance, current European LeukemiaNet (ELN) consensus incorporates molecular mutations to derive AML prognostic risk groups ( Table 62.1 ). In current thinking, allogeneic HSCT is generally recommended when the relapse incidence with non-transplant approaches is expected to be above 35% to 40%.

Numerous groups have evaluated prospectively the relative benefits of allogeneic HSCT versus non-allogeneic therapies (consolidation chemotherapy or autologous HSCT) in adult AML patients 18 to 60 years of age in CR1. Treatment allocation was by biologic assignment as a surrogate for true randomization, with allogeneic HSCT for patients with available human leukocyte antigen (HLA)–matched sibling donors (donor group) and non-allogeneic consolidation for those lacking matched sibling donors (no-donor group). Some studies further stratified post-remission outcomes by cytogenetic risk. Individual studies, when analyzed on an intent-to-treat (ITT) donor versus no-donor basis, typically demonstrated that allogeneic HSCT was effective at improving disease-free survival, but overall survival benefit was hard to document because of graft-versus-host disease (GVHD) and other treatment-related mortality. Moreover, transplantation in CR2 salvages some of the patients who relapsed, thus making overall survival similar with the two approaches.

A meta-analysis of over 6000 patients in 24 biologic treatment assignment trials, however, confirmed that overall survival was in fact better in allogeneic matched related donor transplantation in AML-CR1, with a hazard ratio (HR) of death at 0.90 (95% confidence interval [CI], 0.82 to 0.97). Importantly, when stratified by cytogenetic risk, there was an overall survival benefit of allogeneic transplantation ( Table 62.2 ) for both intermediate-risk (HR, 0.83; 95% CI, 0.74 to 0.93) and poor-risk (HR, 0.73; 95% CI, 0.59 to 0.90) cytogenetic groups, but not for good-risk AML-CR1 (HR, 1.07; 95% CI, 0.83 to 1.38) ( Fig. 62.1 ). These findings were particularly relevant for patients with normal cytogenetic AML, who constitute the largest subgroup and for whom no consensus regarding optimal post-remission treatment was previously available. These studies however did not have the benefit of prognostic assignment based on FLT3, NPM1, or other molecular markers.

Prognostic Factors for Acute Myeloid Leukemia in First Complete Remission

Intermediate-Risk Acute Myeloid Leukemia

The intermediate-risk cytogenetic group is heterogeneous and includes cytogenetically normal disease as well as those with karyotypic abnormalities not meeting criteria for good- or poor-risk AML. It previously constituted the largest risk category, accounting for approximately 40% to 50% of adult AML patients less than 60 years of age, but more recently with the refinement of molecular risk predictors, this category has shrunk. Long-term disease-free survival of approximately 40% to 45% is anticipated after consolidation chemotherapy in intermediate-risk AML. However, identification of mutations with prognostic importance such as mutant FLT3 internal tandem duplication (FLT3-ITD), NPM1, and CEBPA (CCAAT/enhancer-binding protein-α) have further individualized treatment decision-making, including allogeneic transplantation.

Over the past decade, “normal” karyotype AML (NK-AML, the largest category within intermediate-risk AML) has been documented to encompass molecular abnormalities with divergent risk. The presence of an isolated NPM1 or biallelic CEBPA mutation improves disease prognosis to slightly below that of AML with CBF translocations, placing them in the favorable-risk category, and most investigators do not offer HSCT in first remission. The favorable impact of NPM1 mutations appears to persist albeit to a lesser extent in older patients, and allogeneic HSCT may be a consideration in older AML. In contrast, patients with an isolated FLT3-ITD mutation and NK-AML have outcomes similar to those with poor-risk karyotype. In a report evaluating the 2010 ELN risk classification in a large cohort of patients, for those in the “intermediate I” risk group (includes all patients with NK-AML with FLT3 abnormalities and those lacking both FLT3 and NPM1 mutations), relapse-free survival (RFS) was improved with allogeneic HSCT (94 vs. 7.9 months without allogeneic HSCT).

Although the better-prognosis subgroup within NK-AML may not benefit from early allogeneic transplantation, HSCT appears to be the preferred post-remission therapy for other intermediate-risk AML patients at higher risk for relapse, including those lacking favorable gene mutations like NPM1 without FLT3-ITD, or double mutant CEBPA. However, the dissection of disease complexity requires detailed analysis to identify optimal therapy. For instance, Dohner et al. validated the 2017 NPM1/FLT3-ITD ELN AML risk classification, documenting that allogeneic HSCT benefit is restricted to patients in the “adverse-risk” group (NPM1 ^wt /FLT3-ITD ^high ; NPM1 ^mut /FLT3-ITD ^high with high-risk molecular mutations (RUNX1 , ASXL1 , TP53) ; and NPM1 ^wt /FLT3-ITD ^low AML with high-risk molecular markers and/or adverse-karyotype). For the “intermediate-risk” group (NPM1 ^mut /FLT3-ITD ^high ; and NPM1 ^wt /FLT3-ITD ^low , both without adverse-risk karyotype), allogeneic HSCT benefit was restricted to patients who did not receive midostaurin-based treatment on the CALGB 10603/RATIFY trial. Of note, this finding is relevant to AML patients aged 18 to 60 years and requires validation in independent datasets prior to its routine incorporation in clinical decision making. Further, extrapolation to patients aged above 60 years is not warranted, given their overall worse disease outcomes.

Measurable Residual Disease Assessment in Acute Myeloid Leukemia

Sequential monitoring of MRD via sensitive phenotyping (e.g., multicolor flow cytometry [MFC]) or molecular methods (e.g., quantitative PCR [qPCR], digital droplet PCR [ddPCR], deep error-corrected next-generation sequencing [NGS]) to detect AML-associated immunophenotypes or mutations respectively, at a sensitivity of at least 1 × 10 ⁻³ , is applicable to the vast majority of AML and provides a useful guide to post-remission AML management. Indeed, MRD negativity is emerging as an alternative therapeutic endpoint that supplements (and may replace) morphologic AML remission criteria.

As reviewed by Dohner et al., mutational assessment at early post-remission time points can distinguish patients at a differing probability of relapse. As they highlight, MRD status has been found to be a better predictor of relapse risk than the presence of cooperating mutations involving KIT and FLT3 -ITD in CBF-AML, or FLT3 - ITD , DNMT3A , and WT1 in NPM1 -mutated AML. Sequential MRD-monitoring studies have shown that persistent high-level PCR positivity, or a rising level of leukemic transcripts after an initial molecular response, invariably predict relapse. Validated AML-associated mutations for MRD monitoring include RUNX1-RUNX1T1 , CBFB-MYH11 , and PML-RARA translocations ( KMT2A and DEK-NUP214 translocations are also AML-associated). Other likely prognostic MRD mutations include NPM1 , RAS pathway mutations ( FLT3-ITD and -TKD ), N- and K-RAS , PTPN11 or KIT , whose persistence in CR indicates a likely higher risk of relapse. However, in the non-transplant context, the persistence of AML-associated mutations needs to be distinguished from the persistence of underlying clonal-hematopoiesis (CH) or MDS-associated mutations (e.g., DTA: DNMT3A, TET2, ASXL1). Future studies will help further distinguish mutations that are reliable indicators of leukemic clones associated with AML relapse versus mutations associated with preleukemic clones that are poorly predictive of relapse, despite persistence at high levels after chemotherapy and during remission. Whether MRD-based early interventions can prevent overt relapse remains under investigation. Preemptive therapy may be particularly relevant with allogeneic HSCT where MRD status may inform conditioning strategy, or post-hematopoietic cell transplantation (HCT) measures aiming to avoid frank relapse. These data support routine inclusion of MRD assessment to help inform transplant decision making in AML-CR1.

Patients with persistent MRD or with early MRD reoccurrence can receive salvage therapy as a bridge to allogeneic HSCT before hematologic relapse or proceed directly to transplant depending on donor availability and the likelihood of success with salvage therapy. The GIMEMA AML1310 study undertook risk-adapted treatment assignment of young adults with newly diagnosed intermediate-risk AML to allogeneic HSCT based on MFC-based MRD assessments, reporting promising 2-year disease-free survival and overall survival of 67% and 70% respectively for MRD ⁺ intermediate-risk AML patients assigned allogeneic HSCT. Importantly, their finding of similar survival of MRD ⁺ intermediate-risk AML patients (assigned allogeneic HSCT) versus those who were MRD ⁻ (assigned autologous HSCT) suggests that allotransplantation may ameliorate the negative impact of persistent MRD, although other studies suggest it does not fully abrogate the negative effect of unfavorable pre-transplant MRD. For instance, in NPM1 -mutant AML, high-level MRD for NPM1 , or low-level MRD for NPM1 plus presence of FLT3-ITD , had impaired post-transplant survival. More intensive conditioning may offer therapeutic benefit for patients with persistent MRD pre-transplant. Deep error-corrected 13-gene NGS-based reanalysis of the BMTCTN 0901 study randomizing older AML patients to MAC versus RIC documented relapse protection and survival benefit for pre-transplant MRD ⁺ patients assigned MAC HSCT. Importantly, the benefit was restricted to those with persistent AML-associated mutations, but not for those with CH-associated (DNMT3A , TET2 , ASXL1) mutations pre-transplant. These findings need to be validated prior to incorporation in routine clinical practice.

In the post-transplant context, persistent or recurrent recipient MRD positivity is also associated with an increased risk of AML relapse. However, many uncertainties remain, for instance with regards to MRD assessment modalities, thresholds, assessment frequency and timepoints, and the prognostic implications of recurrence or persistence of specific mutations and their recrudescence patterns. Importantly, for post-transplantation testing, detection of either AML- or CH-associated mutations would be concerning for recrudescent donor hematopoiesis and disease. However, therapeutic interventions based on the post-transplant assessment of MRD, ranging from immune suppression (IS) taper, donor lymphocyte infusion (DLI) and modified DLI (mDLI), hypomethylating agents (HMAs), and AML mutation-specific targeted therapies (e.g., FLT3 inhibitors) may offer utility and are currently undergoing prospective evaluation.

Transplantation Regimen Intensity in Acute Myeloid Leukemia

Myeloablative HSCT remains the standard of care for younger adults with AML. However, treatment-related mortality after MAC remains appreciable. Lower-intensity regimens offering less treatment-related toxicity are gaining in popularity. One report retrospectively compared RIC transplantation with chemotherapy in high-risk AML on a donor versus no-donor basis. They identified a leukemia-free survival benefit in the donor group (54% vs. 30%, P = .01). This indicates that both MAC and RIC HSCT offer antileukemic benefit compared to non-transplant therapies.

Several groups compared MAC versus RIC transplantation, documenting similar overall and leukemia-free survival, with some reporting lower treatment-related mortality offset by increased relapse risk, especially for patients not in complete remission at the time of HSCT. A CIBMTR study compared 3731 MAC, 1041 RIC, and 407 nonmyeloablative (NMA) transplantations, reporting adjusted 5-year overall survival (OS) of 34%, 33%, and 26%, respectively. The authors concluded that, although NMA transplantation increased relapse and impaired survival, RIC and MAC regimens had similar overall and leukemia-free survival. While a small randomized trial documented similar outcomes of MAC versus RIC transplantation in AML-CR1, a larger prospective randomized trial of MAC versus RIC transplantation in AML or MDS patients closed after the accrual of 272 of planned 356 patients due to an increased relapse and impaired long-term survival in the RIC arm. Additionally, as discussed previously, post-hoc analysis of this study documented that ablative conditioning can offer therapeutic benefit for patients with detectable pre-transplant MRD, offering relapse protection and survival benefit compared to RIC HSCT.

RIC transplantation does, however, offer the option of curative allogeneic HSCT in older or sicker AML-CR patients for whom the toxicity of MAC transplantation might be prejudicial, especially if combined with novel approaches to augment the GVL effect in the peri- and post-transplant setting. Conversely, patients with good performance status but high AML relapse risk could be preferentially selected for MAC.

Alternative Donor Transplantation in Acute Myeloid Leukemia

HSCT from an HLA-matched sibling donor has been the standard of care for prospective biologic treatment assignment studies evaluating post-remission therapies. Patients lacking sibling donors could be considered for alternative donors: unrelated adult donors, umbilical cord blood (UCB), or haploidentical donor transplantation.