Acute Myeloid Leukemia – Treatment and Research Outlook in 2021, and the MD Anderson Approach

Abstract

Background

Unraveling the pathophysiology of acute myeloid leukemia (AML)has resulted in rapid translation of the information into clinical practice. After 40+ years of slow progress in AML research, the Food and Drug Administration approved, since 2017, 10 agents for different AML treatment indications. Herein we review the research and treatment progress in AML

Methods

Key publications related to AML research and therapy in the English literature since 2000 were reviewed and pertinent information incorporated into the review

Results

The notable subsets of AML include acute promyelocytic leukemia (APL), core-binding factor leukemia (CBF) AML, AML in younger/fit patients for intensive chemotherapy, and AML in older/unfit patients (usually at the age cut-off of 60-70 years). We also consider within each subset whether the AML is primary or secondary (therapy-related, evolving from untreated or treated myelodysplastic syndrome or myeloproliferative neoplasm). In APL, therapy with all-trans retinoic acid and arsenic trioxide results in estimated 10-year survival rates of 80+%. Therapy of CBF-AML with fludarabine, high dose cytarabine and gemtuzumab ozogamicin (GO) results in estimated 10-year survival rates of 75+%. In younger /fit patients, the 3+7 regimen produces poor results (estimated 5-year survival rates of 35%; worse in real-world experience); better regimens incorporating high dose cytarabine during induction, adenosine nucleoside analogs and GO are potentially producing better results. Adding venetoclax, FLT3 and IDH inhibitors into these regimens is resulting in preliminary encouraging data. In older/unfit patients, low-intensity therapy with hypomethylation agents (HMAs) and venetoclax is now the new standard of care. Better low intensity regimens incorporating cladribine, low dose cytarabine and other targeted therapies (FLT3 and IDH inhibitors) are emerging. Maintenance therapy has now a definite role in AML, and oral HMAs with different potential treatment benefits in AML are also now available.

Conclusion

The therapy of AML is evolving rapidly, and the treatment results are improving in all AML subsets, as we gradually incorporate the novel agents and strategies into the original backbone of AML chemotherapy

INTRODUCTION

Understanding of the biology and pathophysiology of acute myeloid leukemia (AML) has accelerated translational discoveries and is contributing to a steady improvement in the outlook and prognosis of AML. ^(1-4) Recent important transitions into clinical practice include small-molecule targeted therapies such as the fms-like tyrosine kinase 3 (FLT3) and isocitrate dehydrogenase (IDH) inhibitors, and the BCL2 inhibitor venetoclax. The “3+7 regimen” (3 days of daunorubicin + 7 days of cytarabine) developed in the late 1970s is an accepted standard of care, producing estimated 5-year survivals of 30%-35% in younger patients (60 years or less),⁽⁵⁾ and 10%-15% in older patients.⁽⁶⁾ This intensive chemotherapy regimen, investigated in cooperative group trials that include highly selected patients, may translate into even worse outcomes in community oncology practices, which see significantly older patients and those with secondary AML and multiple comorbidities (i.e. hypertension, diabetes, cardiac and other organ dysfunctions). Figure 1 shows the MD Anderson outcomes in AML in younger and older patients between 1970 and 2020.

Figure 1.

(A). Survival of younger patients with de novo AML treated at MD Anderson over five decades

Figure 1(B). Survival of older patients with de novo AML treated at MD Anderson over five decades

Acute myeloid leukemia is not one entity, but an umbrella diagnosis comprising multiple sub-types with different prognostic and predictive features, and that can be effectively treated with selected and targeted therapies, which are still being optimized. For example, the chemotherapy-free regimen of all-trans retinoic acid (ATRA) and arsenic trioxide produces estimated cure rates of more than 80% in acute promyelocytic leukemia (APL).^(7-10) The addition of gemtuzumab ozogamicin (GO; CD33-targeted monoclonal antibody conjugated to calicheamicin) to high-dose cytarabine-based chemotherapy in core binding factor (CBF) AML has increased the estimated long-term survival rate from 50% to 75%.^(11-15)

Ongoing research now focuses on multiple AML subsets and treatment combinations. These include 1) novel intensive chemotherapy regimens for younger patients (and fit older patients) adding high-dose cytarabine, nucleoside analogs (fludarabine, cladribine) and targeted agents (FLT3 or IDH inhibitors, venetoclax) during induction and consolidations ; 2) lower-intensity regimens for older/unfit patients utilizing hypomethylating agents (HMAs; azacitidine or decitabine; parenteral or oral formulations) and/or low doses of cladribine-cytarabine rotating with HMAs, to which venetoclax or other targeted inhibitors are added as indicated; 3) novel therapies targeting TP53-mutated AML (e.g. APR246, a TP53 modulator; magrolimab, an anti-CD47 monoclonal antibody that enhances macrophage-mediated phagocytosis) and mixed-lineage leukemia (MLL1)-rearranged disease (menin inhibitors). Combinations of small-molecule targeted therapies, with or without standard chemotherapy, may further increase survival in AML subsets (as was done in APL), and may improve the cure rates in previously incurable AMLs. In addition, the dogma that maintenance therapy has no role in AML has been put to rest by the results of the recent randomized trial demonstrating a survival benefit with oral azacitidine maintenance after intensive chemotherapy.

While many AML experts continue to adhere to the 3+7 regimen as the standard of care, others do not. In fact, the somewhat defeatist mood that prevailed until 2015 has evolved into a highly optimistic vision, as the previously meager AML therapeutic armamentarium has now been enriched by several important anti-AML agents now approved by the Food and Drug Administration (FDA). In this review, we discuss progress in AML research, outline the MD Anderson approaches in 2021, and explore investigational strategies for the coming years.

RELEVANCE OF THE CYTOGENETIC AND MOLECULAR ABNORMALITIES, AND OF MEASURABLE RESIDUAL DISEASE, IN AML IN REMISSION

Cytogenetic abnormalities

A modification of the National Comprehensive Cancer Network (NCCN) cytogenetic-molecular classification of AML is shown in Table 1. The important cytogenetic subsets include: 1) “favorable karyotypes” -- APL, which is characterized by the translocation between chromosomes 15 and 17 [t(15;17) (q22,q21)]; CBF AML, which includes the cytogenetic-molecular subsets of inversion 16 [inv16(p13; q22)] or t(16;16) (p13;q22)] and t(8;21)(q22;q22); 2) “intermediate karyotypes” – essentially diploid (normal) karyotype (about 40%-50% of patients); 3) “unfavorable karyotypes”—these include complex (three or more chromosomal abnormalities) and most MLL translocations (translocations involving 11q23) ; and 4) others. Also, patients with translocations involving chromosome 3q26.2 (EVI1), the location of the MECOM gene, have an extremely poor outcome. ⁽¹⁶⁾ Some studies include particular cytogenetic abnormalities [e.g. single trisomy 8, or single translocation (9;11)] in the intermediate karyotypes (Table 1). ^{(17, 18)} The prognostic significance of a single translocation (9;11) (p22; q23)/KMT2A-MLLT3 has been debated, ^{(19, 20)} but may be intermediate in a small subset of younger patients with de novo AML (not therapy related or secondary). ⁽²¹⁾

Table 1.

Cytogenetic-molecular classification of acute myeloid leukemia (National Comprehensive Cancer Network [NCCN])

NCCN	Cytogenetics	Molecular Abnormalities
Better risk	Inversion (16) or translocation (16;16); Translocation (8;21); Translocation (15;17);	Normal cytogenetics: NPM1 mutation in the absence of FLT3-ITD; or isolated biallelic CEBPA mutation
Intermediate risk	Normal cytogenetics; + 8 alone; Translocation (9;11); 1 Other non-defined	- Translocation (8;21), inversion (16), translocation (16;16): with c-KIT mutation1 - NPM1-mutated and FLT3-ITD mutated (high allelic ratio) - NPM1-wild type and FLT3-ITD wild type - NPM1-wild type and FLT3-ITD mutated (low allelic ratio)
Poor risk	Complex (≥3 clonal chromosomal abnormalities); Monosomal karyotype −5, 5q-, 7, 7q-; 11q23 – non translocation (9:11); Inversion (3), translocations of (3;3), (6;9) or (9;22)	- Normal cytogenetics: with FLT3-ITD mutation - TP53 mutation - RUNX1 mutation - ASXL1 mutation - NPM1-wildtype and FLT3-ITD mutated (high allelic ratio)

¹The NCCN Classification applies generally to younger patients. In older patients, its validation is ongoing, but older patients may have significantly worse outcomes within the same ELN/NCCN risk categories. For example, t (9;11) was shown to be intermediate only in de novo younger AML, but not in older or therapy-related AML. At MD Anderson, we consider any t(---;11q23) to be adverse. We do not consider c-KIT mutation to be adverse in CBF AML . Also, while we considered any FLT3-ITD AML to be non-favorable regardless of the allelic ratio, this is changing rapidly with the incorporation of FLT3 inhibitors into frontline chemotherapy and into post -stem cell transplant maintenance.

Low allelic ratio is <0.5; high allelic ratio is ≥0.5

NCCN=National Cancer Centers Network

Adapted from National Cancer Centers Network (NCCN). Accessed October 9, 2020.

Mutations

Next-generation sequencing identifies one or more somatic mutations in more than 90% of patients with AML. ^(22,23) Frequently mutated genes (more than 5%) include FLT3, NPM1, DNMT3A, IDH1, IDH2, TET2, RUNX1, TP53, NRAS, CEBPA and WT1. Based on functional analysis and known biologic pathways, they are categorized into subsets: myeloid transcription-factor fusions or mutations; NPM1 mutations; tumor-suppressor gene mutations; epigenome-modifying gene mutations; activated signaling-pathway gene mutations; cohesin-complex gene mutations; and spliceosome-complex gene mutations. These mutations exhibit patterns of co-occurrence or mutual exclusivity that help to identify AML pathways of clonal dominance and shifts that may guide targeting therapies.

Mutations may have prognostic and/or predictive values, which may be altered with the introduction of targeted therapies (e.g. adding FLT3 inhibitors to frontline chemotherapy). In addition, the prognostic-predictive significance of mutations is more important in normal-karyotype AML ^{(24, 25)}. Among the favorable and unfavorable karyotype subsets, prognosis is largely defined by the cytogenetic abnormalities.

In normal-karyotype AML, mutations of either a bi-allelic CEBPA (2% or less) or nucleophosmin-1 (NPM1; 50%) in the absence of a FLT3-ITD mutation confer more favorable prognoses. ⁽⁴⁾ In contrast, a FLT3 mutation, particularly the internal tandem duplication (FLT3 ITD) variant, defines a poorer prognosis, particularly with a high FLT3 allelic ratio (AR; ie >50%) and no co-occurring NPM1 mutation. In normal-karyotype NPM1-mutated AML, the presence of a FLT3 mutation (about 50% of patients with a diploid karyotype and NPM1 mutation) predicts a worse outcome historically (before the incorporation of FLT3 inhibitors into frontline AML therapy).

Generally, the burden of a pathogenic mutation is reported as the variant allelic frequency (VAF), which is the percent of the mutated gene over the total (wild type + mutated gene). The exception is the FLT3 mutation, reported as allelic ratio (which creates some confusion). The FLT3-ITD AR is defined as the ratio of the area under the curve of “FLT3-ITD” divided by the area under the curve of “FLT3-wildtype” using a semi-quantitative DNA fragment analysis.⁽²⁶⁾ The AR strongly influenced outcome in several studies of newly diagnosed patients with FLT3-mutated AML who were treated with chemotherapy regimens that did not include FLT3 inhibitors.^(27-29) This may change with the incorporation of FLT3 inhibitors into chemotherapy and into post-allogeneic stem cell transplantation (SCT) maintenance. A higher FLT3-ITD AR (generally defined as ⩾0.5) is associated with worse survival, likely reflecting dominance of the FLT3 clone. At MD Anderson, we add a FLT3 inhibitor to frontline chemotherapy for any level of positivity (even an AR <0.1) to prevent relapse with an expanded FLT3-mutated clone.

Other mutations (ASXL1, RUNX1, TP53 and others) and co-occurrence of multiple mutations may also predict worse outcomes. ^(30-36) Mutations and/or deletions of the tumor suppressor gene TP53 (located on the short arm of chromosome 17) occur in 2%-20% of patients, and are associated older age, complex karyotype and therapy-related disease. ^(31-33)

Several mutations are potentially targetable or respond to specific therapies. The normal-karyotype NPM1-mutated AML is highly responsive to cytarabine-based regimens and to regimens containing HMAs and venetoclax. The FLT3 mutations (30% of AML), including FLT3-ITD and FLT3-TKD (D835 most common), can be targeted with FLT3 inhibitors (midostaurin, gilteritinib, sorafenib, quizartinib). Mutations in the IDH1/2 proteins (20% of patients with AML) can be effectively treated with combinations that contain the IDH inhibitors: ivosidenib for IDH1 and enasidenib for IDH2. In addition, the IDH mutations engender strong BCL-2 dependence for survival, rendering them particularly sensitive to venetoclax-based therapy. ⁽³⁷⁾ Most patients with TP53 mutations do not seem to benefit from intensive chemotherapy and may have similar or improved outcomes, and less toxicity, with lower-intensity approaches. ^{(32,33, 38)} They may also benefit from investigational therapies such as APR246 and magrolimab (discussed later). The cytogenetic-molecular subset of “mixed-lineage leukemia” (translocations involving 11q23; MLL1 rearrangement, now referred to as KMT2A) may respond to novel menin inhibitors (SNDX-5613, KO-539, others). ⁽³⁹⁾ In CBF AML, c-KIT mutations may be associated with worse outcome. ^(40,41) This may be treatment-dependent, since we did not find them to be adverse in patients treated with fludarabine-cytarabine-GO-based regimens. ^(11,12) Potent c-KIT inhibitors (avapritinib, dasatinib) added to chemotherapy may be beneficial. ^(42,43)

The prognostic-predictive value of mutations can be age-dependent and modified by new therapies. For instance, the predictive value of a mutation may be stronger in younger patients (cure rate with intensive chemotherapy/SCT 40%-60%) than in older patients (estimated two-year survival historically less than 20%). Also, the addition of a targeted therapy (eg FLT3 inhibitors or venetoclax) to chemotherapy regimens may alter the prognostic significance of the mutation.

Measurable residual disease in remission

Measuring residual disease in AML in complete remission (CR) is now standard of care. ^(44-50) Detectable measurable residual disease (MRD) at the time of morphologic CR is associated with a higher relapse rate and worse survival. It has been commonly measured using two methodologies, multi-color flow-cytometry (MFC-MRD) and molecular quantification. ^(44-49)

Polymerase chain reaction (PCR) is used to monitor quantitatively certain AML translocations and mutations (e.g. in APL, CBF and NPM1-mutated AML) and is expanding to other molecular subsets (IDH1/2 and FLT3). In APL, monitoring with PCR quantification of promyelocytic leukemia-retinoic receptor alpha (PML-RARA) detects early relapse. ⁽⁵¹⁾ The same applies for CBF AML. Inversion 16 and t(16;16) result in the CBF beta/myosin heavy chain 11 (CBFB/MYH11) fusion gene. The t(8;21) produces the fusion gene of runt-related transcription factor 1 (RUNX1/RUNX1T1). Measurable detection of molecular fusion genes by quantitative PCR in CBF AML, particularly in inversion 16, predicts for relapse. ^(52,53) Patients with the t (8; 21) subtype may have persistent MRD levels below 0.1%, but remain in durable complete remission and and may possibly be cured. Among other subsets of AML patients, monitoring MRD by next generation sequencing (NGS) is informative, for example in patients with NPM1 mutations. ^(54,55) Better outcomes are reported in FLT3- and IDH-mutated AML with molecular clearance. In contrast, the persistence of other mutations by NGS may not be as informative. For example, the persistence of mutations in DNMT3A, TET2 and ASXL1 (DTA mutations, of the “DTA molecular triad”) does not predict for relapse. Combining MFC and PCR modalities may improve the capability of MRD studies to predict for relapse. ⁽⁴⁴⁾

Measurable residual disease in CR may warrant consideration of therapeutic interventions. In APL, therapy at the time of molecular relapse prevented overt hematologic relapse.⁽⁵¹⁾ Allogeneic stem cell transplant (SCT) for persistent MRD in CR in CBF AML improved survival compared with continuation of standard therapy.^(52,53) Interventions that could eradicate MRD in CR may include allogeneic SCT; more intensified chemotherapy regimens; HMAs (parenteral or newly approved oral formulations) plus venetoclax; targeted therapy combinations when indicated for particular molecular abnormalities (FLT3 or IDH inhibitors); antibody therapies (e.g. CD123 or CD33 monoclonal or bispecific antibodies); or immune therapies (e.g. checkpoint inhibitors).

TRANSLATION OF BIOLOGIC INFORMATION INTO CLINICAL PRACTICE AND RESEARCH

As heterogeneous entities, the AML subsets necessarily require different selective therapies, depending on disease biology (cytogenetics, mutations, and pathophysiologic pathways), patient age and co-morbidities, and patient wishes and goals. Next, we will discuss the treatment of AML subsets using FDA-approved agents, as well as approaches with investigational agents.

ACUTE PROMYELOCYTIC LEUKEMIA

Acute promyelocytic leukemia (5%-10% of AML) is characterized by the cytogenetic abnormality t (15; 17), which results in the PML-RAR alpha fusion oncogene and its encoded oncoprotein. The PML-RAR alpha oncoprotein acts as a dominant negative inhibitor of wild-type RAR alpha, causing a maturation block and the clinical-pathologic picture of APL.

Historical perspective with chemotherapy, ATRA and arsenic trioxide

In the 1970s, single-agent anthracyclines (daunorubicin) were first shown to cure APL, at rates of 30%-40%. ⁽⁵⁶⁾ Single agent cytarabine is not curative. ⁽⁵⁷⁾ The addition of cytarabine to anthracyclines (in the 3+7 regimens) did not increase the cure rate substantially, nor did the addition of maintenance therapy with 6-mercaptopurine-methotrexate combinations. ^{(58, 59)} A “differentiation syndrome” with chemotherapy was also reported for the first time in this setting. ⁽⁶⁰⁾ The early mortality from disseminated intravascular coagulopathy (DIC) and bleeding was significant (20%-30%).

In the late 1980s and early 1990s, ATRA and arsenic trioxide were discovered to have major anti-APL activities, through their effect of reversing the maturation block, resulting in a gradual differentiation process. Studies from China, India and Iran of single-agent therapy with ATRA or arsenic trioxide as frontline therapy showed high CR rates and 5-year disease-free survival (DFS) rates exceeding 50%-60%. ^(61-63) This established arsenic trioxide and ATRA as the most effective anti-APL agents. Gemtuzumab ozogamicin was also found to be highly effective. ⁽⁶⁴⁾

Based on the single-agent efficacies of ATRA and arsenic trioxide, ⁽⁶⁵⁾ both agents were added to chemotherapy during induction and/or consolidation in comparative trials that confirmed their added benefits. ^(66-70) In the late 1990s, the combination of idarubicin and ATRA (AIDA regimen) became standard of care in APL. ⁽⁷⁰⁾

The era of ATRA and arsenic trioxide: a chemotherapy-free regimen

In the early 2000s, the MD Anderson group decided cautiously to investigate a non-chemotherapy regimen of ATRA plus arsenic trioxide assalvage (2001), and then as frontline therapy (2002). Gemtuzumab ozogamicin was added for high-risk disease (white cell count> 10x 10⁹/L at diagnosis or during induction). Following the demonstration of the high efficacy of this approach,^(7,8) randomized studies confirmed the superiority of ATRA plus arsenic trioxide over AIDA in low- and intermediate-risk APL.^(9,10,71,72) With ATRA plus arsenic trioxide, the CR rate is 90+%, and the cure rate 80+%. Induction mortality from DIC is low (about 5%), and resistant disease is extremely rare, except in cytogenetic variant-APL (translocations between chromosomes 11 and 17 [PLZF-RAR alpha], or between chromosomes 5 and 17). Patients with high-risk APL have a worse outcome and may benefit from the addition of GO (or anthracylines).

In the non-chemotherapy regimen, ATRA is given at 45 mg/m² orally daily (in 2 divided doses) during induction until achievement of CR, then daily, 2 weeks on and 2 weeks off, for a total of 9 months. Arsenic trioxide is given at 0.15 mg/kg intravenously (IV) daily during induction until complete remission, then daily x 5 every week for 4 weeks, every other month, for a total of 4 courses (total 80 consolidation doses). At MD Anderson, GO 6-9 mg/m² is given for high-risk APL (pretreatment or with rising WBC > 10x10⁹/L during induction) and for PML-RAR alpha persistent MRD (documented twice over 1-2 weeks) 2-3 months into CR. For patients who present with the uncommon picture of DIC with thrombosis (rather than bleeding; may be exacerbated by ATRA), GO 6 mg/m² x1 or idarubicin 6-12 mg/m² daily x 1-2 are the best emergency interventions.

The Medical Research Council (MRC) comparative trial investigated an intermittent dosing schedule of arsenic trioxide 0.3mg/kg on Days 1-5 of each course, then 0.25 mg/kg twice weekly in Weeks 2-8 of Course 1 and Weeks 2-4 of Courses 2-5.⁽⁷¹⁾ Oral formulations of arsenic trioxide may make the treatment of APL more convenient, particularly during the longer-term consolidation ^{(73, 74)}

Figure 2 shows the MD Anderson results in APL among younger and older patients, and the significant improvement in outcomes in the era of ATRA and arsenic trioxide.

Figure 2.

(A). Survival in younger patients with newly diagnosed APL treated at MD Anderson over five decades

Figure 2(B). Survival in older patients with newly diagnosed APL treated at MD Anderson over five decades

Some important, but not well-known considerations in APL management are detailed next. First, granulocyte-colony stimulating growth factors (filgrastim, pegfilgrastim) should never be used in APL as they may induce a florid progression and trigger fatal DIC. ⁽⁷⁵⁾ Second, fluid overload (often confused with “differentiation syndrome”) related to ATRA and arsenic trioxide, as well as the use of high-volume blood product transfusions (fresh frozen plasma, platelets etc.) to prevent the complication of consumptive coagulopathy, can result in pulmonary failure. If not recognized and managed aggressively, pulmonary complications/fluid overload may necessitate intensive care, supplemental oxygen and occasional ventilator support. Interventions include holding ATRA-arsenic trioxide therapy briefly and using intensive diuresis. ⁽⁷⁶⁾ Third, a true differentiation syndrome may occur, possibly resulting in multi-organ failure. This is preventable with the use prophylactic steroids during induction (together with antibiotics and antifungal prophylaxis) and curtailed with interruptions of ATRA and arsenic trioxide therapy during acute episodes. Fourth, among patients with CNS bleeding at diagnosis, the risk of CNS leukemia may increase; two spaced intrathecal cytarabine injections in CR may eliminate this rare complication.

CORE BINDING FACTOR AML

The CBF AML subset constitutes about 10%-15% of adult AML and includes the subsets with chromosomal abnormalities involving inversion 16, t(16;16) and t(8;21).

Historically, CBF AML was treated with cytarabine-anthracycline induction chemotherapy, followed by 1-4 consolidation courses with high-dose cytarabine. The cure rate was 30%-40% with one consolidation and 50+% with 3-4 consolidations.^{(77, 78)} Optimizing the combinations of established drugs (fludarabine plus high-dose cytarabine for 5-6 courses of induction-consolidation; addition of GO to chemotherapy; monitoring and treatment of persistent measurable molecular disease) gradually improved the cure rate from less than 50% to 75+%.^(11-15) A meta-analysis of five randomized trials showed that the addition of GO to chemotherapy improved the estimated 5-year survival from 50% to 75%. ⁽¹⁵⁾ Thus, GO now must be considered an integral part of any CBF AML regimen.

At MD Anderson, we use fludarabine, high-dose cytarabine and GO (FLAG-GO) during induction and consolidations, for a total or up to 6 courses, and modify therapy (e.g. allogeneic SCT, azacitidine-GO-venetoclax) for persistent MRD in CR. Results were better when GO replaced idarubicin, with estimated 5-year survival rates of 80% in both inversion 16 and t(8;21) AML (Figure 3), ^{(3, 12)} and were better in younger patients than in patients 60+ years old (Figure 3). Older patients are treated with lower-dose FLAG-GO/IDA. Patients who cannot tolerate this, or who have persistent molecular disease, may be offered HMA therapy (decitabine, azacitidine) combined with venetoclax/GO, with the treatment duration adjusted according to the MRD results (PCR) or for 12+ months. The MRC trials using the fludarabine, high-dose cytarabine and idarubicin combination (FLAG- IDA +/− GO regimen) also reported cure rates of 80+%. ⁽¹³⁾

Figure 3.

(A). Survival in younger patients with newly diagnosed CBF-AML treated at MD Anderson over five decades

Figure 3(B). Survival in older patients with newly diagnosed CBF-AML treated at MD Anderson over five decades

Frequent mutations noted in CBF AML are FLT3 (15%-20%), c-KIT (25%-30%), RAS (30%-50%) and others. Some historical studies reported that c-KIT or multiple mutations were associated with worse outcomes. ^(40-42) This may be treatment dependent, as it has not been our experience with the FLAG-GO/idarubicin regimen, which may have overcome the adverse effects of the mutations. Targeted therapies may also be considered (avapritinib or dasatinib for c-KIT mutations; FLT3 inhibitors for FLT3 mutations). ^(42,43) Recent studies also suggest that epigenetic mutations (ASXL2 or cohesin/spliceosome mutations) may be adverse.

HOW TO CHOOSE BETWEEN INTENSIVE AND LOWER-INTENSITY CHEMOTHERAPY IN AML?

The median age of patients with AML is 68-70 years, ⁽⁷⁹⁾ yet most of the research with intensive chemotherapy (3+7 and its variations) is conducted in younger patients, and recommended for older patients only if they are considered “fit for intensive chemotherapy.” But outcomes with this approach in oncology community practices may be significantly inferior to those reported in clinical trials. ⁽⁷⁹⁾

In a study of 813 selected patients 60 years and older (median age 67 years) treated with 3+7 (randomization to two doses of daunorubicin), the median survival was 7-8 months and the estimated 3-year survival was 20%. ⁽⁶⁾ The early mortality rate was 11%-12%. This and other experiences ⁽⁸⁰⁾ (from carefully controlled studies in selected patients with good performance status, normal organ function and few co-morbidities) have translated poorly into the community practice. An analysis of the SEER data (a better reflection of real-world experience) in about 29,000 patients with AML showed significantly worse results, even among patients treated more recently (2000-2017). Among patients 40-59 years old with de novo AML (excluding APL and CBF AML), the early (4-week) mortality rate was 27% and the 5-year survival rate 40%. Among patients 70+ years old, the 4-week mortality rate was 45%-50% and the 5-year survival rate less than 5%. ⁽⁷⁹⁾

Evidence suggests the outcomes of patients with AML may not be equal when comparing results from designated cancer centers or academic institutions with a large volume of such patients to those from smaller community practice settings. In a National Cancer Institute (NCI) database of 60,738 patients, the 1-month mortality was 16% in academic centers and 29% in non-academic centers (p<0.001); the 5-year survival rate was 25% versus 15% (p<0.001). ⁽⁸¹⁾ A second study of 7007 patients reported an early mortality rate of 12% in NCI-designated cancer centers versus 24% in non-NCI-designated cancer centers. ⁽⁸²⁾ At MD Anderson, the 4-week mortality with intensive chemotherapy is less than 5%; the early mortality with low-intensity regimens in older AML is 1%-3%. Whenever logistically possible, we encourage AML treatment in leukemia centers of excellence.

Acute myeloid leukemia is a rare and heterogeneous cancer whose management requires cumulative experience in both diagnosis and treatment. It often affects older patients with multiple comorbidities who need intensive chemotherapy in the setting of a compromised marrow and severe cytopenias at diagnosis and throughout therapy. These conditions require the aggressive and consistent use of prophylactic antibiotics, the availability of optimal and prompt supportive care (blood products), skilled emergency centers and treatment facilities, rapid recognition of infections/sepsis and implementation of proper broad-spectrum IV antibiotics, and the timely use of an intensive care unit care when needed. Without these, the risks of serious morbidities, mortality and treatment abandonment are high.

Acute myeloid leukemia in older patients is associated with a distinct disease biology (high incidence of complex karyotype and cytogenetic abnormalities involving chromosomes 5 and 7 (monosomies) and 17; of multiple mutations including TP53 (in 20%); and of secondary and therapy-related AML (20%-30%). At MD Anderson, historical studies with intensive chemotherapy in older patients with AML (60 years and older) produced CR rates of 40%-50%, 4-8 week mortality rates of 26%-36%, median survivals of 4-6 months, and one-year survival rates of less than 30%.^(83,84) By multivariate analysis, independent adverse factors predictive of early mortality were: age 75 years and older; adverse karyotype with three or more chromosomal abnormalities; presence of an antecedent hematologic disorder; poor performance status (ECOG 2–4); creatinine level 1.3 mg/dl or higher; and induction treatment outside a protected environment. The expected 8-week mortality was 10%-19% with the presence of 0-1 adverse factors, and 36%-65% with the presence of 2-5 of them. ⁽⁸³⁾

These poor results led us and others to explore lower-intensity therapy in such patients, and raised the question of how to select patients unfit for intensive chemotherapy. The prevailing approach relies on the leukemia expert’s perception of the patient’s condition. This is highly subjective and the basis of intense discussions among leukemia experts, even within an institution. At MD Anderson, we use the above model. If the expected 4- to 8-week mortality is less than 10%, patients are offered intensive chemotherapy. If it is more than 10%-20%, they are offered lower-intensity approaches. Of interest, one-third of patients may have significant abnormalities detected by computerized tomography (CT) scans of the chest at diagnosis (may reflect infection, leukemic infiltrate, fluid overload, bleeding, other).⁽⁸⁵⁾ Patients with pneumonia at diagnosis have a significantly higher risk of 4-week mortality (15%-20%) with intensive chemotherapy (unpublished). Future studies may need to incorporate pretreatment routine CT chest findings into the predictive models of early mortality.

Historically, many older patients (age 70 or older) with a new diagnosis of AML were offered supportive/palliative or hospice care. ⁽⁸⁶⁾ A MRC randomized trial in this group of patients demonstrated a clear superiority of low-dose cytarabine therapy (20 mg subcutaneously twice daily x 10 days) versus supportive care/hydroxyurea: CR rate 18% versus 1% (p=0.00006); longer survival (odds ratio: 0.60; p=0.0009). ⁽⁸⁷⁾ This study drove home the important message that an active and tolerable treatment can have a significant effect on improving outcome, even among patients deemed only suitable for supportive care.

Next, we will discuss the roles of intensive chemotherapy in younger/fit patients and of lower-intensity chemotherapy in older/unfit patients, as they apply to current standards of care

YOUNGER/FIT PATIENTS WITH AML -- INTENSIVE CHEMOTHERAPY

Summary of the literature using the “3+ 7” anthracycline-cytarabine regimens with high-dose cytarabine consolidation

A series of randomized trials (cytarabine for 5 versus 7 versus 10 days; cytarabine 100 mg/m2 versus 200 mg/m²; different anthracyclines and dose schedules; addition of other agents such as etoposide, 6-mercaptopurine, 6-thioguanine to induction-consolidation) established the 3+7 regimen as a standard of care over the past 40 years: daunorubicin 50-60 mg/m² IV daily x 3, or idarubicin 12 mg/m² IV daily x 3 days; cytarabine 100-200 mg/m² IV continuous infusion daily for 7 days. A Cancer and Leukemia Group B (CALGB) randomized trial reported superior survival with high-dose cytarabine consolidation therapy (3 g/m² IV over 2-3 hours every 12 hours on Days 1, 3 and 5) for 4 courses, compared with lower-dose cytarabine schedules. ⁽⁸⁸⁾ In this study, high dose cytarabine consolidation was followed by four courses of 2+5 chemotherapy, which were omitted in later CALGB studies. This omission may be important because later studies using this regimen in the control arm reported 5-year survival rates of 20%-30% rather than 40%. ⁽⁵⁾ High dose cytarabine became the consolidation standard of care in AML. Other studies investigated lowering the dose of cytarabine (1.5 g/m²), 4-5 versus fewer courses, and the possible benefits of using allogeneic or autologous SCT in first CR. ⁽⁸⁹⁾ The MRC studies suggested that cytarabine doses of 1.5 g/m² and 3g/m² were equivalent , and that outcomes with four or five high-dose cytarabine consolidation courses were equivalent. A study from Korea indicated that a cytarabine dose of 1.5 g/m² or more was better than 1 g/m². ⁽⁹⁰⁾

Better regimens than 3+7

An increasing body of research indicates that there already may be better anti-AML regimens than 3+7. Such regimens incorporate high-dose cytarabine combinations during induction; optimize the choice and dose of the anthracycline (idarubicin; daunorubicin 60mg/m² daily x 3 versus 45 mg/m² or 90 mg/m² daily x 3); add adenosine nucleoside analogs ( fludarabine, clofarabine, cladribine) to the cytarabine-anthracycline combinations; and include the CD33-targeted monoclonal antibody GO in the treatment of favorable and intermediate-risk disease. More recent regimens also incorporate targeted therapies such as FLT3 inhibitors in FLT3-mutated AML (now standard practice), and venetoclax and/or IDH inhibitors in appropriate patients (still investigational). They also now consider the use of oral azacitidine maintenance therapy following its recent FDA approval for maintenance in older AML in first CR.

High-dose cytarabine consolidation is standard of care for AML, ⁽⁸⁸⁾ but is it beneficial during induction? Five studies reported that it is. A meta-analysis of 3 randomized trials in 1,691 patients treated with high-dose cytarabine induction reported improved rates of relapse-free survival (RFS; p=0.03), overall survival (p=0.0005) and event-free survival (EFS; p<0.0001).⁽⁹¹⁾ A Southwest Oncology Group ( SWOG) randomized trial in patients less than 65 years showed a better RFS rate with high-dose cytarabine in younger (less than 50 years; 4-year RFS 33% versus 21%) and older patients (50 to 64 years; 21% versus 9%; p=0.049).⁽⁹²⁾ An Australian randomized trial showed that high-dose cytarabine induction improved the CR duration and RFS.⁽⁹³⁾ An EORTC-GIMEMA randomized trial in 1,942 younger patients (60 years or less) showed that high-dose cytarabine was associated with significantly better rates of CR, EFS, and overall survival among patients 15-45 years. Among patients 45-60 years, high-dose cytarabine was also associated with significant improvements in CR and EFS, as well as a trend for better survival among patients with FLT3-ITD AML or poor prognosis karyotypes.⁽⁹⁴⁾ An Italian randomized trial in 574 patients (median age 52 years; range 16 to 73 years) showed that sequential high-dose cytarabine induction was associated with significantly higher CR and 5-year survival rates.⁽⁹⁵⁾ The MRC trial comparing fludarabine, high-dose cytarabine and idarubicin ( FLAG-IDA regimen) to 3+7 +/− etoposide will be discussed later.

Two randomized trials reported no benefit with high dose cytarabine induction, but their design did not actually address this question. Lowenberg and colleagues randomized 858 younger patients (median age 49 years; range 18 to 60 years) to induction therapy with high-dose cytarabine 1g/m² every 12 hours x 10 versus standard-dose cytarabine 200mg/m² daily x 7, both in combination with idarubicin.⁽⁹⁶⁾ However, all patients received high-dose cytarabine during induction Course 2 ( either 2g/m² every 12 hours x 8 -- total dose 16 g/m²-- for patients randomized to high-dose cytarabine during Course 1; or cytarabine 1g/m² every 12 hours --total dose 12 g/m²-- for patients randomized to standard-dose cytarabine during Course 1). Thus, all patients received high dose cytarabine in at least one of the two induction courses. The recent SWOG S1203 trial randomized patients to: 1) 3+7 induction followed by four consolidations with high-dose cytarabine (3 g/m² twice daily on Days 1, 3, and 5 —total cytarabine 18 g/m²/course x 4 = 72 g/m²), 2) IA +/− vorinostat regimen: idarubicin plus continuous high-dose cytarabine (1.5 g/m² continuous infusion daily x 4) followed by IA consolidations with cytarabine 0.75 g/m² continuous infusion daily x 3 days (2.25 g/m²/course x 4), for a total cumulative cytarabine dose of 16 g/m².⁽⁹⁷⁾ The two arms presumably tested the benefit of high-dose cytarabine induction, but the total dose of cytarabine was 4.5 times higher in the 3+7 arm than in the IA arm. The 3+7 regimen, in fact, delivered more total high dose cytarabine and was, as expected, superior in the CBF AML. However, despite the lower total cytarabine dose given in the IA arm, the results of the two groups were similar among patients with intermediate or adverse karyotypes. The trial design did not address properly the benefit of high-dose cytarabine added to induction (the effect likely nullified by giving more high-dose cytarabine consolidation in the control arm).

The optimal dose-schedule of high-dose cytarabine has been investigated for 30+ years. The established single dose of 3 g/m² may not deliver better anti-AML efficacy and may increase toxicity.^(89,90) At MD Anderson, we use high-dose cytarabine 1.5–2 g/m² daily x 5 (total 7.5- 10 g/m² per course) during induction (3 days during consolidations).

A combination regimen of fludarabine, high-dose cytarabine and idarubicin (FLAG-IDA or FAI), developed at MD Anderson, ⁽⁹⁸⁾ was later evaluated in a randomized trial (MRC AML 15). The FLAG-IDA regimen consists of cytarabine 2 g/m² daily for 5 days, fludarabine 30 mg/m² daily for 5 days, and idarubicin 8-10 mg/m² daily for 3 days. Among patients who tolerated four courses on the FLAG-IDA arm (2 FLAG-IDA + 2 high-dose cytarabine), the 8-year survival rate was 66% versus 47% in the standard 3+7 +/− etoposide arm. ^{(13, 89, 99)} The FLAG-IDA/FAI is intensive and requires cumulative expertise to deliver safely. But it is not more difficult to deliver than allogeneic SCT, and likely provides a 20% benefit in 8-year survival. Candoni and colleagues treated 130 newly diagnosed patients with AML (age less than 65 years) with FLAG-IDA and GO. They reported a CR rate of 82% and an estimated 5-year survival of 52% (10-year survival about 44%). ⁽¹⁰⁰⁾ In a retrospective analysis of a single-center experience using 3 + 7 (n=86) or FLAG ± idarubicin (n=218), patients treated with FLAG ± idarubicin were more likely to achieve remission after one course of induction (74% versus 62%; p=<0.001), had a faster time to achieve CR (30 days versus 37.5; p=<0.001), and had significantly better 3-year OS (54% versus 39%; p=0.01) and DFS rates (49% versus 32%; p=0.01). ⁽¹⁰¹⁾ Delivered effectively, FLAG-IDA/FAI is a multifaceted regimen that explores the benefits of high-dose cytarabine induction-consolidation, the addition of an adenosine nucleoside analog (fludarabine), and the use of idarubicin rather than daunorubicin. Leukemia management expertise (supportive care; antibiotics and antifungal prophylaxis; timely transfusion support; management of toxicities; and early, aggressive treatment of infections/sepsis) allows safe and full delivery of this regimen in specialized leukemia centers (more than 4-5 leukemia experts in an oncology group; large AML referral volume).

Other adenosine nucleoside analogs (clofarabine, cladribine) also have been explored in combinations with standard chemotherapy. Two randomized trials confirmed the benefit of adding cladribine to the 3+7 regimen. The first study randomized 400 patients to induction with 3+7 +/− cladribine and reported that adding cladribine resulted in a higher CR rate (64% versus 46%; p=0.0009) and leukemia-free survival rate (44% versus 28%; p=0.05).⁽¹⁰²⁾ The second study compared three arms, one with 3+7 alone and the other two of them adding cladribine and fludarabine, respectively. The addition of cladribine (but not fludarabine) was associated with higher CR (67.5% versus 56%; p=.001) and 3-year survival rates (45% versus 33%; p=0.02), ⁽¹⁰³⁾ and also improved outcome in FLT3-mutated AML. ⁽¹⁰⁴⁾

At MD Anderson, we use regimens that add the adenosine nucleoside analogs to idarubicin and high-dose cytarabine as frontline induction therapy in younger patients with AML (fludarabine in FAI, FLAG-IDA; clofarabine in CIA; cladribine in CLIA) .⁽¹⁰⁵⁾ We are exploring the addition of venetoclax and other targeted therapies (discussed later).

The optimal choice and dose schedule of anthracycline has been evaluated in multiple studies. Historically, daunorubicin 30-60mg/m² daily x 3 was used for induction therapy. Daunorubicin 45mg/m² daily x 3 for induction was inferior to 90 mg/m² in age-specific subsets ^(5,6), but the 60 mg/m² dose was equivalent to the latter and less toxic. ^(106,107) Idarubicin 12 mg/m² daily x 3 is equivalent, or perhaps superior, to daunorubicin. Studies comparing idarubicin and daunorubicin, including a meta-analysis of five randomized trials, suggested that using idarubicin may be associated with higher CR and survival rates. ^(108-111) With FLAG-IDA/CLIA we reduce the idarubicin dose to 8-10 mg/m² daily x 3 to avoid excessive myelosuppression.

The benefit of GO has been confirmed in a meta-analysis of five randomized trials. The drug, approved in 2000, was withdrawn in 2010 and reapproved in 2017 at a lower dose schedule and in combination with chemotherapy (3 mg/m² x 1 during induction and consolidation; 3 mg/m² on Days 1, 4 and 7 during induction). The negative pivotal trial in the US (SWOG S0106), ⁽¹⁴⁾ which prompted the withdrawal, perhaps had a faulty design. Randomization was to 3+7 with daunorubicin 60 mg/m² daily x 3, versus 3+7 with the addition of GO 6 mg/m² on Day 4, but with daunorubicin at 45mg/m² daily x 3 (presumed to be equitoxic but later found to be suboptimal).⁽¹⁴⁾ The other four randomized trials demonstrated a benefit with GO.⁽¹⁵⁾ The meta-analysis involving 3,325 patients showed that the addition of GO reduced the risk of relapse (p=0.0001), and improved the 5-year survival rate (p=0.01). The benefit was most significant in patients with favorable cytogenetics (increased 5-year survival rate from 50% to 75%; p=0.0006) and intermediate cytogenetics (p=0.005). The 3 mg/m² dose was as effective as 6 mg/m² and was associated with fewer early deaths. ⁽¹⁵⁾

An interesting question is the possible role of lomustine, an alkylating agent, in the treatment of AML. In three French studies involving 847 older patients (older than 60 years), the addition of lomustine 200 mg/m2 orally on Day 1 to idarubicin + cytarabine (n=508), compared with the latter two drugs alone (n=339), was associated with a higher CR rate (68% versus 58%; p=0.002) and longer survival (median 12.7 versus 8.7 months; p=0.004). By multivariate analysis, lomustine was an independent favorable treatment variable for achievement of CR and survival prolongation. ⁽¹¹²⁾

At MD Anderson, we use intensive chemotherapy regimens for younger/fit patients that incorporate high-dose cytarabine and adenosine nucleoside analogs ( fludarabine in FLAG-IDA; cladribine in CLIA) during induction and consolidations, and add targeted therapies as indicated: gilteritinib in FLT3-mutated AML, and venetoclax (7-14 days) in non FLT3- mutated AML. Allogeneic SCT may be offered to patients in CR based on availability of donor, patient age and co-morbidities, pretreatment AML characteristics (cytogenetic, molecular profiles) and MRD status in CR. Allogeneic SCT in first CR should be considered in patients with high-risk disease based on adverse cytogenetics, high FLT3-mutation AR, or MRD positivity by MFC more than 0.1% post first consolidation. Otherwise, patients complete 4-6 courses of consolidation and are offered maintenance therapy with azacitidine and venetoclax on a clinical trial for 2+ years, or targeted therapies as appropriate (e.g FLT3 inhibitors or IDH inhibitors). With this approach, the CR rate in non-selected younger patients is 70%-80%, and the long-term survival rate is 40%-50% (Figure 1). These results antedate the introduction of targeted therapies into the frontline intensive chemotherapy regimens (FLAG/IDA and CLIA), which are now combined with venetoclax for 7-14 days during induction and for 5-7 days in maintenance, as tolerated (detailed later). ^{(113, 114)}

OLDER/UNFIT PATIENTS WITH AML – LOW-INTENSITY THERAPY

Hypomethylating agents

The poor results with intensive chemotherapy and low dose cytarabine in older/unfit patients with AML prompted the search for alternative strategies. Decitabine was originally developed in Europe in the 1970-80s as a cytotoxic drug at doses of 1,000-2,500 mg/m² per course. ⁽¹¹⁵⁾ Its development was abandoned because of severe prolonged and unpredictable myelosuppression, as well as neurotoxicity. In 1992, awareness of the possible differentiation properties of decitabine led the MD Anderson investigators (H.K.) to import the drug to the United States on an investigator-initiated investigational new drug (IND). Between 1992 and 2000, it was developed as an epigenetic HMA at 1/20^th of the myelosuppressive dose -- 10-20 mg/m² daily for 5-10 days. ^(116-119) The initial collaboration was with the Dutch company Pharmachemie BV, but in 1999, Teva acquired the company and abandoned decitabine because it was thought to be a cytarabine-like drug. SuperGen then acquired it and continued its development as an HMA. This led to the Phase 3 trials in myelodysplastic syndrome (MDS), resulting in its FDA approval for higher-risk MDS in 2006. ⁽¹¹⁹⁾ The Phase 3 pivotal trial of decitabine versus low-dose cytarabine in older AML failed to meet the study-designed primary endpoint, ⁽¹²⁰⁾ although the decitabine arm demonstrated a significant survival benefit with the more mature data. Decitabine was approved by the European Medicines Agency (EMA) for the treatment of older AML, but not by the FDA. The AML Phase 3 study randomized 485 patients 65 years or older to decitabine 20mg/m² IV daily x 5 every month versus supportive care or low-dose cytarabine. In a final analysis, the median survival was 7.7 with decitabine versus 5 months with supportive care or low dose cytarabine (p=0.036). Parallel studies were ongoing with azacitidine, resulting in its approval by the FDA for the treatment of higher-risk MDS, ⁽¹²¹⁾ but not for the treatment of older AML. ⁽¹²²⁾ The azacitidine Phase 3 pivotal study in older AML (AZA-AML-001) randomized 488 patients to azacitidine versus three conventional strategies: low-dose cytarabine, intensive chemotherapy or supportive care. Azacitidine therapy was associated with longer survival (median 10.4 versus 6.5 months; p=0.06; hazard ratio 0.85). ⁽¹²²⁾ Today, decitabine and azacitidine are the most commonly used agents for the treatment of older/unfit patients with AML.

Longer durations of decitabine schedules (20mg/m² daily x 10) were also proposed ^{(117, 123)}. Recently, the FDA approved an oral formulation of decitabine plus cedazuridine (cytosine deaminase inhibitor; combination bioequivalent to IV decitabine). ^(124,125) This may result in effective oral therapies for older/unfit patients with AML (decitabine-cedazuridine plus venetoclax) and may improve tolerance and quality of life with effective post-remission outpatient consolidation therapy.

Comparison of intensive chemotherapy versus HMA therapy in older AML showed better results with HMAs. ^(126,127)

Combined low-intensity chemotherapy regimens

Despite the benefits of HMAs, their value in treating older/unfit patients with AML is modest. Because of the anti-AML efficacy of clofarabine, cladribine and low-dose cytarabine, we evaluated a three-drug lower-intensity regimen combining an adenosine nucleoside analog (clofarabine in one study and cladribine in another) with low-dose cytarabine, in alternating cycles with decitabine, over a period of 18 months.^(128,129) Among 248 patients (median age 69; range 48-85 years) treated with the two regimens, the overall response rate was 66%, the CR rate 59%, the early (4-week) mortality rate 2%, the median survival 12.5 months, and the estimated 2-year survival rate 29%. Among patients with normal karyotype, the median survival was 19.9 months and the estimated 2-year survival rate 45%.^{(128, 129)} Compared with single-agent HMAs (considered a standard of care in older/unfit AML), the double-nucleoside analog/HMA lower-intensity therapy suggested improved results and represents a novel, well-tolerated, effective backbone upon which approaches that add venetoclax and other targeted therapies can be built (discussed later).

EXCITING DISCOVERIES IN AML, PRESENT AND FUTURE

VENETOCLAX

Venetoclax and HMAs/low dose cytarabine

One strategy to target AML involves activation of the intrinsic or mitochondrial pathway of apoptosis, regulated by the BCL-2 family of proteins. Survival/apoptosis involves a dynamic balance of pro-apoptotic effectors (Bak, Bax) and anti-apoptotic proteins (BCL-2, BCL-XL, MCL-1). The latter are overexpressed in AML. Small-molecule “BH3 –mimetics” bind to the anti-apoptotic proteins in the BH3 domain and liberate pro-apoptotic proteins, thus triggering apoptosis. The earlier generation of BH3 mimetics were associated with unacceptable on-target toxicities, including thrombocytopenia.

The BCL-2 inhibitor venetoclax is a more advanced and highly potent BH3-mimetic molecule that retains specificity for BCL2, but without affinity for BCL-XL or MCL-1. It is active against several cancers (chronic lymphocytic leukemia; other lymphoproliferative disorders) and is under investigation in others (acute lymphoblastic leukemia, MDS, lymphoma and myeloma subsets). AML blasts and stem cells depend on BCL-2 for survival and preclinical studies confirmed venetoclax’s activity in AML, ⁽¹³⁰⁾ leading to a Phase 2 single-agent venetoclax study for relapsed AML that showed modest activity. ⁽¹³¹⁾ Responses were more frequent in the IDH-mutated subtype, confirming the preclinical hypothesis that IDH-mutant patients were particularly venetoclax-sensitive. ^{( 131)} Venetoclax was then combined with HMAs and low-dose cytarabine in single-arm and, later, randomized trials in older/unfit patients with newly diagnosed AML. The positive results from the single-arm trials (azacitidine/venetoclax combination: overall response rates 67%; estimated median survival 17.5 months; 2-year survival 40%) led to FDA accelerated approval of venetoclax in combination with HMAs or low-dose cytarabine for the treatment of these patients. ^{(132, 133)}

The VIALE-A Phase 3 pivotal trial randomized newly diagnosed AML patients 75+ years/unfit for intensive chemotherapy to azacitidine +/− venetoclax. Among 431 patients randomized (2:1) to azacitidine plus venetoclax (n=286) or azacitidine alone (n=145), those who received venetoclax had significantly improved survival (median survival 14.7 versus 9.6 months; p<0.001). The response rates (66.4% versus 28.3%; p<0.001) and CR rates (29.7% versus 17.9%; p<0.001) were also better.⁽¹³⁴⁾ A second randomized study of low-dose cytarabine +/− venetoclax (211 patients; 2:1 randomization in favor of the combination) showed similar findings: median survival 8.4 versus 4.1 months, p=0.04; overall response rate 48% versus 13%, p<0.001; CR rate 27% versus 7%, p<0.001, all in favor of the combination.⁽¹³⁵⁾

A single-arm trial of decitabine (10-day induction; maintenance 5 day/month) with venetoclax (14-21 days) in older patients (median age 72 years; range 70-78 years) with newly diagnosed de novo AML showed an overall response rate (CR+CRi) of 84%, a CR rate of 67%, a 4-week mortality rate of 0%, and a median survival of 18.1 months. ⁽¹³⁶⁾

Low -intensity and intensive chemotherapy combinations with venetoclax

A study of the combination of cladribine/cytarabine/venetoclax alternating with azacitidine/venetoclax is ongoing in older/unfit patients with newly diagnosed AML. ⁽¹³⁷⁾ Among 48 patients treated so far (median age 68 years; range 57-84 years), the CR rate is 77% (overall response rate 94%). The MRD negativity rate is 80%, the 4-week mortality is 0%, and estimated 1-year survival is 70%.

In younger/fit patients with newly diagnosed AML, we are investigating intensive chemotherapy (FLAG/IDA, CLIA) in combination with venetoclax (7-14 days during induction; 5-7 days in maintenance). ^{(113, 114)} Among 28 patients treated so far with FLAG/idarubicin-venetoclax, the overall response rate is 93%, and the MRD negativity rate in CR is 92%.⁽¹¹³⁾ Among 31 patients treated with CLIA-venetoclax, the overall response rate is 90%, and the estimated 1-year survival is 78%.⁽¹¹⁴⁾

FLT3 INHIBITORS

FLT3 inhibitors in AML salvage

Investigations of FLT3 inhibitors have now spanned close to two decades. These have included, among others, midostaurin, sorafenib, gilteritinib and quizartinib. Type 1 inhibitors (midostaurin, gilteritinib) are active against both FLT3-ITD and FLT3-TKD mutations. Type 2 inhibitors (sorafenib, quizartinib) are effective only against FLT3-ITD mutations. Newer FLT3 inhibitors may be more effective than earlier ones. ⁽¹³⁸⁾

Therapy with single-agent gilteritinib (type 1 FLT3 inhibitor; dual FLT3-AXL inhibitor) 120 mg daily resulted in composite CR (CRc) rates of 45%-50% in relapsed/refractory FLT3-mutated AML.⁽¹³⁹⁾ The Phase 3 ADMIRAL trial randomized (2:1) 371 patients with relapsed FLT3-mutated AML to gilteritinib 120 mg daily (n=247) or investigator-choice salvage chemotherapy (both high- and low-dose chemotherapy) (n=124).⁽¹⁴⁰⁾ Gilteritinib therapy was associated with a significantly longer survival (median survival 9.3 versus 5.6 months; hazard ratio 0.637; p=0.0007), and higher rates of CR (21% versus 11%; p= 0.013), CR/CRh rate (34% versus 15%), and CRc rate (54% versus 22%).⁽¹⁴⁰⁾ This led to the FDA approval of single-agent gilteritinib as salvage therapy for FLT3-mutated AML. Ongoing studies are combining gilteritinib with HMA therapy and with intensive chemotherapy as well as with venetoclax in frontline, salvage and maintenance strategies.

Quizartinib is a potent type 2 FLT3 inhibitor. In the Phase 1 studies in refractory-relapsed AML, a maximum tolerated dose of 200 mg orally daily was proposed; the dose-limiting toxicity was QTc prolongation. ⁽¹⁴¹⁾ In FLT3-ITD AML, the overall response rate was 50%. This was confirmed in a large Phase 2 study in relapsed-refractory older and younger patients with FLT3-mutated AML.⁽¹⁴²⁾ Later studies evaluated quizartinib 90-135 mg daily, then 30-60 mg daily, to reduce the incidence of QTc prolongation.⁽¹⁴³⁾ A total of 67 patients with refractory- relapse FLT-3 ITD AML were randomized to quizartinib 30 mg or 60 mg daily. The CRc rate was 50% in both arms and the median survivals 6-8 months. The rate of grade 3 QTc prolongation was 3% in both arms (grade 2 11%-17%). ⁽¹⁴³⁾ The randomized Phase 3 QUANTUM-R study evaluated quizartinib versus investigator-choice salvage chemotherapy in 367 patients with relapsed-refractory FLT3-ITD mutated AML. Quizartinib was better than chemotherapy: CRc rate 48% versus 27%; median survival 6.2 months vs. 4.7 months, p=0.0177). However, quizartinib was not granted FDA approval, due in part to concerns over treatment equipoise and robustness of the survival benefit (it was approved in Japan in 2019). Combination studies of quizartinib in relapsed and frontline FLT3-ITD AML are ongoing.

Certain activating point mutations (such as FLT3-ITD F691L) are resistant to current FLT3 inhibitors. Third-generation compounds (FT10101, crenolanib) are under development potentially to overcome these and other mechanisms of resistance (e.g. emergence of MAPK-pathway mutations -- RAS, RAF, PTPN11, NF1). ⁽¹⁴⁴⁾

Combination therapy of FLT3 inhibitors with agents that induce apoptosis may enhance cytotoxicity against FLT3-mutated and wild-type clones and potentially delay or prevent resistance to FLT3 inhibitor-based therapies. Preclinical data indicated strong synergism between venetoclax and FLT3 inhibitors. Ongoing studies are evaluating combinations of HMAs plus gilteritinib, gilteritinib plus venetoclax, and triplet-therapy (HMAs/gilteritinib/venetoclax).

Frontline therapy with FLT3 inhibitors

A Phase 2 RATIFY trial randomized 717 younger patients (<60 years) with newly diagnosed FLT3-mutated AML (median age 48 years; range 18-60; 77% FLT3-ITD, 23% FLT3-TKD) to 3+7 +/− midostaurin. ⁽¹⁴⁵⁾ The addition of midostaurin resulted in a significant survival benefit (median survival 74.7 versus 25.6 months, p=0.009; estimated 5-year survival 50% versus 42%). The benefit was noted in FLT3- ITD low AR (< or = 0.70), FLT3- ITD high AR (> 0.70) and TKD-mutated AML. At MD Anderson, a matched-cohort analysis showed the benefit of adding sorafenib to idarubicin-cytarabine in FLT3-mutated AML. ⁽³³⁾ In our study of CLIA + FLT3 inhibitor (sorafenib/midostaurin), the CR rate was 86% and the estimated one-year survival 70%. ⁽³³⁾

Several trials are evaluating newer-generation FLT3 inhibitors with intensive chemotherapy. In a study of 3+7 plus gilteritinib, the CR rate was 90.5%, the overall response rate 94%, and the median response duration 14.1 months. ⁽¹⁴⁶⁾ Sorafenib added as maintenance therapy post allogeneic SCT in FLT3-mutated AML improved survival and/or relapse-free survival. ^{(147, 148)} The combination of azacitidine and sorafenib in older patients with FLT3-ITD AML resulted in a CR-CRi rate of 78%. ⁽¹⁴⁹⁾

Of interest, several non-targeted chemotherapy strategies have shown benefits in FLT3-mutated AML, including induction regimens containing high-dose cytarabine, cladribine and high-dose daunorubicin. ^{(94, 104, 150)}

IDH INHIBITORS

The IDH 1-2 mutations induce neomorphic IDH enzyme activity, which causes aberrant production of the 2-hydroxyglutarate (2-HG) onco-metabolite. The 2-HG competitively inhibits alpha-ketoglutarate (αKG), leads to dysregulated epigenetic function, a hypermethylated phenotype, and a block in maturation leading to AML. ⁽¹⁵¹⁾ Enasidenib is an orally bioavailable small molecule inhibitor of mutant IDH2, and ivosidenib of mutant IDH1.

In a Phase 1-2 study, 239 patients with IDH2-mutated AML (176 relapsed-refractory) were treated with enasidenib 50-650 mg orally continuously daily. A subset of 109 patients with refractory-relapsed AML received enasidenib at the recommended Phase 2 dose of 100 mg daily, with an overall response rate of 40.3%, a CR rate of 19.3%, a median response duration of 5.8 months, and a median survival of 9.3 months. ⁽¹⁵²⁾ This resulted in the FDA approval of enasidenib 100 mg daily as single-agent therapy in IDH2-mutated refractory-relapsed AML. Grade 3-4 side effects included elevation of indirect bilirubin (12%) and differentiation syndrome (7%, responsive to steroids)

In a Phase 1-2 study, 258 patients with IDH1-mutated AML (179 refractory-relapsed) were treated with ivosidenib. In the Phase 2 efficacy portion that included 125 patients, ivosidenib 500 mg daily produced an overall response rate of 41.6%, a CR/CRh of 30.4%, a CR of 21.6%, a median overall response duration of 8.2 months, and a median survival of 8.8 months. Grade 3-4 side effects included prolongation of QT interval (7.8%) and differentiation syndrome (3.9%). ⁽¹⁵³⁾ Based on this, ivosidenib 500 mg daily was approved by the FDA for the treatment of relapsed-refractory IDH1-mutated AML (as well frontline therapy of IDH1-mutated AML in patients unfit for intensive chemotherapy).

With both IDH inhibitors, RAS/RTK pathway co-mutations and/or high mutational burden (> 6 mutations) were associated with worse results, suggesting the importance of combination therapy. ⁽¹⁵⁴⁾

In a randomized Phase 2 study, 101 older patients with newly diagnosed IDH2-mutated AML (median age 74 years; range 62-85 years) were randomized (2:1) to azacitidine plus enasidenib (n=68) versus azacitidine alone (n=33). The combination produced better results: CR rate 50% versus 12% (p=0.0002); overall response rate 68% versus 42% (p=0.015); median EFS 17.2 months versus 10.8 months (HR 0.59; p=0.13). The median overall survival was impressive but similar in both arms (22.0 months versus 22.3 months), likely because of the availability of effective salvage options (including enasidenib, which was used in at least 24% of patients on the azacitidine-alone arm). ⁽¹⁵⁵⁾

In another single-arm trial, 134 younger/fit patients with newly diagnosed IDH-mutated AML (47 with IDH1 mutations; 87 with IDH2 mutations) received 3+7 chemotherapy plus ivosidenib (for IDH1-mutated) or enasidenib (for IDH2-mutated). ⁽¹²⁰⁾ In IDH1-mutated AML, the overall response rate was 93% and the estimated 1-year survival 79%. In IDH2-mutated AML, the overall response rate was 73% and the estimated 1-year survival 75%. ⁽¹⁵⁶⁾ A randomized, placebo-controlled trial of intensive chemotherapy with ivosidenib or enasidenib versus placebo is ongoing in Europe.

Other IDH inhibitors are under development. Olutasidenib (FT-2102; IDH1 inhibitor) was investigated in a Phase 1-2 trial in patients with IDH1-mutated AML. Thirty-two patients were treated with single-agent FT-2102, and 46 were treated with FT-2102 and azacitidine. The overall response rate was 39% (CR 15%) with single-agent FT-2102 and 54% (CR 23%) with the combination. ^{(157, 158)}

CPX-351

CPX-351 is a nano-scale liposome that contains a fixed 5:1 molar ratio of cytarabine and daunorubicin. ⁽¹⁵⁹⁾ Following the encouraging Phase 1-2 trial results in the subset of secondary AML, a Phase 3 trial in newly diagnosed secondary AML randomized 309 patients to CPX-351 versus 3+ 7. Therapy with CPX-351 was associated with a significantly longer survival (hazard ratio 0.69; p=0.005), and better response rates (CR rate 38% versus 26%, p=0.035; CR+CRi rate 48% versus 33%, p=0.016). CPX-351 was also associated with a longer duration of myelosuppression. The feasibility of later allogeneic SCT was higher in patients achieving CR post CPX-351 (20% versus 12%); their survival was also longer post SCT. This resulted in the FDA approval of CPX-351 as frontline therapy for secondary AML. ^{(160, 161)} Ongoing studies are combining CPX-351 with venetoclax, GO and other targeted therapies.

GLASDEGIB

The Hedgehog (Hh) signaling pathway plays critical roles in embryogenesis and stem cell maintenance. Dysregulation in the Hh pathway can result in the development, maintenance and expansion of leukemic stem cells, which may play an important role in AML pathogenesis, persistence and progression. ⁽¹⁶²⁾

Glasdegib is an orally bioavailable selective inhibitor of the Smoothened (SMO) receptor, a component of the Hh signaling pathway. Following pre-clinical and Phase 1-2 trials, a Phase 2 study investigated low-dose cytarabine +/− glasdegib 100 mg daily. The addition of glasdegib was associated with a significant survival prolongation (median survival 8.8 months versus 4.9 months; 12-month survival 59.8% versus 38.2%).⁽¹⁶³⁾ The FDA approved glasdegib (with low-dose cytarabine) for the treatment of newly diagnosed AML in unfit patients 75 year or older.⁽¹⁶⁴⁾ Glasdegib combinations with azacitidine and intensive chemotherapy are under investigation.

APR-246 in TP53-MUTATED AML

TP53-mutated AML is associated with older age, therapy-related disease, complex cytogenetics, and poor prognosis. With HMA plus venetoclax therapy in older/unfit patients with TP53-mutated AML, the response rate is 55%, but the median survival is only approximately 6 months. ^(134,136)

The investigational drug APR-246 may restore the transcriptional activity of unfolded wild type or mutant p53, leading to induction of apoptosis in TP53-mutated cancers. ⁽¹⁶⁵⁾ The early experience with azacitidine plus APR-246 is producing encouraging results in newly diagnosed older/unfit patients with TP53-mutated AML. An American study treated 55 patients with TP53-mutated disease (40 MDS, 11 AML, 4 CMML). Among 45 evaluable patients, the response rate was 87%, with 24 of the 39 responders achieving CR (53%). Median survival was 11.6 months. ⁽¹⁶⁶⁾ A second study from France treated 52 patients with TP53-mutated disease (34 MDS, 18 AML) with the same combination and reported an overall response rate of 58% (CR 37%). The median overall survival was about 12 months. ⁽¹⁶⁷⁾ A Phase 3 randomized study of azacitidine +/− APR-246 in MDS and AML with 20%-30% blasts has completed accrual.

MAGROLIMAB (CD47 ANTIBODY)

The CD47 protein functions as a macrophage checkpoint, providing a potent “do not eat me” signal. This allows tumor cells to evade detection and immune destruction by macrophages. CD47 is upregulated in AML, and associated with a poor prognosis. ^(168,169) Magrolimab (Hu5F9-G4) is a humanized monoclonal antibody that binds CD47 and blocks its interaction with SIRPα, its ligand on phagocytic cells, resulting in elimination of cancer cells. ⁽¹⁷⁰⁾

The combination of azacitidine plus magrolimab was evaluated in older/unfit patients with newly diagnosed AML or with intermediate-higher risk MDS. Among 34 evaluable patients with AML, the objective response rate was 65%, (40% CR, 12% CRi). Among patients with abnormal pre-treatment karyotype, 50% achieved a complete cytogenetic response. Among patients with TP53-mutated AML, the overall response rate was 71% (15 of 21 patients; CR rate 42%). The estimated median survival was 18.9 months in wild type patients and 12.9 months in mutated patients. ⁽¹⁷¹⁾

MAINTENANCE THERAPY

Maintenance therapy is established as beneficial in many cancers, including acute lymphocytic leukemia. However, for many years, studies in AML did not confirm a survival benefit with maintenance therapy. This changed when the recent positive results reported with oral azacitidine (CC-486) shattered the dogma. In an international multi-center trial (QUAZAR AML-001), 472 older patients (55+ year old; median age 68 years) with AML (non-favorable karyotype) in first CR for less than 4 months were randomized to CC-486 300 mg orally x 14 days every month (n=238) or placebo (n=234). The median survival was 24.7 months with CC-486 versus 14.8 months with placebo (hazard ratio 0.69, p=0.0009). The median RFS was 10.2 versus 4.8 months. As a result, the FDA approved CC-486 as maintenance therapy for this indication in September 2020. ⁽¹⁷²⁾

A second study (HOVON97) randomized 116 older patients (60+ years) with AML in CR post two courses of intensive chemotherapy to azacitidine 50 mg/m² subcutaneously x 5 days/month for 12 courses (n=56) versus observation (n=60). The 12-month DFS was 64% with azacitidine versus 42% with observation (p=0.04). ⁽¹⁷³⁾

It is important to emphasize that the two FDA-approved “oral HMAs,” decitabine-cedazuridine and CC-486, are vastly different. Decitabine- cedazuridine is 100% absorbed and approved as an alternative to IV decitabine in MDS (35 mg flat dose daily x 5/course) based on a study that was designed specifically to demonstrate dose proportionality. CC-486, on the other hand, is poorly absorbed, producing an AUC that is 10%-30% of IV azacitidine. Its FDA approval is as a two-week course every month for maintenance therapy in older AML in first CR. A formulation of oral azacitidine-cedazuridine that would be 100% absorbed is under development as an alternative to subcutaneous/IV azacitidine. Current studies are investigating combination of these agents e.g. oral decitabine- cedazuridine plus venetoclax) and longer dose schedules (mimicking the decitabine 10-day schedule) in AML.

Maintenance therapy may also be beneficial in the post-SCT setting. In the SORMAIN trial, 83 patients with FLT3-ITD AML post-allogeneic SCT were randomized to sorafenib 200 mg-400 mg twice daily versus placebo for two years. The two-year PFS rate was 85% with sorafenib versus 53% with placebo (p=0.04). Survival was longer with sorafenib (hazard ratio 0.447; p=0.03). ⁽¹⁴⁸⁾ In the pivotal (ADMIRAL) trial that evaluated gilteritinib versus salvage chemotherapy in FLT3-mutated AML (detailed earlier), 51 patients achieving a response post gilteritinib and undergoing allogeneic SCT either resumed gilteritinib post SCT (n=35) or did not (n=16). The median survival was longer with gilteritinib resumption (16.2 months versus 8.4 months; hazard ratio 0.387; p=0.024). ⁽¹⁴⁰⁾ The ongoing randomized RADIUS trial and BMT-CTN 1506 study are prospectively investigating post-SCT maintenance with midostaurin and gilteritinib, respectively, versus standard of care.

ALLOGENEIC AND AUTOLOGOUS STEM CELL TRANSPLANTATION

A meta-analysis of multiple randomized trials demonstrated, on average, a benefit of allogeneic SCT in AML in first CR. ⁽¹⁷⁴⁾ This has been questioned in previous randomized trials for several reasons. The limited number of patients in each study may have overlooked modest but clinically significant benefits with SCT. The lead-time bias to allogeneic SCT and the fact that many patients allocated to allogeneic SCT could not undergo the procedure for various reasons (infections, organ dysfunction, new chemotherapy-related morbidities, AML relapse, others) further confused the analyses. Finally, some patients allocated to chemotherapy in first CR may have benefited from allogeneic SCT in second CR. A MRC study reported that the benefits of chemotherapy versus allogeneic SCT in first CR were similar when the benefit of allogeneic SCT in second CR was considered. ⁽¹⁷⁵⁾

Today, allogeneic SCT is standard of care in first CR, and is considered based on several factors. These include the presence of an adverse karyotype or high FLT3-mutated AR at diagnosis; persistent MRD in CR; and a lower risk of SCT-associated mortality (based on favorable characteristics such as age, co-morbid conditions, suitability of donor and degree of matching).

With the availability of multiple effective targeted therapies in AML, allogeneic SCT should be considered as part of a total strategy of chemotherapy/targeted therapy/SCT/post-SCT maintenance in order to improve the AML cure rates further. Further investigations of post-SCT maintenance strategies to reduce the risk of relapse need to be considered in this continuum, including HMAs (both parenteral and oral), FLT3 and IDH inhibitors, venetoclax, and others.

Allogeneic SCT should be considered in patients with refractory-relapsed AML who achieve subsequent CR, and also may be the best salvage option in patients with refractory-relapsed AML who have persistent disease with less than 20% marrow blasts (long-term survival 10%-20%). ⁽¹⁷⁶⁾

Autologous SCT is still considered occasionally in the setting of APL and CBF AML in second CR, with negative molecular studies in collected stem cells. Otherwise, it has been largely abandoned in the United States for AML because of the perceived lack of benefit. Some European researchers consider autologous SCT in first CR based on randomized trials showing that it provides equivalent results to multiple chemotherapy consolidations (usually fewer than 4). With increasing knowledge about the negative impact of persistent MRD in CR, it is conceivable that with autologous SCT, host marrows may have been reinfused with significant persistent AML disease burden, thus perhaps causing relapses and negating its possible benefit. Future studies could evaluate again the benefit of autologous SCT using collected MRD-negative cells.