Signal Transduction and Targeted Therapy
`
`www.nature.com/sigtrans
`
`OPEN
`REVIEW ARTICLE
`Targeting multiple signaling pathways: the new approach
`to acute myeloid leukemia therapy
`Jenna L. Carter1,2, Katie Hege1, Jay Yang3,4, Hasini A. Kalpage 5, Yongwei Su3,4,6, Holly Edwards3,4, Maik Hüttemann5,
`Jeffrey W. Taub1,4,7,8 and Yubin Ge 1,3,4
`
`Acute myeloid leukemia (AML) is the most common form of acute leukemia in adults and the second most common form of acute
`leukemia in children. Despite this, very little improvement in survival rates has been achieved over the past few decades. This is
`partially due to the heterogeneity of AML and the need for more targeted therapeutics than the traditional cytotoxic
`chemotherapies that have been a mainstay in therapy for the past 50 years. In the past 20 years, research has been diversifying the
`approach to treating AML by investigating molecular pathways uniquely relevant to AML cell proliferation and survival. Here we
`review the development of novel therapeutics in targeting apoptosis, receptor tyrosine kinase (RTK) signaling, hedgehog (HH)
`pathway, mitochondrial function, DNA repair, and c-Myc signaling. There has been an impressive effort into better understanding
`the diversity of AML cell characteristics and here we highlight important preclinical studies that have supported therapeutic
`development and continue to promote new ways to target AML cells. In addition, we describe clinical investigations that have led
`to FDA approval of new targeted AML therapies and ongoing clinical trials of novel therapies targeting AML survival pathways. We
`also describe the complexity of targeting leukemia stem cells (LSCs) as an approach to addressing relapse and remission in AML
`and targetable pathways that are unique to LSC survival. This comprehensive review details what we currently understand about
`the signaling pathways that support AML cell survival and the exceptional ways in which we disrupt them.
`
`Signal Transduction and Targeted Therapy (2020)5:288
`
`; https://doi.org/10.1038/s41392-020-00361-x
`
`1234567890();,:
`
`INTRODUCTION
`Epidemiology of acute myeloid leukemia (AML)
`There are about 20,000 new cases of AML diagnosed each year in
`the United States.1,2 AML can affect people of all ages, but it is
`much more common in older adults with the age-adjusted
`incidence for those aged ≥65 years being 20.1 per 100,000 person-
`years compared with 2.0 per 100,000 person-years for those aged
`<65 years. The median age at diagnosis is 68 years and is most
`frequently diagnosed among people aged 65–74 years. Addition-
`ally,
`incidence is modestly increased in males compared with
`females and in Caucasians compared with other ethnic groups.3
`In most patients, the precise inciting event(s) leading to AML is
`unknown, but a genetic origin is strongly implicated. Environ-
`mental factors including exposure to chemicals such as benzene
`have also been associated with AML. Patients with a prior history
`of myelodysplastic syndromes (MDS) or myeloproliferative neo-
`plasms (MPN) and those who have previously received radiation
`and/or chemotherapy are also at increased risk of developing
`AML. Collectively, patients with AML with an antecedent myeloid
`disorder and those with therapy-related AML are considered to
`have secondary AML. In large population-based studies, 25% of
`AML cases are considered secondary and these are associated
`
`with lower rates of response to therapy and an inferior prognosis
`compared with de novo AML.4 AML can also be seen in patients
`with inherited genetic syndromes such as Fanconi’s anemia,
`Bloom syndrome, Down syndrome, and others.5–7 There has also
`been an increased interest in the study of inherited predisposi-
`tions to myeloid malignancies such as those seen with familial
`mutations of CEBPA, DDX4, and RUNX1.8
`Premalignant evidence of clonal hematopoiesis can be found in
`healthy individuals, as evidenced by acquired mutations in genes
`such as DNMT3A, TET2, and ASXL1.9 The prevalence of such clonal
`hematopoiesis of indeterminate potential (CHIP) increases with
`age and is associated with an increased risk of developing a
`subsequent hematological malignancy, particularly MDS or AML.
`However, the large majority of people who have CHIP never
`develop a hematologic cancer.10
`
`AML pathobiology
`transformation of myeloid
`AML results from the malignant
`precursor cells that are driven by a number of acquired genetic
`abnormalities. The biology is complex and involves a number of
`interdependent genetic mechanisms and pathways. The two-hit
`model of leukemogenesis hypothesizes that many cases of AML
`
`1Cancer Biology Graduate Program, Wayne State University School of Medicine, Detroit, MI, USA; 2MD/PhD Program, Wayne State University School of Medicine, Detroit, MI, USA;
`3Department of Oncology, Wayne State University School of Medicine, Detroit, MI, USA; 4Molecular Therapeutics Program, Barbara Ann Karmanos Cancer Institute, Wayne State
`University School of Medicine, Detroit, MI, USA; 5Center for Molecular Medicine and Genetics, Wayne State University School of Medicine, Detroit, MI, USA; 6National Engineering
`Laboratory for AIDS Vaccine, Key Laboratory for Molecular Enzymology and Engineering, The Ministry of Education, School of Life Sciences, Jilin University, Changchun, China;
`7Division of Pediatric Hematology/Oncology, Children’s Hospital of Michigan, Detroit, MI, USA and 8Department of Pediatrics, Wayne State University School of Medicine, Detroit,
`MI, USA
`Correspondence: Jeffrey W. Taub (jtaub@med.wayne.edu) or Yubin Ge (gey@karmanos.org)
`These authors contributed equally: Jenna L. Carter, Katie Hege
`
`Received: 24 July 2020 Revised: 21 September 2020 Accepted: 23 September 2020
`
`© The Author(s) 2020
`
`CELGENE 2071
`APOTEX v. CELGENE
`IPR2023-00512
`
`

`

`Targeting multiple signaling pathways: the new approach to acute myeloid. . .
`Carter et al.
`
`2
`
`are the result of the cooperation of two types of mutations:
`mutations that result in unhindered cell proliferation (class I
`mutations such as FMS-like tyrosine kinase 3 (FLT3), NRAS, c-KIT)
`and mutations that
`result
`in the arrest of normal myeloid
`differentiation (class II mutations such as RUNX1-RUNX1T1, CEBPA,
`TP53).11 Although this two-hit model
`is a simplification of the
`biology of AML, it serves as a useful conceptual framework.
`Cytogenetic abnormalities can be found in 50–60% of AML
`cases and strongly correlate with prognosis.12 For example, the
`core-binding factor (CBF) leukemias with the balanced transloca-
`tions t(8;21)(q22;q22), inv(16)(p13;q22), and t(16;16)(p13;q22) are
`associated with a favorable prognosis while AML with deletion 7,
`a monosomal karyotype, or a complex karyotype are associated
`with a dismal prognosis.13 The overwhelming majority of AML
`cases are also typified by genetic mutations such as NPM1, FLT3,
`IDH2 and TP53.14 Together
`isocitrate dehydrogenase 1 (IDH1),
`these acquired cytogenetic and molecular abnormalities encode
`transcription factors, tumor suppressors, DNA repair proteins,
`signaling molecules, regulators of apoptosis, and other diverse
`mechanisms, which promote leukemia development.
`AML appears to be maintained by a pool of self-renewing
`leukemia stem cells (LSCs) that are typically characterized by a
`CD34+CD38−CD123+ immunophenotype.15 AML stem cells are
`highly resistant to chemotherapy since they are primarily in the G0
`phase of the cell cycle and preferentially express multiple drug-
`resistant proteins such as P-glycoprotein and Bcl-2.16,17 The
`persistence of these LSCs predisposes to relapse even if bulk
`AML cells are effectively eliminated by treatment. As discussed
`below, there is great interest in therapies targeting LSCs in order
`to prevent relapse.
`Patients with AML usually present with manifestations of
`cytopenias due to bone marrow (BM) failure such as fatigue, fevers,
`infection, or bleeding. AML can sometimes present with signs and
`symptoms of hyperleukocytosis, which usually affects the pulmon-
`ary and central nervous systems. Extramedullary collections of
`leukemic blasts (e.g., myeloid sarcoma) is much less common but
`can occur in almost any organ, with skin (aka. leukemia cutis) and
`lymph node involvement being the most common.18 The diagnosis
`of AML is typically made by documenting ≥20% myeloblasts in a
`BM biopsy or aspirate specimen using adjunctive tests such as flow
`cytometry, cytogenetics, and fluorescence in situ hybridization to
`confirm and categorize the leukemia.18 Molecular profiling of AML
`is now considered mandatory in the era of precision medicine.
`There are a number of platforms available but next-generation
`sequencing panels focused on the most critical and common gene
`mutations are now routinely being employed. At a minimum, the
`European LeukemiaNet recommends testing for FLT3-ITD, FLT3-TKD,
`TP53, NPM1, RUNX1, ASXL1, and CEBPA mutations.19 One could also
`argue that screening for IDH1 and IDH2 mutations should be
`considered essential particularly at the time of relapse due to the
`availability of IDH1 and IDH2 inhibitors.
`
`AML classification
`The original FAB (French–American–British) classification of AML
`was the first attempt to systematically categorize this disease and
`(FAB M0–M7)
`divided AML into groups
`largely based on
`morphology and a few histochemical stains. The modern World
`Health Organization (WHO) classification is based on a combina-
`tion of morphology,
`immunophenotype, clinical characteristics,
`and genetics with the goal of identifying distinct biologic entities
`of AML with defined molecular pathways.20 The WHO classifica-
`tion recognizes six major categories of AML:
`(a) AML with
`recurrent genetic abnormalities; (b) AML with myelodysplasia-
`related features; (c) therapy-related AML and MDS; (d) AML, not
`otherwise specified;
`(e) myeloid sarcoma; and (f) myeloid
`proliferations related to Down syndrome.
`There are currently 11 genetic subtypes of AML recognized in
`the WHO classification including t(8;21)(q22;q22), inv(16)(p13;q22),
`
`t(16;16)(p13;q22), and several others. AML with the following gene
`mutations have also been included: NPM1, CEBPA (biallelic), BCR-
`ABL1, and RUNX1. AML with NPM1 or biallelic CEBPA mutations are
`considered favorable while AML with RUNX1 mutations are
`unfavorable.21,22
`Although AML with FLT3 mutation is not included in the WHO
`classification as a distinct entity, it is the most commonly (~30% of
`AML) mutated gene in AML and its presence predicts an
`unfavorable prognosis.23 FLT3 internal tandem duplication (FLT3-
`ITD) mutations result in a constitutively active FLT3, a transmem-
`brane tyrosine kinase, which in turns results in the growth and
`proliferation of leukemia cells.24 Because of its association with
`high rates of
`relapse, allogeneic hematopoietic stem cell
`transplant (SCT)
`is generally recommended in first remission.
`FLT3-ITD mutations are also an example of the complex interplay
`of genetic abnormalities seen in AML and their diverse effects on
`outcomes. Many of these mutations are often found in the same
`patient. NPM1 mutations can often co-exist with FLT3-ITD
`mutations resulting in a genotype with an intermediate-risk
`prognosis, depending on the FLT3 allelic ratio.25
`About 5–10% of AML patients have acute promyelocytic
`leukemia (APL) with PML-RARA fusion gene. This is characterized
`by a reciprocal translocation between chromosomes 15 and 17 (t
`(15;17)(q24;q21)) resulting in the production of a PML-RARA fusion
`gene. APL remains the paradigm of the genetic classification and
`treatment of AML given its disease-defining molecular signature
`and excellent outcomes with targeted therapies. APL is clinically
`characterized by disseminated intravascular coagulation and
`hyperfibrinolysis, which can result in a potentially fatal hemor-
`rhagic diathesis. However,
`if managed promptly and appropri-
`ately, the majority of patients are cured with treatment regimens
`that
`include a combination of
`targeted biologic therapies
`including all-trans retinoic acid and arsenic trioxide.26 Due to the
`unique characteristics of APL with PML-RARA fusion gene, this
`entity is not specifically covered in the remainder of this review.
`
`Treatment of AML
`The standard treatment for newly diagnosed AML remained static
`for many decades and was divided into induction therapy and
`consolidation therapy (Fig. 1). The goals of induction therapy are
`achievement of a complete morphologic remission, which results
`in the restoration of normal hematopoiesis and allows for
`subsequent therapy that maximizes the probability of long-term
`remission and potentially a cure.
`A combination of a daunorubicin and cytarabine was introduced
`approximately half a century ago and remained the standard
`therapy for most patients until very recently (Fig. 1). The most
`common iteration of
`this combination consists of 7 days of
`infusional cytarabine and 3 days of daunorubicin, the so-called
`“7+3” regimen. Remission rates are reported between 30 and 80%
`depending on patient and disease-related factors but long-term
`survivals and cure rates are appreciably lower due to relapses. This
`intensive chemotherapy approach is accompanied by a number of
`potential complications,
`including prolonged marrow aplasia,
`profound cytopenias, need for transfusional support, and risks of
`neutropenic infection and sepsis. Mortality rates during induction
`
`Fig. 1 History of AML therapies. Timeline of approved clinical
`therapies in the United States for the treatment of AML
`
`Signal Transduction and Targeted Therapy (2020)5:288
`
`

`

`range from 5% reported in clinical trials of younger patients to >20%
`reported in analyses of real-world data from the SEER database.27,28
`In patients who achieve a complete remission (CR), the cure rate
`approaches zero in the absence of some form of post-remission
`therapy. The antimetabolite cytarabine has been the mainstay of
`consolidation therapy for decades. A higher dose of cytarabine
`such as HIDAC improves survival in younger patients and in those
`who carry favorable cytogenetics.29 SCT is often used as post-
`remission therapy since its application is associated with the
`lowest rates of leukemia relapse. This benefit is balanced by risks
`of transplant-related mortality, including age, pre-existing comor-
`bidities, and other molecular lesions, and by risks of long-term
`complications in the form of graft-vs-disease. Retrospective
`analyses have clearly shown a survival benefit in favor of SCT in
`patients with AML who have adverse-risk genetic abnormalities
`(intermediate- or poor-risk AML).30 The absence of randomized
`data makes definitive conclusions in the intermediate-risk patient
`population less than clear-cut; however, most experts would favor
`SCT in eligible patients in the absence of contraindications.
`Older patients (>65 years) with AML have a much poorer
`prognosis compared with their younger counterparts primarily due
`to an increased prevalence of adverse-risk genetic features that
`promote resistance to chemotherapy.31 Intensive chemotherapy in
`this population is associated with lower rates of remission, increased
`toxicities, higher rates of early mortality, and dismal
`long-term
`survival rates in the absence of transplantation.31,32 For this reason,
`less intensive therapies such as single-agent hypomethylating
`agents (HMAs) and low-dose cytarabine (LDAC) have been used
`prior to the development and recent approval of novel agents.
`HMAs (azacitidine and decitabine) are commonly used in the
`United States for the first-line treatment of AML. However, their use
`as a single-agent therapy for AML has always been met with a lack
`of enthusiasm given the modest activity of these drugs. HMAs are
`associated with a CR rate of approximately 17.8 and 27.8% for
`decitabine and azacitidine, respectively, a duration of remission of
`10.4 months (azacitidine), and an overall survival (OS) of 7.7 and
`10.4 months for decitabine and azacitidine, respectively, compared
`to no responses in supportive care patients.33,34 Overall results with
`LDAC are likely to be worse, though it has improved OS and higher
`rates of supportive care in elderly adults who cannot receive
`conventional therapy.35,36 However patients with adverse cytoge-
`netics have no benefit in remission or survival on LDAC.36 Single-
`agent HMA use has not brought clear improvements in survival in
`AML patients and limited Food and Drug Administration (FDA)
`this use.33,34 Fortunately, novel drugs such as
`approval
`for
`venetoclax, an inhibitor of the antiapoptotic protein Bcl-2, appear
`to markedly potentiate the activity of HMA or LDAC making
`venetoclax-based combinations a more attractive option for older
`patients.37,38 HMAs are continued to be used in combination
`therapies and will be highlighted extensively in this review.
`Unlike acute lymphoblastic leukemia (ALL), there has been no
`role for maintenance therapy in AML. However, maintenance
`with an oral form of azacitidine, CC-486, was recently reported to
`improve survival compared with placebo in patients who
`completed standard 7+3 induction and cytarabine consolida-
`tion.39 This therapy was just recently (September 2020) approved
`for maintenance therapy in adults with AML in first remission
`and is likely to be considered standard of care for these patients
`(Fig. 1). In the subset of patients with FLT3 mutations, tyrosine
`kinase inhibitors (TKIs) are often used as posttransplantation
`maintenance despite the lack of regulatory approval. Addition-
`ally, there are several clinical trials investigating new potential
`AML maintenance therapies
`(NCT01546038, NCT03564821,
`NCT03881735, NCT03515512, NCT03728335, NCT03932643).
`
`AML clinical outcomes
`The OS of all AML patients is around 30% but it is quite dependent
`on age: long-term survivals for patients less than and older than
`
`Signal Transduction and Targeted Therapy (2020)5:288
`
`Targeting multiple signaling pathways: the new approach to acute myeloid. . .
`Carter et al.
`
`3
`
`65 years is about 50% and 10%, respectively.40 Survival rates in
`children are about 60–65%. Age acts as a surrogate for a panoply
`of patient and disease-related factors.41 Patient variables include
`performance status, comorbidities, and impaired organ function.
`Disease-related factors include clinical characteristics and biologi-
`cal features, such as cytogenetics and genetic mutations. Patients
`can be divided into adverse-,
`intermediate-, and favorable-risk
`groups based on cytogenetics and molecular features; however,
`much heterogeneity exists even within these subgroups.19 There
`has been a consistent and substantial improvement in the survival
`of younger patients over the past several decades despite the lack
`of drug approvals during this time period.40 The improvement is
`likely due to better supportive care measures and implementation
`and improvement of the risk-adapted use of transplantation.
`Significant improvements in the elderly AML population have
`been much harder to come by. The adverse-risk biology of AML in
`older adults is associated with increased drug resistance leading
`to early mortality, shorter remissions, and inferior survivals. This
`has led to a therapeutic nihilism in some treating physicians when
`faced with the prospect of treating older patients with either
`intensive chemotherapy, which is deemed too toxic, or less
`intensive therapies with less than desired effectiveness. Fortu-
`nately,
`recent drug approvals have improved the treatment
`landscape of AML, particularly in older patients, so that almost
`all newly diagnosed patients should be offered some form of
`therapy.
`
`Newly approved AML therapies
`For many years, treatment options for AML remained stagnant,
`with the standard 7+3 regimen of cytarabine and daunorubicin
`and SCT being the only option for patients (Fig. 1). The difficulty in
`developing new therapies for AML attributes mainly to myelo-
`suppression, which probably is the biggest hurdle overall
`in
`developing new drugs for this deadly disease. This is due to a
`significant amount of overlap in processes and signaling between
`AML cells and normal hematopoietic cells.42 Fortunately, the past
`few years have seen the development of more targeted therapies
`to better address the pathobiology and heterogeneity of AML and
`includes the FDA approval of midostaurin, gemtuzumab ozoga-
`micin, CPX-351, enasidenib, ivosidenib, gilteritinib, venetoclax, and
`glasdegib (Fig. 1). This surge in approved AML therapies has
`diversified the treatment options for patients and marks a turning
`point in how AML is approached.
`
`FLT3 inhibitors: midostaurin and gilteritinib. As mentioned, FLT3
`mutations are common in AML, occurring in approximately 30% of
`patients, and can be due to ITD mutations or point mutations in
`the tyrosine kinase domain (TKD).43 Constitutive activation of the
`mutant FLT3 supports tumorigenesis in hematopoietic precursor
`cells, so inhibition of FLT3 has been heavily pursued as a targeted
`therapeutic option for these patients.43 Midostaurin and gilter-
`itinib are type I FLT3 inhibitors and are effective against FLT3-ITD
`and FLT3-TKD.44 Midostaurin was approved by the FDA for
`therapy in adult patients with newly diagnosed FLT3-mutated AML
`in April 2017 and was followed by the approval of gilteritinib for
`the treatment of adult patients with relapsed/refractory (R/R) AML
`with FLT3 mutations in November 2018 (Fig. 1).45,46
`
`IDH1 and IDH2 inhibitors: enasidenib and ivosidenib. Mutations in
`IDH1 and IDH2 occur in about 20% of AML patients and are
`common in other malignancies, such as glioblastoma.47 These
`mutant enzymes have become promising targets for new
`therapies, as they promote tumorigenesis in cells through
`production of the oncometabolite, 2-hydroxyglutarate (2-HG).48
`Inhibition of the mutant IDH1 and IDH2 enzymes is able to reduce
`the production of 2-HG to normal physiologic levels, and this
`promotes the differentiation of leukemia cells.48 Enasidenib (AG-
`221), an IDH2 mutant inhibitor, and ivosidenib (AG-120), an IDH1
`
`

`

`Targeting multiple signaling pathways: the new approach to acute myeloid. . .
`Carter et al.
`
`4
`
`mutant inhibitor, were approved for treatment of R/R AML in
`August 2017 and July 2018, respectively (Fig. 1).49,50
`
`Bcl-2 inhibitor: venetoclax. AML has been shown to be dependent
`on dysregulations of the apoptotic pathway,
`including over-
`expression of Bcl-2, which is an important antiapoptotic protein.51
`This has supported the development of Bcl-2 inhibitors to
`promote the induction of apoptosis in AML cells and has led to
`the discovery of venetoclax, a potent and selective Bcl-2
`inhibitor.52,53 After being approved for use in chronic lymphocytic
`leukemia (CLL) (2016), demonstrating promising results in early
`clinical trials, and being well tolerated in older patients, venetoclax
`was FDA approved in November 2018 for the treatment of AML
`(Fig. 1).37,54,55 It was approved for use in combination with
`azacitidine or decitabine or LDAC for the treatment of newly
`diagnosed AML patients aged ≥75 years or who have comorbid-
`ities that preclude the use of intensive induction chemotherapy.
`Currently, venetoclax is being studied in numerous clinical trials in
`combination and single-agent therapies.
`
`Hedgehog (HH) pathway inhibition: glasdegib. Aberrant activation
`of the HH signaling pathway has been shown to be increased in
`AML cells and has also been correlated with worse prognosis
`and drug resistance in AML.56,57 Numerous studies have demon-
`strated that targeting this pathway showed antitumor activity and
`combination with current therapies improves efficacy.56 This led
`to the development of glasdegib, a HH pathway inhibitor that
`works by binding to and inhibiting the transmembrane protein
`Smoothened (SMO).58 Glasdegib was FDA approved in November
`2018 for use in combination with LDAC for the treatment of newly
`diagnosed AML patients aged ≥75 years or who have comorbid-
`ities that preclude the use of intensive induction chemotherapy
`(Fig. 1).59
`
`Antibody–drug conjugate (ADC): gemtuzumab ozogamicin. CD33
`has been found to be highly expressed on the leukemia cells of
`most AML patients and has developed as a targetable antigen.60
`Gemtuzumab ozogamicin is an ADC of a CD33-directed huma-
`nized monoclonal antibody (mAb) that is covalently linked to N-
`acetyl gamma calicheamicin (cytotoxic drug). The antibody
`portion localizes to CD33 antigens found on myeloid leukemia
`blasts and calicheamicin is internalized and induces double-strand
`breaks in DNA and cell death. Gemtuzumab ozogamicin was
`originally approved for monotherapy in CD33+ AML patients
`aged ≥60 years in May 2000 but was withdrawn from the market
`in 2010 due to concerns about toxicity. Additional studies using
`gemtuzumab ozogamicin at lower doses in combination with
`currently approved therapies confirmed its safety. This led to the
`approval of gemtuzumab ozogamicin in September 2017 for the
`treatment of CD33-positive AML in adults and pediatric patients
`aged ≥2 years (Fig. 1).61
`
`Cytotoxic therapy: CPX-351. CPX-351 is a liposomal formulation of
`cytarabine and daunorubicin, two standard of care chemotherapy
`drugs used in the treatment of AML. CPX-351 utilizes liposomal-
`encapsulated delivery system to avoid the first-pass metabolism,
`enhancing the pharmacodynamics and pharmacokinetics (PK) of the
`drugs and potentially leading to greater efficacy.62 Given its success
`in clinical trials in improving patient response rate and survival, CPX-
`351 was FDA approved in August 2017 for the treatment of adults
`with newly diagnosed AML with myelodysplasia-related changes or
`therapy-related AML (Fig. 1).62,63
`
`TARGETING SIGNALING PATHWAYS IN AML
`Apoptotic pathways
`Apoptosis is essential for ensuring the homeostasis of healthy
`tissue via two highly regulated pathways—the mitochondrial
`
`(intrinsic) pathway and death receptor (DR; extrinsic) pathway. The
`extrinsic pathway of apoptosis is initiated by DRs or tumor
`necrosis factor (TNF)
`family receptors, which are cell surface
`receptors that are activated by ligand interactions. Activation of
`these receptors by external stimuli results in the recruitment and
`activation of caspase 8 and ultimately leads to cell death. The
`intrinsic pathway is usually initiated in a cell-autonomous way
`when a cell undergoes stress and is unable to repair damages. The
`intrinsic apoptosis pathway is governed by the Bcl-2 family
`proteins, which consists of highly regulated proapoptotic and
`antiapoptotic proteins. Proapoptotic effectors Bak and Bax are
`necessary to initiate apoptosis in the cell and, when activated, will
`form pores on the outer membrane of mitochondria. This process
`is referred to as mitochondrial outer membrane permeabilization
`and results in release of cytochrome c, a proapoptotic factor, from
`the mitochondrial
`intermembrane space into the cytosol. Cyto-
`chrome c promotes assembly of the apoptosome and activation of
`caspase 9—ultimately leading to cell death. Antiapoptotic
`proteins (e.g., Bcl-2, Bcl-xL, and myeloid cell leukemia 1 (Mcl-1))
`bind, sequester, and inhibit proapoptotic proteins (Bak/Bax) to
`prevent apoptosis. BH3-only proteins (e.g., Bid, Bim, Bad, and
`Noxa) enhance apoptotic activity through activation of proapop-
`totic effectors or neutralization of antiapoptotic proteins. The
`imbalance of
`interactions between these proapoptotic and
`antiapoptotic proteins control the caspase cascade and cell death
`(Fig. 2a).64
`Apoptosis is frequently dysregulated in cancer. Cancer cells
`evade apoptosis through different mechanisms, but many involve
`overexpression of antiapoptotic proteins (Bcl-2, Mcl-1, and Bcl-xL)
`or loss of expression of proapoptotic proteins (Bak/Bax).65 These
`dysregulations promote cancer cell survival during treatment as
`many cancer therapies are dependent on intrinsic apoptosis
`induction to promote cancer cell death.51,65,66 It has been well
`established in preclinical studies that both Bcl-2 and Mcl-1 are
`frequently overexpressed in AML cells.67 Overexpression of Bcl-2
`and/or Mcl-1 is related to poor prognoses and are also associated
`with chemotherapy resistance. This understanding has led to the
`development of many therapeutics to target the dysregulated
`intrinsic apoptotic pathway for the treatment of AML, and here we
`summarize the targeting of Bcl-2 and Mcl-1.
`
`The antiapoptotic protein Bcl-2 was found to be
`Bcl-2 inhibitors.
`overexpressed and critical for leukemogenesis of AML.67 It is
`frequently overexpressed in leukemia progenitor and stem cells
`and in myeloid leukemia blasts in comparison to normal
`hematopoietic cells.68 True to its nature of preventing apoptosis,
`Bcl-2 has also been shown to play a role in resistance to
`chemotherapy of AML.69 Based on the importance of Bcl-2, many
`pharmaceutical companies have developed methods to target Bcl-
`2 for the treatment of AML.
`Oblimersen is a single-stranded anti-sense oligonucleotide
`(ASO) that is complementary to the first six codons of the open
`reading frame of the Bcl-2 mRNA sequence. The binding of
`oblimersen to the Bcl-2 mRNA targets the duplex for cleavage by
`RNase H and prevents translation into the Bcl-2 protein (Fig. 2c).70
`In preclinical studies, oblimersen was shown to decrease Bcl-2
`mRNA to nearly undetectable ranges and showed promising
`efficacy in in vivo leukemia xenograft models.70,71 Following the
`promising preclinical data, oblimersen was the first drug targeting
`Bcl-2 to be used for clinical trials in AML. A phase I trial combined
`oblimersen with standard of care chemotherapy (cytarabine and
`daunorubicin) and showed that 48% of patients achieved CR, with
`a median survival of 12.6 months. BM samples from patients who
`achieved CR had higher expression levels of Bcl-2 prior
`to
`induction therapy and showed significant decreases in Bcl-2
`mRNA following treatment. Oblimersen had no additional
`toxicities compared to the known standard chemotherapy
`toxicities and no cardiac toxicities were noted.72 A phase III
`
`Signal Transduction and Targeted Therapy (2020)5:288
`
`

`

`Targeting multiple signaling pathways: the new approach to acute myeloid. . .
`Carter et al.
`
`5
`
`Fig. 2 Targeting antiapoptotic proteins induces apoptosis in AML. a Antiapoptotic proteins Mcl-1 and Bcl-2 bind and sequester apoptotic
`effector proteins Bak/Bax to prevent Bak/Bax oligomerization and subsequent induction of apoptosis. b BH3 mimetics bind to the BH3-
`binding site of antiapoptotic proteins, Bcl-2 and Mcl-1, and release Bax/Bak to promote oligomerization and MOMP, which leads to
`subsequent induction of apoptosis. c Oblimersen is an anti-sense oligonucleotide that binds specifically to Bcl-2 mRNA to prevent Bcl-2
`translation and promote AML cell apoptosis. Selinexor (KPT330) inhibits XPO1 expression and subsequently decreases Mcl-1 stability to
`promote AML cell apoptosis. CDK9 inhibitors prevent the transcription of Mcl-1 gene to promote apoptosis of AML cells
`
`intensive induction and consolidation
`randomized trial of
`chemotherapy ± oblimersen was performed in 503 untreated
`older AML patients (>60 years).73 The study found that the
`addition of oblimersen to induction and consolidation therapy
`resulted in no significant improvement in CR rates, OS, or disease-
`free survival. These results limited further study of oblimersen in
`AML, but investigation into Bcl-2 inhibition continued.
`With a greater understanding of the intrinsic apoptotic pathway
`came a new method of targeting Bcl-2 through the development
`of “BH3 mimetic” small-molecule inhibitors. As mentioned above,
`BH3 only proteins (BIM, PUMA, BIC, NOXA, BID, etc.) activate
`apoptosis by binding to the BH3-binding site of antiapoptotic
`proteins and activating BAX/BAK to oligomerize in mitochondrial
`outer membrane. The identification of this function led to the
`development of small molecules that could mimic BH3 only
`proteins and have been thus termed BH3 mimetics (Fig. 2b).66 The
`effort to develop these small molecules have been immense and
`has now resulted in seven BH3 mimetics entering clinical trials and
`the FDA approval of the Bcl-2 inhibitor, venetoclax (Fig. 1).66
`The first BH3 mimetic tested in clinical trial of AML was
`obatoclax, which was
`thought
`to inhibit Bcl-2 and other
`antiapoptotic proteins (Bcl-xL, Bcl-w, Bcl-b, A1, and Mcl-1)
`(Fig. 2b).74 In preclinical studies, obatoclax was shown to inhibit
`cell growth and induce apoptosis in both AML cell
`lines and
`primary patient samples75 and was able to potentiate the
`cytotoxic effect of cytarabine in AML cells.76 A phase I study of
`obatoclax included patients with myelodysplasia or refractory
`leukemia, including AML, ALL, and CLL. This study looked at the
`efficacy of obatoclax via continuous intravenous (IV) infusion at
`increasing doses and frequencies.77 Obatoclax was well tolerated,
`but only one AML patient achieved CR and there was no
`improvement in disease progression. Another phase I/II trial of
`single agent obatoclax in untreated older patients (≥70 years) with
`AML was performed t