In the Pipeline: Emerging Therapy for Acute Myeloid Leukaemia

Abstract

Acute myeloid leukaemia (AML) is an aggressive, heterogenous, and age-related haematological malignancy with dismal prognosis. Conventional therapy for AML consists of frontline induction therapy with cytarabine infusion for 7 days and administration of anthracyclines, most commonly daunorubicin, for 3 days (7 + 3), followed by subsequent consolidation with chemotherapy or allogeneic haematopoietic stem cell transplant (HSCT) for high-risk disease. However, the age-related nature of AML implies that a significant portion of patients are unfit for such intensive regimens and can only be put on palliative treatment. Increasing emphasis is being put on maximizing specificities and potencies of novel agents while minimizing treatment-related toxicities, entailing a future of personalized-therapy in AML. This chapter reviews recently approved agents and agents still in the pipeline for the treatment of AML both in the frontline and the relapsed/refractory setting.

16.1 Introduction

Acute myeloid leukaemia (AML) is an aggressive, heterogenous, and age-related haematological malignancy with dismal prognosis. Conventional therapy for AML consists of frontline induction therapy with cytarabine infusion for 7 days and administration of anthracyclines, most commonly daunorubicin, for 3 days (7 + 3), followed by subsequent consolidation with chemotherapy or allogeneic haematopoietic stem cell transplant (HSCT) for high-risk disease. However, the age-related nature of AML implies that a significant portion of patients are unfit for such intensive regimens and can only be put on palliative treatment. Although treatment options for AML have remained stagnant for a long time, exciting progress has been made during recent years, with the U.S. Food and Drug Administration (FDA) approving nine novel agents indicated for this disease (Table 16.1). Increasing emphasis is being put on maximizing specificities and potencies of novel agents while minimizing treatment-related toxicities, entailing a future of personalized-therapy in AML.

Table 16.1 Summary of recently approved agents in acute myeloid leukaemia (AML)

16.2 Novel Chemotherapeutic Formulations

16.2.1 CPX-351

CPX-351 (Vyxeos) is an FDA-approved liposomal formulation of daunorubicin and cytarabine in a 5:1 molar ratio. While the combination of cytarabine and anthracyclines (7 + 3) has long been the conventional treatment for AML, their administration in the form of a liposomal capsule significantly prolongs their half-life and efficacy [1].

After encouraging results in a phase I study, subsequent phase II and III trials were carried out [2]. In a phase II study comparing CPX-351 with 7 + 3 in newly diagnosed AML patients, remarkable clinical benefit of CPX-351 was demonstrated, especially among patients with secondary AML, which is associated with poor prognosis [3]. In another phase II trial among relapsed or refractory (r/r) patients, CPX-351 induced superior responses when compared to standard salvage chemotherapy [4]. In a phase III trial, CPX-351 showed significantly prolonged survival compared to 7 + 3 induction [5]. These promising results were consistently replicated in subsequent trials [6, 7]. Side effects of CPX-351 are generally similar to those of 7 + 3, including myelosuppression, cardiotoxicity, with the exception of slower recoveries of neutrophil and platelet counts [3,4,5]. Combination of vyxeos with gemtuzumab ozogamicin (GO) and FLT3 inhibitors (quizartinib, midostaurin) also demonstrated clinical and preclinical efficacy respectively [8, 9].

Trials of CPX-351 as monotherapy or in combination with Ivosidenib, enasidenib, venetoclax, gilteritinib, midostaurin, quizartinib, palbociclib, glasdegib, GO, or fludarabine are underway (NCT04230239, NCT03988205, NCT03629171, NCT04668885, NCT04269213, NCT03555955, NCT04049539, NCT04493164, NCT03825796, NCT04075747, NCT04209725, NCT04038437, NCT03826992, NCT04293562, NCT04128748, NCT03844997, NCT04231851, NCT03878927, NCT03904251, NCT03672539, NCT02272478, NCT04425655). Studies comparing CPX-351 with other intensive chemotherapy regimens are also ongoing (NCT03897127, NCT04061239, NCT04293562, NCT04195945, NCT04802161).

16.3 Targeting Tyrosine Kinases

Tyrosine kinases regulate a wide range of cellular pathways and are crucial to signal transduction. Their aberrant activities can contribute to leukaemogenesis via promoting proliferation, impeding differentiation, and inhibiting apoptosis. Therefore, various agents have been developed against these kinases for the treatment of AML (Figs. 16.1, 16.2, and 16.3).

Fig. 16.1

Downstream events of aberrant FLT3 signalling. AKT protein kinase B, ERK extracellular-signal-regulated kinase, FLT3-ITD FLT3 internal tandem duplication, FLT3-TKD FLT3 tyrosine kinase domain mutations, FLT3 Fms-like tyrosine kinase 3, MEK mitogen-activated protein kinase kinase, mTOR mammalian target of rapamycin complex, PI3K phosphoinositide 3-kinase, RAF rapidly accelerated fibrosarcoma, Ras rat sarcoma viral oncogene homolog, STAT5 signal transducer and activator of transcription 5

Fig. 16.2

Actions of type I and type II FLT3 inhibitors. FL FLT3 ligand, FLT3 Fms-like tyrosine kinase 3

Fig. 16.3

Agents targeting tyrosine kinases. AXL anexelekto, BTK Bruton tyrosine kinase, c-KIT cluster of differentiation 117, c-MET mesenchymal-epithelial transition factor, FL FLT3 ligand, FLT3 Fms-like tyrosine kinase 3, HGF hepatocyte growth factor, SCF stem cell factor, SFK Src family kinases, SYK spleen-associated tyrosine kinase

16.3.1 FLT3 Inhibitors

Figures 16.1, 16.2, and 16.3 and Table 16.2 summarize the role of FLT3 inhibitors and the major FLT3 inhibitors in development. Readers should refer to Chap. 12 of this title for further discussion.

Table 16.2 Summary of the characteristics of major novel FLT3 inhibitors [10, 11]

16.3.2 c-KIT Inhibitors

c-KIT, also known as CD117, is an RTK expressed in haematopoietic cells for their normal development. Upon binding of stem cell factor (SCF), c-KIT dimerizes and undergoes autophosphorylation, which activates downstream PI3K/AKT/mTOR, JAK-STAT, and Ras/RAF/MAPK pathways, as well as Src family kinases (SFKs) [12, 13]. The expression of c-KIT is found in 60–80% of AML and its mutation is especially prevalent in core binding factor (CBF) AML [14]. Mutations in c-KIT mainly occur in exon 8 and exon 17, with the latter being associated with a more inferior clinical outcome [12]. Aberrant activation of c-KIT results in increased proliferation, reduced apoptosis, and subsequent leukaemogenesis [12].

Dasatinib and radotinib are multi-kinase inhibitors with potent activity against c-KIT. These agents induced apoptosis in c-KIT-positive AML cell lines and showed activity in downregulating other leukaemogenic pathways in various preclinical studies [15]. Dasatinib also showed synergistic efficacy with navitoclax against AML cells with NUP98-NSD1 and FLT3-ITD [16]. Addition of dasatinib to standard chemotherapy and its use as single agent maintenance therapy in patients with CBF AML showed favourable outcomes and a tolerable safety profile [17,18,19]. A phase III randomized controlled trial of chemotherapy with or without dasatinib in CBF AML patients is underway (NCT02013648). Other c-KIT inhibitors which are not actively evaluated for use in AML include imatinib, SU5416, and SU6668 [20, 21].

16.3.3 AXL Inhibitors

Anexelekto (AXL) is a member of the TYRO3, AXL, and MER (TAM) RTK family [22]. It is expressed on a multitude of cells and tissues and is crucial for the normal function of various haematopoietic cell types [22, 23]. Binding of Gas6 to AXL induces its dimerization and subsequent activation of PI3K, Ras, Src, and JAK/STAT pathways, resulting in cellular proliferation and migration [22]. In AML, AXL may be activated via mechanisms independent of Gas6 [22]. Aberrant signalling of AXL also acts as a key mediator of resistance against FLT3 inhibitors [23].

Similar to FLT3 inhibitors, AXL inhibitors are divided into two types. Type I AXL inhibitors bind to the ATP-binding site of the active AXL receptor [22]. Bemcentinib (BGB324) is a highly specific, potent, and safe small molecule type I inhibitor of AXL which showed efficacy against both FLT3-WT and FLT3-mutant AML cell lines [24, 25]. Due to promising results of its combination with LDAC in recent trials, bemcentinib has received fast track designation from the FDA [26, 27]. A phase II study regarding the use of bemcentinib in AML is currently underway (NCT03824080). Other type I inhibitors include gilteritinib and sunitinib. Type II inhibitors bind to the AXL receptor in its inactive form [22]. Among them, merestinib (LY2801653) is a potent and orally available inhibitor of AXL, FLT3, MNK, MET/RON, and other oncoproteins [28, 29]. It was proven to be safe in r/r AML patients in a phase I clinical trial [30]. Other novel AXL inhibitors with impressive preclinical efficacies against AML cell lines include the AXL/Mer dual inhibitors ONO-9330547 and ONO-7475, with ONO-7475 currently in a phase I/II trial as monotherapy or in combination with venetoclax (NCT03176277) [31,32,33].

16.3.4 c-MET Inhibitors

The MET RTK family consists of two members, c-Met and RON. Upon binding of their respective ligands (HGF for c-Met, MSP for RON), their tyrosine kinase domain activates and initiates signal transduction via PI3K, AKT, Β-catenin, Ras/MAPK, and JAK/STAT pathways [34]. Evidence of their expression in AML blasts led to studies evaluating their potential roles as therapeutic targets [34].

SU11274 is a c-Met inhibitor which demonstrated anti-leukaemic efficacy in preclinical studies [34,35,36]. Crizotinib also exhibited activity against AML cells, but seemed to induce resistance via a compensatory increase in HGF expression [35].

16.3.5 SYK Inhibitors

Spleen-associated tyrosine kinase (SYK) is a cytoplasmic tyrosine kinase with diverse biological activities, including roles in adaptive immune receptors signalling [37]. In AML, its increased expression is shown to be associated with inferior clinical outcomes [38]. Upon activation by FLT3-ITD or other upstream pathways, SYK undergoes phosphorylation and initiates a series of downstream signalling pathways, ultimately contributing to leukaemogenesis [39].

Preclinical studies with R406, an active metabolite of the SYK inhibitor fostamatinib, showed efficacy against AML cell lines by inducing differentiation and inhibiting proliferation [39]. Entospletinib (GS-9973) showed efficacy as monotherapy as well as in combination with chemotherapy in two early phase trials [40, 41]. TAK-659, a dual inhibitor of SYK and FLT3, also exhibited anti-leukemic activity in murine models and showed promising efficacy and safety profile in a phase Ib/II study in r/r AML patients [42, 43].

16.3.6 BTK Inhibitors

Bruton tyrosine kinase (BTK) is a member of the Tec family kinases [44]. This family of non-receptor, cytoplasmic kinases are mainly expressed on the surfaces of haematopoietic stem cells (HSCs) and other haematopoietic cells [44]. BTK, in particular, also plays a critical role in the development of B lymphocytes and is considered to be a key mediator in B-cell neoplasms [45, 46]. In AML, aberrant signalling of SFK, SYK, and PI3K leads to BTK activation and downstream activation of NFKB and other kinase pathways, resulting in leukaemogenesis. Emerging evidence of high BTK expression and constitutive activation in AML cells has led to interests on its potential role as a therapeutic target [47, 48]. In addition, FLT3-ITD may act as one of the upstream events leading to BTK autophosphorylation, implying the potential of BTK inhibitors for treating FLT3-ITD-positive AML [47].

Ibrutinib (CI-32765) is an irreversible inhibitor of BTK. In preclinical studies, it showed efficacy against AML cell lines by inhibiting downstream NFKB signalling, SDF1/CXCR4-mediated migration, and SDF1-induced activation of the AKT/MAPK pathway [48, 49]. Mutations in FLT3, NPM1, and DNMT3A were shown to be associated with increased sensitivity to ibrutinib [50]. In leukaemic blasts obtained from c-KIT-positive AML patients, ibrutinib also inhibited activation of BTK by c-KIT and their adhesion to bone marrow stromal cells [51]. Furthermore, specific inhibition of FLT3-ITD by ibrutinib in leukaemic cell lines has been reported, supporting the hypothesis that BTK inhibition may be efficacious against FLT3-ITD-positive AML [52]. Its combination with the recently approved Bcl-2 inhibitor, venetoclax, also showed promising results in preclinical studies [53]. However, a phase II clinical trial with ibrutinib monotherapy or in combination of azacitidine or cytarabine showed limited efficacy [54].

CG-806 is a dual FLT3/BTK inhibitor with remarkable activity and safety against AML cell lines and murine models [55]. This agent is currently evaluated in a phase I clinical trial (NCT04477291). Other novel BTK inhibitors with promising preclinical results include ARQ351 and abivertinib (AC0010) [56,57,58].

16.3.7 SFK Inhibitors

The non-receptor Src family of kinases include LYN, HCK, BLK, FGR, FYN, LCK, SRC, and YES [59]. In AML cells, FYN, LYN, HCK, and FGR are commonly expressed. Aberrant upstream signalling of FLT3, c-KIT, and other RTKs result in their activation, subsequently causing STAT5, Ras, and PI3K induction [59].

Bosutinib is an SFK inhibitor primarily used in the treatment of chronic myeloid leukaemia (CML). Recently, studies showed that its combination with all-trans retinoic acid (ATRA) enhances sensitivity of AML cell lines to ATRA, thus promoting differentiation of AML blasts [60]. It will be evaluated in a subsequent phase Ib trial in combination with glasdegib (NCT04655391).

RK-20449 is a selective HCK inhibitor which showed efficacy against chemotherapy-resistant AML cells in murine models [61]. An FGR inhibitor, TL02–59, also showed anti-leukaemic activity in a preclinical study [62]. Other SFK inhibitors with impressive preclinical evidences include PP2, dasatinib, ponatinib, PD180970, and SKI-606 [59, 63]. Although SAR103168 showed efficacy against AML cell lines in preclinical studies, results from a subsequent phase I trial were disappointing [64, 65].

16.4 Targeting the Hedgehog Pathway

The hedgehog (Hh) pathway is an essential mediator of embryonic development. In the canonical Hh pathway, Hh ligand binds to the transmembrane protein Patched (PTCH) to alleviate its inhibition on Smoothened (Smo), another transmembrane protein (Fig. 16.4). Smo then activates downstream glioma transcription factors (GLI) to stimulate gene transcription and proliferation [66]. In the non-canonical Hh pathway, activation of GLI is induced by other upstream pathways instead of Smo activation, such as PI3K/AKT/mTOR, RAS/RAF/MEK/ERK, protein kinase C (PKC), and many others [67]. In AML, the Hh pathway and oncogenic GLI activity may be constitutively activated, which is associated with radio- and chemo-resistance as well as poor prognosis [68,69,70]. Notably, crosstalk between the Hh pathway and FLT3-ITD has been discovered, prompting contemplations on its therapeutic role in FLT3-mutant AML [67].

Fig. 16.4

Agents targeting the hedgehog pathway. AKT protein kinase B, ERK mitogen-activated protein kinase, GLI glioma transcription factors, Hh hedgehog, MEK mitogen-activated protein kinase kinase, mTOR mammalian target of rapamycin, PI3K phosphoinositide 3-kinase, PTCH patched, RAF rapidly accelerated fibrosarcoma, Ras rat sarcoma, Smo smoothened

16.4.1 Smo Inhibitors

Inhibition of Smo is the most widely studied among all potential therapeutic targets in the Hh pathway. Glasdegib (PF-04449913) is an FDA-approved selective Smo inhibitor. After encouraging results from preclinical studies, glasdegib was further studied in phase I clinical trials, where it was proven to be effective and tolerable in AML patients [71,72,73]. Subsequent trials investigating combinations of glasdegib with LDAC, decitabine, or standard chemotherapy all demonstrated clinical effectiveness [74,75,76,77]. Notably, addition of glasdegib to LDAC prolonged survival by nearly twofold compared to single agent LDAC, but did not increase toxicity in a multi-centre randomized phase II trial [76, 77]. Major side effects of glasdegib include febrile neutropenia, anaemia, and gastrointestinal disturbances [72,73,74,75,76,77]. Future trials include combination therapies with chemotherapy, LDAC, CPX-351, decitabine, azacitidine, GO, gilteritinib, ivosidenib, enasidenib, venetoclax, bosutinib, avelumab, and OX40 (NCT0341617, NCT02038777, NCT04231851, NCT04051996, NCT02367456, NCT04093505, NCT04655391, NCT03390296).

Sonidegib (LDE225) is another Smo inhibitor which demonstrated efficacy against doxorubicin-resistant AML cell lines and exhibited synergism with azacitidine in preclinical studies [78]. Its single agent therapy and combination with azacitidine or decitabine have been studied in phase I and phase II trials (NCT02129101, NCT01826214) [79].

Vismodegib (GDC-0449) also showed anti-leukaemic activity in preclinical studies, but had limited efficacy as monotherapy in a subsequent trial [80, 81]. Similarly, another trial of its use in combination with cytarabine was terminated due to the minimal responses observed among patients (NCT01880437).

16.4.2 GLI Inhibitors

Given that GLI activation can occur independently of Smo, direct inhibition of GLI is an attractive strategy against resistance to Smo inhibitors [67]. GANT61 is a GLI inhibitor which inhibited proliferation and induced apoptosis of AML in preclinical studies [82]. Its combination with sunitinib also prolonged survival of FLT3-mutant mice [83]. These optimistic results warrant clinical studies for GLI inhibitors in AML patients.

16.5 Targeting Apoptotic Pathways

16.5.1 BCL-2 Family Inhibitors

The anti-apoptotic B-cell lymphoma 2 (BCL-2) family prevents cellular apoptosis via the inhibition of proapoptotic proteins, such as BAX and BAK. Examples of members include Bcl-2 (B-cell lymphoma 2), myeloid cell leukaemia sequence 1 (MCL-1), and B-cell lymphoma-extra-large (Bcl-xL) [84]. Their actions are counteracted by the pro-apoptotic subfamily of BCL-2 (Fig. 16.5). In AML, their overexpression has been identified in multiple studies, which implies their influence on impairing apoptosis and promoting survival of leukemic cells [84, 85].

Fig. 16.5

Agents targeting apoptotic pathways. BAK Bcl-2-antagonist/killer 1, BAX apoptosis regulator BAX, Bcl-2, B-cell lymphoma 2, Bcl-xL B-cell lymphoma extra-large, BIM Bcl-2-like protein 11, BOK Bcl-2-related ovarian killer, DR4/5 death receptor 4 or 5, MCL-1 myeloid cell leukaemia sequence 1, NOXA Phorbol-12-myristate-13-acetate-induced protein 1, PUMA p53 upregulated modulator of apoptosis, tBID truncated BID, TRAIL TNF-related apoptosis-inducing ligand

16.5.2 Bcl-2 Inhibitors

Bcl-2 inhibitors exert anti-leukaemic activity by mimicking the BH3 domain of the pro-apoptotic BCL-2 proteins and freeing them from the anti-apoptotic BCL-2 protein, which induces apoptosis [85].

Despite unsatisfactory results in early trials with oblimersen and obatoclax, efforts on investigation of Bcl-2 inhibition were persistent, which led to the development of venetoclax [85]. Venetoclax (ABT-199) is an FDA-approved, potent, and selective Bcl-2 inhibitor. Venetoclax was proven to be effective and tolerable in preclinical and clinical studies, both as monotherapy and in combination with HMAs (azacitidine, decitabine) or cytarabine, both in newly diagnosed and r/r patients [86,87,88,89,90,91,92,93,94,95]. Notably, combination of venetoclax with HMAs induced remarkable responses in a wide range of patients, including those with high-risk cytogenetic features and mutant-TP53 [95, 96]. Preclinical studies also elucidated their efficacies in targeting LSCs via inhibition of complex 2 of the ETC [97].

Full approval of venetoclax by the FDA was prompted by the phase III randomized placebo-controlled VIALE-A and VIALE-C trials, which evaluated the use of venetoclax in combination with azacitidine and LDAC, respectively. Both trials illustrated improvements of survival outcomes and remission rates upon the addition of venetoclax, along with tolerable increases in haematological toxicities [98, 99]. However, it should be noted that these benefits did not reach statistical significance in the VIALE-C study. Major side effects of venetoclax include febrile neutropenia and thrombocytopenia [98, 99].

Finally, multiple novel combinations with venetoclax are also being studied to overcome resistance. Among them, agents downregulating activity of MCL-1 are intensively evaluated owing to associations between MCL-1 upregulation and venetoclax resistance [85]. Multiple trials of venetoclax as monotherapy or in combination with other agents are ongoing (Table 16.3).

Table 16.3 Venetoclax in future clinical studies as monotherapy or in combination with other agents

Other Bcl-2 inhibitors currently engaged in clinical trials include VOB560, S 055746 (BCL201), S6548, and APG2575 (NCT04702425, NCT02920541, NCT03755154, NCT04501120).

16.5.2.1 Bcl-2/Bcl-xL Dual Inhibitors

ABT-737 demonstrated promising efficacy against AML cell lines in preclinical studies, but its clinical development has been limited by an unfavourable pharmacokinetic profile [85, 100]. A derivative of this agent, navitoclax (ABT-263), possesses superior pharmacokinetic properties, though its clinical investigation is still not of interest due to the major adverse effect of thrombocytopenia [85].

16.5.2.2 MCL-1 Inhibitors

MCL-1 is another attractive therapeutic target in AML due to its overexpression in AML and association with venetoclax-resistance. AZD5991 is an MCL-1 inhibitor which demonstrated synergistic actions with bortezomib against AML xenograft in a murine study and is currently evaluated in combination with venetoclax in r/r AML patients in a phase I/Ib/IIa trial (NCT03218683) [101]. AMG 176 and AM-8621 both showed single agent efficacy and synergistic activity with venetoclax, though only AMG 176 is selected for further clinical investigations as monotherapy and in combination with azacitidine or venetoclax given its superior pharmacokinetic profile (NCT02675452, NCT03797261) [102, 103]. Another agent, AMG 397, also showed favourable preclinical results and will be evaluated in r/r AML patients in a phase I trial (NCT03465540) [104]. In addition, S63845 demonstrated excellent anti-leukaemic efficacy as single agent and in combination with venetoclax, daunorubicin, or S55746 (Bcl-2 inhibitor) in preclinical studies [105,106,107]. A related agent, S64315, has been evaluated in AML patients in a phase I trial and will undergo further testing in combination with azacitidine, venetoclax, or VOB560 (NCT02979366, NCT04629443, NCT03672695, NCT04702425). Other MCL-1 inhibitors with preclinical efficacies against AML include Compound 42, VU661013, MIMI, and Cardone compound 9 [108,109,110,111].

16.5.3 TRAIL Inducers

TNF-related apoptosis-inducing ligand (TRAIL) induces p53-independent apoptosis upon binding to its cell surface receptors, namely death receptors (DR) 4 and 5 [112]. Imipridone compounds have been found to promote TRAIL transcription and expression, subsequently inducing apoptosis. Among them, ONC201 demonstrated potent anti-leukaemic effect against AML cells and LSCs, both as monotherapy and in combination with cytarabine or azacitidine [113,114,115]. Interestingly, its therapeutic activity relies on both the induction of TRAIL activity and stimulation of an integrated stress response (ISR) [113,114,115]. It is currently evaluated as monotherapy or in combination with LDAC, and as single agent post-HSCT maintenance in AML patients in phase I/II trials (NCT02392572, NCT03932643). ONC212, a more potent derivative of ONC201, exhibited single agent activity and synergism with venetoclax against AML cell lines and murine models [116, 117].

16.6 Targeting the TP53 Pathway

TP53 encodes the tumour suppressor p53 and is among the most commonly mutated genes in all human malignancies [118]. WT p53 promotes cell cycle arrest, inhibits proliferation, and induces cellular apoptosis upon cellular stress [119]. Its activity is counteracted by mouse double minute 2 (MDM2), an E3 ligase which induces proteasomal degradation of p53 with the aid of MDM4 (Fig. 16.6). In AML, TP53 mutations are associated with resistance to chemotherapeutic agents and dismal prognosis, which warrants the development of novel targeted therapies against this entity [120].

Fig. 16.6

Agents targeting the TP53 pathway. ATO arsenic trioxide, GSH glutathione, MDM2 mouse double minute 2, MQ methylene quinuclidinone, mut-p53 mutant p53, ROS reactive oxygen species, WT wild type

16.6.1 Mutant TP53 Inhibitors

APR-246, a methylated analogue of p53 reactivation and induction of massive apoptosis (PRIMA-1), is a pro-drug of methylene quinuclidinone. Upon conversion into its active form, APR-246 restores the active conformation of p53 and its ability to induce apoptosis and cell cycle arrest in leukemic cells [120, 121]. APR0246 can also exert anti-tumour effect in a p53-independent manner via depletion of anti-oxidants and induction of oxidative stress [122]. Synergism with azacitidine in inducing G0/G1 cell cycle arrest, apoptosis, and downregulation of FLT3 signalling was also reported [123]. This agent demonstrated remarkable clinical efficacy in combination with azacitidine in TP53-mutant AML patients in an ongoing phase 1b/2 study [124] and is being further investigated in other trials (NCT03072043, NCT03931291).

Arsenic trioxide (ATO), an agent primarily used for the treatment of acute promyelocytic leukaemia, also demonstrated ability to induce proteasomal degradation of mutant p53 and restore normal function of WT p53 [125, 126]. Its combination with ascorbic acid selectively induced oxidative stress and apoptosis in TP53-mutant leukemic cells in a recent study [127]. In addition, this agent exhibited activity against NPM1-mutant AML cells by inducing mutant protein degradation in multiple studies [128,129,130]. The use of ATO as single agent and in combination with decitabine or all-trans-retinoic-acid (ATRA) is currently explored in patients with TP53 or NPM1 mutations in a number of clinical studies (NCT04689815, NCT03855371, NCT03031249).

16.6.2 MDM2 Inhibitors

Increased activity of MDM2 is associated with reduced p53 activity [118]. Therefore, inhibition of binding between MDM2 and p53 prevents degradation of p53 and restores its tumour suppressor functions [131]. Nutlins are the earliest selective inhibitors of MDM2 to be discovered, with nutlin 3 being widely used in preclinical studies investigating effects of MDM2 inhibition [131]. A small molecule MDM2 inhibitor, RG7112, demonstrated anti-leukaemic efficacy as monotherapy and in combination with cytarabine in AML patients [132, 133]. Another agent, idasanutlin (RG7388), is a potent, selective, and orally available second generation MDM2 inhibitor. Clinical studies of this agent as monotherapy and in combination with cytarabine had impressive responses. This agent was generally tolerable with gastrointestinal toxicity as a significant side effect [134]. In addition, idasanutlin exhibited synergistic activity with venetoclax in a preclinical study, which led to the initiation of a phase 1/1b trial with favourable results [135, 136]. Combination of idasanutlin with venetoclax or chemotherapy will be further evaluated in a phase 1/2 clinical trial (NCT04029688). Synergism between idasanutlin and XPO inhibitors (selinexor, eltanexor) was also discovered in a preclinical study [137].

Disappointingly, RO6839921, the pegylated prodrug of idasanutlin, showed inferior effectiveness compared to idasanutlin in a recent study and will not undergo further clinical development [138].

Another MDM2 inhibitor, AMG 232 (KRT232), showed modest clinical activity in combination with trametinib, a MEK inhibitor. This combination regimen was tolerable and common adverse effects include nausea, gastrointestinal disturbances, and poor appetite [139]. This agent will be tested in combination with cytarabine and venetoclax; cytarabine; decitabine; or with TL-895 (TKI) in subsequent trials (NCT04190550, NCT04113616, NCT04669067).

Siremadlin (HDM201) showed promising activity in a phase I trial with cytopenias and tumour lysis syndrome as the most significant side effects. It will undergo evaluation with midostaurin in r/r patients with TP53 and FLT3 mutations, as well as with MBG453 (TIM3 inhibitor) or venetoclax in AML patients (NCT04496999, NCT03940352) [140]. Another MDM2 inhibitor, Milademetan (DS-3032b), has been evaluated as monotherapy in a phase 1 trial and is currently evaluated in combination with azacitidine, or LDAC with or without venetoclax (NCT03671564, NCT02319369, NCT03634228). Finally, APG-115 is currently evaluated with azacitidine or cytarabine in a phase 1 trial (NCT04275518).

16.7 Targeting the PI3K/AKT/mTOR Pathway

The phosphoinositide 3-kinase (PI3K)-Protein kinase B (AKT)-mammalian target of rapamycin (mTOR) pathway is crucial to cellular metabolism and can be activated by a myriad of upstream pathways [141]. In AML, upregulation of this pathway supports leukaemic cell activities and can occur as a result of aberrant upstream tyrosine kinases signalling or constitutive activation [141]. Unfortunately, increased activity of this pathway seems to be associated with decreased survival [141]. Thus, pharmacological inhibition of this pathway is a logical and attractive novel strategy in AML (Fig. 16.7).

Fig. 16.7

Agents targeting the PI3K/AKT/mTOR pathway. AKT protein kinase B, mTOR mammalian target of rapamycin, PDK 3-phosphoinositide-dependent protein kinase-1, PI3K phosphoinositide 3-kinase, TSC1 tuberous sclerosis complex 1, TSC2 tuberous sclerosis complex 2

Although PI3K/AKT/mTOR inhibition demonstrated anti-leukaemic efficacies in preclinical studies, these results did not translate into meaningful clinical benefits [141]. mTORC1 inhibitors, including sirolimus, everolimus (RAD001), deferolimus (AP23573, MK-8669), and temsirolimus (CCI-779), have been tested in multiple clinical trials as monotherapies or in combination with chemotherapy regimens among AML patients with mostly limited success [142,143,144,145,146,147]. Although dual inhibition of PI3K and mTORC1 was proposed as a mechanism against resistance to mTORC1 inhibitors [148], two dual PI3K/mTOR inhibitors, gedatolisib (PF-05212384) and BEZ235, did not improve patient survival as single-agent and as an adjunct to chemotherapy, respectively [149, 150]. Other strategies to overcome resistance, such as dual mTORC1/mTORC2 inhibition, are being explored for the treatment of AML [148].

16.8 Targeting Metabolic Pathways

Mitochondrial activity is fundamental to supporting cellular metabolisms of almost all types of body cells. This carries paramount significance for the treatment of AML due to the presence of mitochondrial abnormalities, which can be exploited for selective AML cells targeting [151] (Fig. 16.8). In addition, other aberrant metabolic pathways discovered in LSCs are also being explored as targets for LSC eradication [151].

Fig. 16.8

Agents targeting metabolic pathways. 2-HG 2-hydroxyglutarate, Acetyl-CoA acetyl-coenzyme A, ADP adenosine diphosphate, ATP adenosine triphosphate, CPT1α carnitine palmitoyl transferase 1a, FAD flavin adenine dinucleotide, FADH flavin adenine dinucleotide hydrogen, IDH1 isocitrate dehydrogenase 1, IDH1m mutant IDH1, IDH2 isocitrate dehydrogenase 2, IDH2m mutant IDH2, NAD nicotinamide adenine dinucleotide, NADH nicotinamide adenine dinucleotide hydrogen, NADPH nicotinamide adenine dinucleotide phosphate hydrogen, α-KG α-ketoglutarate

16.8.1 IDH1/2 Inhibitors

This role of IDH1/2 inhibition and development of IDH1/2 inhibitors are further discussed in Chap. 11.

16.8.2 Oxidative Phosphorylation Inhibitors

Leukaemic stem cells (LSCs) reply on oxidative phosphorylation (OXPHOS) for their metabolism rather than anaerobic glycolysis, which is the predominant metabolic pathway in normal HSCs [152]. Since integrity of the mitochondrial electron transport chain (ETC) is essential for OXPHOS, its inhibition can disrupt metabolic activities of LSCs. IACS-010759, an inhibitor of complex 1 of the ETC, demonstrated selective anti-leukemic activity as monotherapy and synergism with venetoclax and vinorelbine, a microtubule destabilizer, against AML cells and xenograft models while sparing normal haematopoietic cells [153,154,155]. Compared to its predecessor BAY 87–2243, IACS-010759 also has a superior safety profile [152]. It is currently being studied in r/r AML patients in a phase I trial (NCT02882321). Another ETC complex 1 inhibitor, mubritinib (TAK-165), also exhibited activity against AML cells in a preclinical study [156].

16.8.3 Fatty Acid Oxidation Inhibitors

Fatty acid oxidation (FAO) generates acetyl coenzyme A (Acetyl-CoA) for the TCA cycle, and ultimately, OXPHOS [152]. The rate limiting step in FAO is catalysed by carnitine palmitoyl transferase 1a (CPT1a), thus, inhibition of this enzyme selectively impedes metabolism of leukaemic stem cells [152]. ST1326 is a CPT1a inhibitor which induced growth arrest, mitochondrial disruption, and apoptosis in various leukaemic cell lines, with the highest activity towards AML cells [157].

16.9 Targeting the Proteasome

The proteasome is a multimeric protein complex which mediates degradation of ubiquitinated proteins (Fig. 16.9). It controls a wide range of cellular activities, including cell cycle progression and survival [158]. Aberrant activities of the proteasome contribute to leukaemogenesis through various mechanisms, such as the activation of NF-κB signalling via degradation of its regulatory protein IκBα. Inhibition of the proteasome attenuates these pathways and induces autophagy of abnormal proteins, such as FLT3-ITD [158].

Fig. 16.9

Agents targeting the proteasome. BAX apoptosis regulator BAX, BID BH3 interacting-domain death agonist, IκB α inhibitor of NF-κB alpha, NAE neural precursor cell expressed, developmentally downregulated 8 activating enzyme, NEDD8 neural precursor cell expressed, developmentally downregulated 8, Ub ubiquitin, UBC12 ubiquitin-conjugating enzyme 12

16.9.1 Proteasome Inhibitors

Bortezomib inhibits the 26S subunit of proteasome complex 2 [159]. This agent has been shown to exert anti-tumour activity via stabilization of p53, p27, IκBα, pro-apoptotic proteins BID and BAX, and other signalling proteins [159]. After demonstrating anti-leukaemic activity in preclinical studies, it was tested in AML patients in a number of clinical trials as monotherapy and in combination with other agents, including chemotherapy, hypomethylating agents, and HDAC inhibitors [158, 160]. Although it was minimally effective as a single agent, its combination regimens successfully induced remissions in varying portions of patients, with the highest response rates when added to intensive chemotherapy. Although bortezomib was generally tolerable, the risks of bortezomib-related peripheral neuropathy and potentially, pulmonary toxicity, are concerning [158, 160]. Other side effects of this agent include febrile neutropenia, nausea, and gastrointestinal disturbances [158, 160]. A phase 2 trial evaluating its role as a chemo-sensitizing agent is underway (NCT04173585).

16.9.2 NAE Inhibitors

Neural precursor cell expressed, developmentally downregulated 8 (NEDD8)-activating enzyme (NAE) promotes conjugation of NEDD8 to proteins, which results in their ubiquitination by Cullin-RING E3 ubiquitin ligase (CRL) and subsequent proteasomal degradation [161, 162].

Pevonedistat (MLN4924) is a first-in-class small molecule inhibitor of NAE. In preclinical studies, it downregulated NF-κB signalling, triggered oxidative stress, and caused apoptosis in AML cells [161, 162]. In view of its synergistic action with belinostat in inducing DNA SSBs and apoptosis in AML cells [163], this combination regimen will be tested in a phase I study in r/r AML patients (NCT03772925). The combination of pevonedistat and venetoclax also showed synergism in a preclinical model and yielded promising preliminary results in a phase I/II study [164, 165], prompting other phase I to III trials regarding this regimen (NCT04172844, NCT04266795, NCT03862157). Synergism between pevonedistat and LSD1 inhibitors was also demonstrated in another murine study [166]. These optimistic results paved way to phase I and randomized phase II trials evaluating the combination of pevonedistat and azacitidine, where it was effective and provided superior survival over azacitidine monotherapy along with a favourable safety profile [167, 168]. Common side effects of this agent include fever, peripheral edema, dyspnea, febrile neutropenia, nausea, gastrointestinal disturbances, and transaminitis. Pevonedistat will be evaluated in combination with LDAC (NCT03459859), cytarabine, and idarubicin (NCT03330821), HMAs (NCT04712942, NCT04090736, NCT03009240).

16.10 Targeting Nuclear Transport

16.10.1 XPO1 Inhibitors

Exportin 1 (XPO1), or chromosome maintenance protein 1 (CRM1), is a nuclear exporter responsible for the export of substances from the nucleus [169]. Aberrant activity of XPO1 contributes to the pathogenesis of AML via shuttling tumour suppressors, such as NPM1 and p53, into the cytoplasm, which perturbs their functions [169] (Fig. 16.10). Upregulation of XPO1 is also associated with FLT3 mutations and confers inferior prognosis in AML [169].

Fig. 16.10

Mechanism of actions of XPO1 inhibitors. IκB inhibitor of NF-κB, NPM1 nucleophosmin 1, p27 tumour protein 27, XPO1 exportin 1

Small molecule inhibitors of XPO1, known as selective inhibitors of nuclear export (SINE) or KPT-SINE, have diverse anti-leukaemic functions. These orally available agents irreversibly bind to the cysteine⁵²⁸ residue of XPO1 and alter its conformation, preventing export of tumour suppressors. They also induce differentiation via upregulation of the myeloid differentiation marker CD11b and downregulate WT and mutant FLT3 as well as c-KIT [169]. In addition, their strong activity against NPM1-mutant blasts is highlighted by a lower IC50 compared to NPM1-WT blasts [169]. An early KPT-SINE, KPT-185, demonstrated downregulation of FLT3 and induction of apoptosis in AML cell lines, while its analogue, KPT-276, prolonged survival in murine models [170].

Selinexor (KPT-330), a first generation SINE, demonstrated preclinical synergism with topoisomerase inhibitors (idarubicin, daunorubicin, mitoxantrone, etoposide), cytarabine, and sorafenib [171,172,173]. As monotherapy, Selinexor produced modest responses among patients in a phase I trial, but the subsequent randomized phase II Selinexor in Older Patients with Relapsed/Refractory AML (SOPRA) trial was terminated due to a failure of meeting the expected survival endpoint [174, 175]. Selinexor has been tested with multiple agents, including 7 + 3 induction (daunorubicin/idarubicin and cytarabine), fludarabine and cytarabine, cladribine, cytarabine, G-CSF (CLAG), high-dose cytarabine (HDAC) and mitoxantrone, and decitabine, where it induced excellent responses among patients [172, 176,177,178,179,180,181,182,183,184]. In combination with sorafenib, it also exhibited anti-leukemic efficacy in FLT3-mutant AML patients [185]. The use of selinexor as post-HSCT maintenance therapy has been explored with optimistic results in a phase I trial [186]. However, due to the CNS-penetrating properties of selinexor, its therapy is associated with dose-limiting toxicities such as cerebellar toxicity, anorexia, weight loss, and nausea [169, 174, 175]. Other major side effects include gastrointestinal disturbances, myelosuppression, and asymptomatic hyponatraemia [172, 174,175,176,177,178,179,180,181,182,183,184,185]. Preclinical studies also suggested that it may exert undesirable activity against normal haematopoietic cells [172]. Nevertheless, selinexor is currently studied as monotherapy in r/r paediatric AML, in combination with standard chemotherapy or with venetoclax in adult patients, and as post-transplant maintenance therapy (NCT02091245, NCT02403310, NCT02835222, NCT03955783, NCT02485535).

Eltanexor (KPT-8602) is a second-generation SINE with similar potency as selinexor. It is suggested to have an improved safety profile due to a lower degree of CNS penetration and reduced effect on normal haematopoiesis [187]. It exhibited potent single-agent anti-leukaemic effect and synergism with venetoclax in preclinical studies [187,188,189,190].

16.11 Targeting Epigenetic Pathways

Epigenetic regulators, such as DNMTs and HDACs, regulate transcription via controlling DNA methylation and acetylation [191] (Fig. 16.11). Aberrant activities of these pathways result in transcription of oncoproteins and/or transcriptional silencing of tumour suppressors, resulting in leukaemogenesis [191].

Fig. 16.11

Agents targeting epigenetic pathways. 5hmC 5-hydroxymethylcytosine, 5mC 5-methylcytosine, Ac acetyl group, Arg arginine residue, BET bromodomain and extra-terminal domain, BRD bromodomain-containing protein 4, DNMT DNA methyltransferase, DOT1L disruptor of telomeric silencing 1-like, EED embryonic ectoderm development, EZH drosophila enhancer of zeste homolog, HDAC histone deacetylase, HMA hypomethylating agents, LSD1 lysine specific demethylase 1, Me methyl group, Me3 trimethyl group, MLL mixed-lineage leukemia, PRC2 polycomb repressive complex 2, PRMT protein arginine methyltransferase, SUZ12 suppressor of Zeste 12, TET ten-eleven-translocation

16.11.1 Hypomethylating Agents

Hypomethylating agents exert anti-leukemic activities by inhibition of DNA methyltransferases (DNMT), causing demethylation and reactivation of tumour suppressor genes [192]. Azacitidine and decitabine have been extensively studied and are widely used in AML patients. To enhance the ease of administration, an oral formulation of azacitidine (Onureg, CC-486) was developed and has recently been FDA-approved for the treatment of AML. In phase I trials, oral azacitidine demonstrated efficacy in DNA demethylation with a prolonged duration compared to subcutaneous azacitidine and a favourable safety profile, with common side effects being myelosuppression and gastrointestinal disturbances [193, 194]. In a subsequent randomized placebo-controlled phase III trial evaluating its use as maintenance therapy, oral azacitidine was significantly more effective at providing survival benefits [195, 196]. More randomized studies of oral azacitidine compared with placebos as maintenance therapies are ongoing (NCT04173533, NCT01757535).

Guadecitabine (SGI-110) is a deoxyguanosine analogue of decitabine with resistance to cytidine deaminase (CDA), thus prolonging its activity. Several trials of this agent in AML patients showed remarkable responses with tolerable toxicities, such as myelosuppression and infections [197,198,199]. However, subsequent phase III trials had disappointing results [200]. It is currently undergoing evaluation with talazoparib in r/r AML patients and with donor lymphocyte infusion (DLI) in post-HSCT patients (NCT02878785, NCT03454984, NCT02684162).

ASTX727, an oral formulation of decitabine with a cytidine deaminase inhibitor, cedazuridine, is currently compared with intravenous decitabine in a phase III randomized trial (NCT03306264). Its combinations with venetoclax, ivosidenib, enasidenib, and ASTX 660, a dual antagonist of cellular inhibitor of apoptosis protein (cIAP) 1 and X-linked inhibitor of apoptosis protein (XIAP), are also undergoing evaluation in clinical trials (NCT04657081, NCT04746235, NCT04774393, NCT04155580).

16.11.2 HDAC Inhibitors

Histone deacetylase (HDAC) and histone acetyltransferase (HATs) mediate deacetylation and acetylation of both histone and non-histone proteins. They are integral to the regulation of numerous cellular activities, such as gene transcription [201]. In AML, aberrant activation of HDAC by oncoproteins impairs the tumour suppressor function of p53, inhibits cellular differentiation, mediates aberrant signaling pathways (e.g. c-MYC), and induces abnormal proliferation [201]. Thus, the efficacies of multiple HDAC inhibitors have been studied in AML (Table 16.4) [201]. HDAC inhibitors can be classified into hydroxamines, benzamides, cyclic peptides, aliphatic acids, and electrophilic ketones according to their spectrum of activities and molecular structures [201]. Among them, vorinostat, panobinostat, and belinostat appear to be the most clinically promising. These agents are generally safe with only mild side effects, such as fatigue, nausea, and gastrointestinal disturbances.

Table 16.4 HDAC inhibitors and their developments

16.11.3 LSD1 Inhibitors

Lysine specific demethylase 1 (LSD1) controls demethylation of H3K4 and can function both as a transcription activator and repressor [299]. Inhibition of LSD1 was shown to promote differentiation of AML cells [299]. Multiple agents targeting this enzyme have been studied as potential therapies for AML.

Tranylcypromine (TCP) is a selective LSD1 inhibitor which induced differentiation of AML cell lines and demonstrated synergistic effect with ATRA [300]. In a subsequent phase I/II trial, this combination was proven to be effective in AML patients [301]. This agent was tolerable, with hypotension, orthostatic dysregulation, vertigo, confusion, and cytopenias as its major adverse effects. Another trial regarding these two agents in AML is ongoing (NCT02717884).

Various analogues of TCP also demonstrated preclinical activities against AML cells [302,303,304,305,306,307,308,309,310,311,312,313]. Notably, iadademstat (ORY-1001) exhibited remarkable preclinical anti-leukemic efficacy and was effective and tolerable as monotherapy in AML patients in a phase I trial [314, 315]. A phase II trial regarding its combination with azacitidine is underway (EudraCT No.: 2018–000482-36). Another agent, GSK2879552, synergized with ATRA to exert anti-leukaemic efficacy in preclinical studies, but disappointing survival benefits from a phase I trial led to termination of the study (NCT02177812) [316]. Another LSD1 inhibitor, CC-90011, is also undergoing evaluation in combination with venetoclax and azacitidine (NCT0474884).

16.11.4 BET Inhibitors

Bromodomain and extra-terminal domain (BET) is a family of epigenetic readers responsible for regulating gene transcriptions [317]. Importantly, bromodomain-containing protein 4 (BRD4) is a member of this family which has been identified as a crucial mediator of various oncogenic pathways [299]. JQ-1 is a selective BRD4 inhibitor with potent preclinical anti-leukaemic efficacy as monotherapy and in combination with other agents, including cytarabine, ATRA, azacitidine, and ponatinib [318,319,320,321]. BI 894999 is another BRD inhibitor which also demonstrated marked single-agent anti-leukaemic activity and synergism with LDC000067, a CDK9 inhibitor in a preclinical study [322]. In addition, birabresib (OTX015/MK-8628) showed preclinical activity against AML cells as monotherapy and therapeutic synergy with either panobinostat or azacitidine [323]. It is now undergoing evaluation as monotherapy in a phase I/II trial (NCT02698189).

16.11.5 TET Inhibitors

Ten-eleven-translocation (TET) enzymes inhibit DNA methylation via oxidizing 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC) [299]. In AML, mutant-TET causes hypermethylation of various gene loci, resulting in impaired differentiation and uncontrolled proliferation [299]. Ascorbic acid serves as a co-factor for TET2 to restore its normal activity and is frequently found to be deficient in AML patients. It showed anti-leukemic efficacy in preclinical studies and synergized with decitabine to prolong patient survival in a clinical trial [324,325,326]. A phase II trial of azacitidine in combination with ascorbic acid is currently underway (NCT03397173).

16.11.6 Menin-MLL Inhibitors

Mixed-lineage leukemia (MLL) is a lysine methyltransferase which methylates H3K4, while menin functions as its co-factor. MLL translocations result in generation of oncoproteins and are generally markers of poor prognosis. Small molecule inhibitors with preclinical efficacies against MLL complexes include MM-401, MI-503, MI-463, and MIV-6R [327,328,329,330]. Strikingly, these agents also exhibited potent activity against NPM1-mutant AML cell lines, possibly due to the reliance of mutant NPM1 on Menin-MLL1 interactions for its aberrant gene expression [331]. In particular, MI-503 and MI-3454 selectively targeted MLL1-rearranged and NPM1-mutant cells and prolonged survival in murine models [331, 332].

16.11.7 DOT1L Inhibitors

Disruptor of telomeric silencing 1-like (DOT1L) is a histone methyltransferase mediating the methylation of H3K79 [299]. Since its function is integral to the oncogenic activities of MLL fusion complexes, it can be used as a potential target against MLL-rearranged AML [299]. Pinometostat (EPZ5676) is a DOT1L inhibitor with remarkable preclinical efficacy against MLL-rearranged cell lines and showed modest single-agent clinical activity along with a favourable safety profile [333,334,335]. It is currently being tested in combination with standard chemotherapy in MLL-rearranged AML patients (NCT03724084). SYC-522 is another agent with preclinical efficacy against MLL-rearranged AML [336].

16.11.8 EZH Inhibitors

Drosophila enhancer of zeste homolog (EZH) is a subunit of polycomb repressive complex (PRC) 2, which regulates trimethylation of H3K27 and mediates gene transcription [299]. Interestingly, they can function both as a tumour suppressor and oncoprotein in AML [299]. The selective EZH2 inhibitor 3-Deazaneplanocin A (DZNep) and its analogue D9 both showed efficacy against MLL-rearranged AML cells [337,338,339]. UNC1999 is a dual inhibitor of EZH1 and 2 with preclinical efficacy against AML models with MLL gene rearrangement [340]. Finally, valemetostat (DS-3201) is another dual EZH1/2 inhibitor currently evaluated as monotherapy in a phase I trial (NCT03110354).

16.11.9 PRMT Inhibitors

Protein arginine methyltransferases (PRMTs) are mediators of arginine methylation of histone as well as non-histone proteins and their overexpression is frequently found in AML [299]. AMI-408 is a specific inhibitor of PRMT1 with growth suppressive effect on AML cell lines and murine models [341]. ERZ015666, an inhibitor of PRMT5, induced differentiation of AML cells and showed efficacy in murine models with MLL rearrangements [342]. GSK3326595 is another PRMT5 inhibitor currently undergoing evaluation in combination with azacitidine in a phase I trial (NCT03614728).

16.12 Targeting DNA Damage Response Pathways

DNA damage response (DDR) is essential for the maintenance of genomic stability via halting cell cycle progression for DNA repair [343]. In the case of substantial DNA damage beyond repair, the apoptotic cascade would be initiated [343]. Studies have shown that AML cells have defective DDR mechanisms and are thus more susceptible to combined inhibition of chemical and DDR pathways [344]. Importantly, IDH-mutant AML is proposed to be sensitive to further inhibition of DDR due to their intrinsic defects in homologous recombination (HR).

16.12.1 PARP Inhibitors

Poly (ADP-ribose) polymerases (PARP) are a superfamily of 18 enzymes responsible for DNA single strand breaks (SSBs) repair and survival of cells with DNA damage [345]. Some subtypes of AML, such as those with IDH1/2 and FLT3 mutations, are proposed to be more sensitive to the effects of PARP inhibitors [345]. PARP inhibitors are nicotinamide analogues which function via the inhibition of DNA SSB repair by PARP and induction of cytotoxic allosteric effects by trapping PARPs to damaged DNA [345, 346].

Olaparib is a potent and selective PARP inhibitor which showed excellent potency against AML cell lines and synergistic activity with two anti-CD33 antibody drug conjugates, GO and IMGN779, in preclinical studies [347, 348]. Olaparib is in a trial as monotherapy for r/r IDH-mutant AML (NCT03953898).

Other PARP inhibitors with promising preclinical activities against AML include veliparib, talazoparib (BMN-673), niraparib, rucaparib, and PJ34 [345, 349, 350]. These agents showed synergistic activity against AML cell lines in combination with IMGN632 (anti-CD123 antibody drug conjugate), MS275 (HDAC inhibitor), entinostat (MS275, HDAC inhibitor), and AZD1775 in preclinical studies [351,352,353,354]. Among them, results of veliparib as single agent or in combination with temazolomide (alkylating agent) or topotecan and carboplatin in r/r ALM patients were impressive [355, 356]. Two trials regarding the use of these two combinations in AML patients are ongoing (NCT00588991, NCT01139970). Talazoparib also demonstrated potent efficacy against IDH1-mutant AML cells [357]. Trials of talazoparib as monotherapy and in combination with decitabine are currently underway (NCT03974217, NCT02878785).

16.12.2 ATR Inhibitors

Ataxia telangiectasia and Rad3-related kinase (ATR) is responsible for detecting DNA SSBs. It subsequently activates downstream repair pathways or apoptotic cascades depending on the extent of DNA damage [343]. VX-970 and AZ20 are two ATR inhibitors which demonstrated single agent efficacy against AML cell lines [358, 359]. AZ20 also synergistically induced anti-leukemic activity with cytarabine in another preclinical study [360].

16.12.3 ATM Inhibitor

The function of ataxia telangiectasia mutated kinase (ATM) resembles that of ATR except for its detection of double strand breaks (DDBs) instead of SSBs in DNA [343]. AZD0156, an ATM inhibitor, prolonged survival of MLL-rearranged mice in a preclinical study [359]. In another study, KU-59403 also induced apoptosis in AML cell lines [361].

16.12.4 CHK Inhibitors

Checkpoint kinase (CHK) 1 and 2 inhibit CDK 1 and 2 and cause cell cycle arrest upon activation by ATR and ATM [343]. Their overexpression in AML is associated with inferior prognosis [362]. Prexasertib (LY2606368), MK-8776 (SCH900776), and rabusertib (LY2603618) are CHK inhibitors which synergistically induced apoptosis in combination with CPX-351 in TP53-WT and TP53-deleted AML cells [363]. Rabusertib also exhibited synergism with venetoclax against AML cells [364]. MK-8776 demonstrated activity at overcoming chemotherapeutic resistance and synergized with cytarabine and vorinostat [362, 365, 366]. However, the combination of MK-8776 with cytarabine did not provide survival benefit over single agent cytarabine in r/r AML patients in a subsequent trial [367]. A phase I trial of prexasertib in combination with cytarabine and fludarabine is underway (NCT02649764).

16.12.5 WEE1 Inhibitors

Wee1-like protein kinase (WEE1) is activated by CHK and induces cell cycle arrest by inhibition of CDK1 and 2 [343]. Adavosertib (AZD1775, MK-1775) exhibited synergism with panobinostat and olaparib, respectively, in AML cell lines [224, 354]. It also synergistically overcame cytarabine-resistance when combined with cytarabine in leukemic cells [368]. Unfortunately, a trial of adavosertib as monotherapy was terminated due to safety concerns and another trial of its combination with belinostat was terminated for unspecified reasons (NCT03718143, NCT02381548).

16.13 Targeting the Cell Cycle

The cell cycle is a 4-phased process and progression through each phase is under strict regulation by several mediators, including cyclin-dependent kinases (CDKs) and cell cycle checkpoints. Aberrant progression of the cell cycle results in uncontrolled proliferation and leukaemogenesis [151].

16.13.1 CDK Inhibitors

Cyclin-dependent kinases (CDKs) are regulators of cell cycle progression which are activated upon binding of cyclins. Among them, transcriptional CDKs (CDK7, 8, 9) are mainly responsible for regulating transcription (Fig. 16.12). Inhibition of CDKs can halt the cell cycle and inhibit aberrant gene expression, giving rise to anti-leukemic effects. Information regarding various CDK inhibitors is summarized in Table 16.5. Among them, palbociblib and alvocidib are the most widely studied in AML. These agents have an excellent safety profile with myelosuppression as a significant side effect.

Fig. 16.12

Agents targeting DNA damage responses and cell cycle. ATM ataxia telangiectasia mutated kinase, ATR ataxia telangiectasia and Rad3-related kinase, AURK aurora kinase, CDC25A cell division cycle 25 A, CDC25C cell division cycle 25 C, CDK1 cyclin-dependent kinase 1, CDK2 cyclin-dependent kinase 2, CDK4/6 cyclin-dependent kinase 4/6, CHK1 checkpoint kinase 1, CHK2 checkpoint kinase 2, DSB double strand breaks, P phosphate group, PARP poly (ADP-ribose) polymerases, PLK1 polo-like kinases 1, RSK p90 ribosomal S6 kinase, SSB single strand breaks, WEE1 Wee1-like protein kinase

Table 16.5 Developments of cyclin-dependent kinase (CDK) inhibitors

16.13.2 Aurora Kinase Inhibitors

The aurora kinase (AURK) family consists of three members, AURK A, B, and C. These enzymes are responsible for entry into the M phase and normal progression of mitosis [393]. In AML, their overexpression has been observed and is associated with poor-risk cytogenetics. Alisertib (MLN8237) is an AURKA inhibitor with oral bioavailability. Investigational use of this agent in a phase II study in combination with induction chemotherapy among poor-risk AML patients illustrated its clinical effectiveness and safety [394]. Barasertib (AZD1152) is an AURKB inhibitor which demonstrated anti-leukaemic efficacy along with a less desirable safety profile in a phase I/II study in AML patients, with major side effects being febrile neutropenia and oral mucositis [395]. In another trial, it was tested in combination with LDAC and showed favourable outcomes and tolerability [396]. A trial of barasertib as monotherapy or in combination with venetoclax and/or azacitidine is underway (NCT03217838).

16.13.3 PLK Inhibitors

Polo-like kinases (PLKs) promote cell cycle progression via inducing degradation of WEE1 and activating CDK1 [397, 398]. It also inhibits apoptosis by activating Bcl-xL [398]. In AML, its overexpression is frequently observed [399].

Rigosertib (ON01910) is a dual inhibitor of PLK1 and PI3K. In addition to exhibiting preclinical anti-leukemic efficacy, it was effective and tolerable as monotherapy and in combination with azacitidine in clinical trials [400,401,402]. Major adverse events were gastrointestinal disturbances, myelosuppression, and pneumonia. A phase II study of oral rigosertib in combination with azacitidine is underway (NCT01926587).

Volasertib (BI6727) is a selective PLK1/2/3 inhibitor. Encouraging results from preclinical studies in AML models paved way for subsequent clinical trials in AML patients [403]. In summary, volasertib was safe and effective as monotherapy and demonstrated synergism in combination with LDAC and decitabine, respectively, among AML patients in phase I and II trials [404,405,406,407]. However, responses of its combination with LDAC did not meet expectations in the randomized phase III POLO-AML-2 trial [408]. Significant side effects of volasertib include myelosuppression and fatigue. It is currently undergoing evaluation as monotherapy or in combination with cytarabine in several trials (NCT00804856, NCT01721876). Another oral PLK1 inhibitor, onvansertib (NMS-1286937), demonstrated impressive efficacy and safety in combination with decitabine, but limited activity with LDAC in a phase Ib study [409].

BI2536 also exhibited anti-leukemic effect in a preclinical study and had modest single agent activity in AML as reported in a phase I/Ib trial [410,411,412]. Other PLK1 inhibitors with preclinical efficacies against AML include TAK-960 and NMS-P937 [413, 414].

16.13.4 CDC25 Inhibitors

Cell division cycle 25 (CDC25) is a protein phosphatase which modulates cell cycle progression via dephosphorylation of CDKs [415]. In a preclinical study, several CDC25 inhibitors, namely NSC95397, ALX1, ALX2, ALX3, and ALX4, inhibited proliferation of AML cells, but did not demonstrate cytotoxic effects [415].

16.13.5 RSK Inhibitor

p90 Ribosomal S6 Kinase (RSK) is a downstream mediator of the Ras/MAPK/ERK pathway and controls a wide range of cellular pathways, including the promotion of cell cycle progression via activation of CDC25 and CDK1 [416]. In AML, upregulation of RSK has been discovered in patient samples and is indicative of poor prognosis. BI-D1870 is an RSK inhibitor which exerts potent anti-leukemic activity via S phase cell cycle arrest, impeding mitotic exit, and induction of DNA damage [416, 417]. It was effective as monotherapy and showed synergism with vincristine in AML cell lines [416, 417].

16.14 Targeting the Bone Marrow Microenvironment

The bone marrow microenvironment (BMM) plays crucial roles for the normal development of HSCs and other haematopoietic cells. In AML, the complex interactions between leukaemic cells and the BMM are integral to their development and disease progression [418]. With overwhelming evidence suggesting the substantial abnormalities in the BMM of AML patients, multiple therapeutic strategies to target these aberrant pathways are being explored (Fig. 16.13) [418].

Fig. 16.13

Agents targeting the bone marrow microenvironment. CD44 cluster of differentiation 44, CXCR4 C-X-C chemokine receptor type 4, SDF1 stromal-derived factor 1

16.14.1 SDF1/CXCR4 Inhibitors

C-X-C chemokine receptor (CXCR) type 4 is a HSC surface G-protein-coupled chemokine receptor for stromal-derived factor 1 (SDF1), also known as CXCR12, which is produced by mesenchymal stromal cells. Their interactions promote survival, quiescence, and marrow homing of HSCs. Leukaemic cells exploit this mechanism by upregulating their expressions of CXCR4, which grants them chemoresistance due to protection by marrow stromal cells.

Plerixafor (AMD3100) is a small molecule inhibitor of CXCR4 commonly used as an off-label stem cell mobilizing agent. Promising results from preclinical studies prompted several trials of plerixafor in combination with chemotherapies [419, 420], decitabine [421], as well as with G-CSF with or without sorafenib [422]. These studies all showed impressive survival outcomes and demonstrated the remarkable potential of plerixafor as a chemo-sensitizing and AML blast mobilizing agent. It will be tested as a chemo-sensitizing agent prior to pre-transplant conditioning in a phase II trial (NCT02605460).

16.14.2 E-Selectin Inhibitors

E-selectins are molecules expressed by vascular endothelial cells which mediate cellular adhesion. Leukaemic cells express CD44, the ligand for E-selectins, to promote their engraftment in the bone marrow. Uproleselan (GMI-1271) is an inhibitor of e-selectin which showed preclinical efficacy in overcoming chemoresistance and synergism with chemotherapeutic agents [423]. Its use in several clinical trials in combination with chemotherapy yielded profound response rates, excellent tolerability, and even reduction in risks of mucositis [424,425,426]. Phase III trials evaluating comparing responses to chemotherapy with or without uproleselan are underway (NCT03616470, NCT03701308).

16.15 Immunotherapy

Immunotherapy represents a new era of therapies in AML and has been intensively studied in recent years. Broadly, these strategies can be classified into antibody-based or T/NK-cell-based depending on their mechanism of actions. The former involves targeting cell surface antigens of leukemic cells, while the latter relies on activation of immune responses against leukaemic cells. Compared to conventional chemotherapy, they are generally more tolerable due to reduced toxicity on normal cells.

16.15.1 Antibody-Based Immunotherapies

16.15.1.1 Antibody-Drug Conjugates

Antibody-drug conjugates (ADJs) are synthesized via the conjugation of cytotoxic agents to antibodies against various cell surface antigens of AML cells or LSCs. Upon cell surface receptor binding, they are endocytosed and release their cytotoxic moieties to induce leukaemic cell death (Fig. 16.14).

Fig. 16.14

Mechanism of actions of antibody-based immunotherapies. ¹³¹I Iodine-131, ²²⁵Ac Actinium-225, ⁹⁰Yt Yttrium-90, CD123 cluster of differentiation 123, CD33 cluster of differentiation 33, CD45 cluster of differentiation 45, DNA deoxyribonucleic acid

16.15.1.1.1 Anti-CD33 ADJs

CD3 is expressed primarily on LSCs and not in normal haematopoietic cells [152]. Thus, targeting this cell surface antigen allows selective eradiation of LSC while sparing normal haematopoietic cells [152]. Gemtuzumab ozogamicin (GO; Mylotarg) is an FDA-approved anti-CD33 ADJ with the cytotoxic agent calicheamicin as a conjugate. GO was first FDA-approved for the treatment of AML in 2000, but was withdrawn in 2010 in view of its non-superior survival benefit compared to standard 7 + 3 induction and high risks of toxicities, such as veno-occlusive diseases (VODs), and hepatotoxicity [427]. Despite these discouraging events, GO was continually studied at fractionated and lower doses with optimistic results. Notably, the phase III randomized ALFA-0701 study showed that the addition of GO to standard induction chemotherapy provided marked survival benefits with only a slight increase in risks of VODs [428, 429]. Another randomized phase III trial (AML-19) also reported that GO improved patient survival to a larger extent than best supportive care [430]. Following these encouraging results, GO was re-approved by the FDA for newly diagnosed and relapsed AML patients. However, it should be noted that subsequent controlled trials still reported higher risks of VODs and early mortality with GO therapy than in control groups [431]. Apart from VODs, other toxicities of GO include haemorrhage, infections, gastrointestinal disturbances, febrile neutropenia, and myelosuppression.

Further trials of GO include its use as monotherapy (NCT03737955) or in combination with pracinostat (NCT03848754) and venetoclax (NCT04070768); talozoparib (NCT04207190); OX40, venetoclax, avelumab, glasdegib, and azacitidine (NCT03390296); mitoxantrone and etoposide (NCT03839446), CPX-351 (NCT03904251, NCT03878927, NCT03672539), midostaurin and standard induction therapy (NCT03900949, NCT04385290), CLAG (NCT04050280), CLAG, and mitoxantrone (CLAG-M) (NCT03531918); fludrabine, cytarabine, G-CSF, idarubicin (NCT00801489); cytarabine, daunorubicin, erwinase, and etoposide (NCT04326439). It will also be tested as induction therapy followed by glasdegib (NCT04093505) or non-engraftment donor leukocyte infusion (NCT03374332),

Vadastuximab talirine (SGN33A) is another anti-CD33 ADJ linked to a pyrrolobenzodiazepine dimer. In preclinical studies, it exhibited remarkable anti-leukaemic activity in a diverse panel of cell lines, including those with TP53-mutant and multi-drug-resistant phenotypes [432]. It induced remarkable responses as monotherapy in AML patients in a number of trials, with some achieving MRD negativity [433, 434]. Although it also showed potent efficacy and induced MRD-negativity in combination with azacitidine in a phase I trial [435, 436], the subsequent randomized-phase III CASCADE trial was discontinued due to increased mortality in the experimental arm [437]. Treatment-related deaths were attributed to severe infections rather than VODs [437].

IMGN779 also targets CD33 and is conjugated to DGN462, an alkylating agent. It showed anti-leukaemic activity against multiple AML cell lines and murine models, with the highest activity in cells harbouring FLT3-ITD [438, 439]. Its use in a phase I trial yielded impressive response rates and tolerability [440].

16.15.1.1.2 Anti-CD123 ADJs

CD123 functions as an interleukin (IL)-3 receptor and mediates downstream proliferation induced by IL-3 [152]. It was also found to be highly expressed on LSCs and is an attractive target for eliminating leukaemic colony forming activities [152].

Tagraxofusp (SL-401) consists of an anti-CD33 antibody conjugated to part of the diphteria toxin [152]. It will be evaluated as monotherapy or in combination with venetoclax with or without azacitidine in AML patients (NCT04342962, NCT03113643).

IMGN632 is conjugated to DNA mono-alkylating portion of the indolinobenzodiazepine pseudodimer. It showed encouraging activity and favourable safety profile in a phase I trial [441]. Since it demonstrated synergism with azacitidine and venetoclax in a preclinical study [442], its combination with venetoclax, azacitidine, or both agents will be tested in a phase I/II trial (NCT04086264). Its use as monotherapy in AML patients will also be tested (NCT03386513).

16.15.1.2 Radioimmunotherapy

Radioimmunotherapy (RIT) involves the use of monoclonal antibodies linked with radionuclides, which then bind to leukaemic cell surface antigens and continually release ionizing radiation, resulting in selective anti-leukaemic effects (Fig. 16.14) [443]. In AML, RITs mainly utilize Iodine(I)-131 and Yttrium(Yt)-90 and are usually studied as pre-transplant conditioning regimens.

¹³¹I-anti-CD45 RIT markedly improve post-transplant outcomes in various clinical trials in combination with various conditioning regimens, including total body irradiation (TBI), busulfan and cyclophosphamide, as well as fludarabine and low-dose TBI [444,445,446,447]. The use of ⁹⁰Y-anti-CD45 RIT also resulted in remarkable survival outcomes and prolonged donor engraftment [448, 449].

While the above two agents emit β-radiation, ²²⁵Actinium(Ac)-lintuzumab (²²⁵Ac-anti-CD33) emits short-ranged α-radiation. Although it induced blast reduction in patients, no remissions were seen in a phase I trial combining this agent with LDAC in newly diagnosed patients [450]. ²²⁵Ac-lintuzumab and ²¹¹Astatine(At)-anti-CD45 will undergo further testing either as therapy for r/r patients or as part of conditioning regimens in multiple clinical trials (NCT03867682, NCT03441048, NCT03670966, NCT03128034).

16.15.2 T-Cell-Based Immunotherapies

T cells are integral to the normal functioning of the adaptive immune system. In particular, cytotoxic T cells are responsible for the elimination of cells carrying abnormal antigens, including leukaemic cells. However, these activities often impaired in AML, giving rise to immune evasion of leukaemic blasts. Thus, intensifying the anti-tumour responses of T cells is an attractive strategy against AML.

16.15.2.1 Immune-Related Adverse Events

Although immune-cell-based immunotherapies are generally considered to be more tolerable than conventional therapies, their resulting alterations in immune responses cause a distinct group of side effects termed “immune-related adverse events”. They can present in a multitude of ways, including as skin rash, pneumonitis, and colitis, among others [451]. Fortunately, the majority of these events are tolerable and not lethal. In addition, cytokine release syndrome (CRS) is especially common with the use of multivalent antibodies and CAR-T, with presentations ranging from mild flu-like symptoms to severe multi-organ failures and encephalopathy [452]. Cautious monitoring and proper management of CRS are keys to preventing significant morbidities and mortality.

16.15.2.2 Immune Checkpoint Inhibitors

Immune checkpoints (ICs) inhibit aberrant T-cell responses against normal body cells and are paramount to self-tolerance. However, leukemic cells can also express checkpoint ligands, which cause anergy of T-cells upon their binding, resulting in immune evasion and uncontrolled proliferation. Therefore, inhibition of these pathways allows reactivation of immune responses against leukemic cells (Fig. 16.15).

Fig. 16.15

Mechanisms of actions of immune checkpoint inhibitors and OX40 agonists. CD47 cluster of differentiation 47, CD80/86 cluster of differentiation 80 or 86, CTLA4 cytotoxic T-lymphocyte-associated protein 4, MHC I major histocompatibility complex class I, OX40 tumor necrosis factor receptor superfamily, member 4, PD-1 programmed death 1, PD-L1 programmed death ligand 1, SIRPα signal regulatory protein alpha, TIM3 T cell immunoglobulin and mucin domain-containing protein 3

16.15.2.2.1 PD-1/PD-L1 Inhibitors

Programmed death 1 (PD-1) is expressed on the surface of T cells while its ligand, Programmed death ligand 1 (PD-L1), is expressed on leukaemic cells [453]. Nivolumab, an anti-PD-1 antibody, induced impressive responses in combination with azacitidine in older r/r AML patients in a phase II trial [454]. Addition of ipilimumab to this regimen further improved survival outcomes at the cost of increased toxicities and immune-related adverse events [455]. The above study is still currently ongoing (NCT02397720). Another trial of nivolumab with cytarabine or idarubicin showed remarkable remission rates with measurable residual disease (MRD) negativity in more than half of the cohort [451]. Nivolumab is generally tolerable with mostly immune-related adverse events, such as skin rash, transaminitis, and nephritis. However, another trial of nivolumab as post-transplant therapy demonstrated minimal efficacy and unacceptable adverse effects [456]. In view of these encouraging results, it will be further evaluated in multiple trials, including combination with azacitidine in r/r paediatric AML patients, as monotherapy in post-transplant relapsed patients, and as post-chemotherapy or post-HSCT maintenance as monotherapy or in combination with ipilimumab (NCT03825367, NCT01822509, NCT02275533, NCT02532231, NCT03600155, NCT02846376). Its combination with NY-ESO-1 vaccination and decitabine will also be tested in a clinical trial (NCT03358719).

Pembrolizumab is another anti-PD-1 antibody which has been tested in combination with azacitidine, decitabine, and following HDAC in AML patients; their respective trials all showed optimistic outcomes and tolerable adverse effects which were mostly immune-related [457,458,459]. Further trials will evaluate this agent in combination with decitabine, galinpepimut-S, azacitidine, venetoclax, with intensive chemotherapy as frontline therapy, with azacitidine in NPM1-mutant AML patients with molecular relapse, and as monotherapy in patients with post-HSCT relapse (NCT03969446, NCT03761914, NCT04284787, NCT04284787, NCT03769532, NCT02981914, NCT03286114).

Avelumab, an anti-PD-L1 mAb, showed anti-leukaemic efficacy and tolerability in combination with decitabine in a phase I trial and is currently involved in a phase I/II trial with venetoclax, PF-04518600, glasdegib, GO, and azacitidine (NCT03390296) [460].

However, durvalumab (MEDI-4736), another anti-PD-L1 antibody, did not provide additional survival benefits when added to azacitidine in a randomized phase II study [461].

16.15.2.2.2 CTLA-4 Inhibitors

Cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) is expressed on the surface of T cells. It competitively binds to CD80 or CD86 expressed by leukaemic cells with a higher affinity than CD28 [453], inducing T cell anergy. Ipilimumab is an anti-CTLA-4 antibody which induced responses in AML patients who experienced relapse after HSCT, with graft-versus-host disease (GVHD) as the dose limiting toxicity in some patients [462]. It is undergoing evaluation in combination with decitabine, as monotherapy in patients with post-transplant relapse, as post-HSCT maintenance either as monotherapy, in combination with nivolumab, or in combination with donor lymphocyte infusion (NCT02890329, NCT01822509, NCT03600155, NCT02846376, NCT03912064).

16.15.2.2.3 TIM-3 Inhibitors

T cell immunoglobulin and mucin domain-containing protein 3 (TIM-3) receptors on T cells are activated by the binding of galectin-9 on leukemic cell surface [453]. Sabatolimab (MBG453) is an anti-TIM3 antibody. In a phase Ib trial, its combination with either azacitidine or decitabine showed promising anti-leukaemic activity [463]. This agent was tolerable, with myelosuppression and immune-related adverse effects being major side effects. Trials will further explore its combination with azacitidine and venetoclax, HDM201, and decitabine (NCT04150029, NCT03940352, NCT03066648). Its use in MRD-positive post-transplant patients will also be tested (NCT04623216).

16.15.2.2.4 CD47 Inhibitors

CD47 functions as an immune checkpoint by binding to Signal regulatory protein alpha (SIRPα) receptors on macrophage and preventing phagocytosis of CD47-positive cells [464]. The anti-CD47 mAb magrolimab is currently undergoing clinical evaluation in combination with azacitidine with optimistic preliminary results from a phase Ib trial [465]. It has been granted fast track designation by the FDA and will be tested in combination with azacitidine and venetoclax in a phase I/II trial (NCT04435691). A phase III trial comparing magrolimab combined with azacitidine against standard therapy is also underway (NCT04778397).

16.15.2.3 Targeting Co-Stimulatory Pathways

16.15.2.3.1 OX40 Agonists

OX40 is a cell surface receptor predominantly expressed by activated T cells, while its ligand, OX40L, is widely expressed by activated antigen presenting cells. The binding of OX40L to OX40 provides a co-stimulatory signal necessary for further T cell activation, clonal expansion, and anti-leukaemic immune responses [466] (Fig. 16.15). (PF-8600) is an anti-OX40 agonist monoclonal antibody currently investigated in combination with venetoclax, avelumab, glasdegib, GO, and azacitidine in a phase I/II trial (NCT03390296).

16.15.2.4 Multivalent Antibody Therapies

Multivalent antibody therapies facilitate interactions between immune cells and leukaemic cells. These recombinant antibodies are constructed by the combination of antibodies targeting these two types of cells and thus carry specificity against multiple antigens. This allows them to bring immune cells to the proximity of leukaemic cells for exerting anti-tumour effects (Fig. 16.16). Broadly, they can be divided into non-IgG-like and IgG-like, where only IgG-like multivalent antibodies retain the Fc region to promote additional immune pathways, such as antibody-dependent cell-mediated cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC) [467]. Currently, bivalent antibodies are the most widely studied in AML.

Fig. 16.16

Mechanism of actions of multivalent antibodies. BiTE bispecific T cell engager, CD3 cluster of differentiation 3, CD123 cluster of differentiation 123, DART dual-affinity retargeting antibodies, Fab antigen-binding fragment, Fc fragment crystallizable, FcRn neonatal Fc receptor, NK natural killer, scFv single chain variable fragment

16.15.2.4.1 Non-IgG like Multivalent Antibodies

16.15.2.4.1.1 BiTE

Bispecific T-cell engagers (BiTEs) contain both the heavy and light chain variable domains (VH and VL) of the single chain variable fragments (scFv) from two antibodies targeting T cells (e.g. CD3) and leukaemic cells (e.g. CD33, CD123), respectively.

AMG330 is an anti-CD33 x anti-CD3 BiTE with encouraging preclinical efficacy and is now currently evaluated in a trial with r/r or MRD-positive AML patients (NCT02520427) [468, 469]. AMG673 is another anti-CD33 x anti-CD3 BiTE. Preliminary results from an ongoing first-in-human phase I study revealed anti-leukaemic efficacy (NCT03224819) [470]. Manageable adverse effects, such as CRS and myelosuppression, were observed. AMV564 has the same antigen specificity as the above BiTEs, but is tetravalent, i.e. it contains two VH and VL chains from each types of antibody, which further promotes target binding. Favourable preclinical data promoted a phase I trial, which showed promising preliminary results (NCT03144245) [471, 472].

16.15.2.4.1.2 Dart

Dual-affinity retargeting (DART) antibodies are similar in principle to BiTEs, except the VH and VL chains from the two antibodies are cross-linked to further increase efficiency [467]. Flotetuzumab (MGD-006) is anti-CD3 x anti-CD123 DART which induced T-cell activation against CD123-positive leukaemic blasts in preclinical studies. Preliminary results from the subsequent in-human trial showed potent anti-leukaemic activity and manageable side effects, which mainly included CRS [473].

16.15.2.4.2 IgG-Like Multivalent Antibodies

XmAb14045 is an anti-CD123 x anti-CD3 multivalent antibody. Its long serum half-life of 6.2 days can be attributed to the binding of its bispecific Fc domain to neonatal Fc receptor (FcRn), which prevents its degradation [452, 467]. This agent demonstrated preclinical anti-leukaemic efficacy and induced T-cell activation. Although a phase 1 study in AML patients was suspended following occurrences of patient mortalities and major toxicities, including CRS and pulmonary edema, the partial suspension was shortly lifted by the FDA [474, 475]. A subsequent phase I study reported encouraging efficacy with manageable adverse events, such as CRS (NCT02730312) [476, 477]. The clinical activity and tolerability of XmAb14045 will be further elucidated in another trial (NCT02730312).

16.15.2.5 Chimeric Antigen Receptor T Cells Therapy

Chimeric antigen receptor T (CAR-T) cells therapy involves the use of genetically engineered T cells expressing chimeric antigen receptors (CAR) against leukaemic cell surface antigens. After infusion of CAR-T cells, binding of CAR to leukaemic cells triggers cytotoxic responses and leukaemic cell death (Fig. 16.17) [427].

Fig. 16.17

Mechanism of action of chimeric antigen receptor (CAR) T cells. NKG2D natural killer group 2D, NKG2D-L NKG2D ligand

CYAD-01 is a CAR-T cell product expressing the natural killer group 2D (NKG2D) fused to a CD3ζ signalling domain. NKG2D is normally expressed by natural killer cells, CD8+ T cells and NK-T cells. It is activated upon binding to NKG2D ligand (NKG2D-L) expressed by leukaemic cells, while a co-stimulatory signal is provided by DNAX-activating protein 10 (DAP10) [478]. Astonishingly, CYAD-01 is capable of inducing a co-stimulatory signal via DAP10-independent pathways. In a phase I trial, it was determined to possess anti-leukaemic activity. However, frequent adverse effects, including CRS and pneumonitis, were observed [479]. Another phase I trial with the use of NKG2D CAR-Tx cells in AML patients is underway (NCT04658004).

CAR-T cell products can also be engineered to target more than one leukaemic antigens to further improve potency. Notably, a compound CAR-T (cCAR-T) product constructed to target C-type lectin-like molecule-1 (CLL1) and CD33 was evaluated in a phase I study with remarkable results, with seven out of nine patients achieving remission and MRD-negativity [480, 481].

Other CAR-T products with preclinical successes in AML include c-Kit-targeting and FLT3 scFv-targeting CAR-Ts [482, 483]. CAR-T is undergoing intensive studies in numerous clinical trials, including novel CAR-T strategies such as donor-derived CAR-T (NCT04766840), CD123/CLL1 CAR-T (NCT03631576), CD38-targeted CAR-T (NCT04351022), and IL3 CAR-T (NCT04599543), among many others.

16.15.3 NK Cell-Based Immunotherapies

NK cells are paramount effectors of anti-tumour immune responses. After binding of an antibody to a cell surface antigen, the Fc receptors of NK cells bind to the Fc region of the antibody, resulting in activation of NK cells and release of cytotoxic materials for target cell killing. This process is known as antigen-dependent cellular cytotoxicity (ADCC). NK cell-based immunotherapies aim at harnessing the cytotoxic activity of intrinsic or foreign NK cells against leukaemic cells (Fig. 16.18).

Fig. 16.18

Summary of NK cell-based immunotherapies. CAR chimeric antigen receptor, NK natural killer

16.15.3.1 Unconjugated Antibodies

Unlike ADJs, unconjugated antibodies exert cytotoxicity by stimulating ADCC as well as CDC activated by their Fc domains. Daratumumab is an anti-CD38 mAb which demonstrated anti-leukaemic activity against AML cell lines via induction of ADCC and CDC [484]. Interestingly, it also targets leukaemic blasts via perturbing cellular metabolism [485]. It is currently studied as monotherapy and in combination with FT538 (NCT04714372, NCT03067571). A study of daratumumab in combination with DLI for patients who relapsed post-HSCT is also underway (NCT03537599).

Isatuximab is another anti-CD38 mAb with potent anti-leukaemic activity in a preclinical study [486]. It will undergo further evaluation in combination with chemotherapy in r/r paediatric AML patients (NCT03860844).

Talacotuzumab (CSL362) is a CD123 which demonstrated high potency against CD123 in preclinical studies [487]. However, its uses as monotherapy or in combination with decitabine were only minimally effective in multiple clinical trials and caused high incidences of treatment termination [488, 489].

16.15.3.2 CAR-NK Cells Therapy

CAR-NK cells are engineered to express receptors which enhance ADCC and are administered in conjunction to unconjugated antibodies [490]. FT538 is a CAR-NK cell product expressing an IL-5 receptor alpha fusion protein and high affinity non-cleavable CD16 [490]. It demonstrated promising preclinical efficacy against multiple myeloma cells and will undergo further testing in a phase I trial with daratumumab in r/r AML patients (NCT03067571) [491].

16.15.4 Vaccination

Vaccination of tumour-associated antigens is a strategy created to induce antigen presentation of dendritic cells to T cells, which then generate anti-leukaemic immune responses and prolonged immunological memory against leukaemic cells (Fig. 16.19) [427]. They can potentially prevent future relapses by inducing eradication of all remaining abnormal blasts in the haematopoietic system. Given the dismal prognosis of r/r AML, their development carries substantial significance for patients.

Fig. 16.19

Mechanism of action of vaccination for the treatment of AML. DC dendritic cell, WT-1 Wilm’s tumour 1

Wilm’s tumour 1 (WT-1) antigens are attractive targets for peptide vaccination due to their high expression in leukaemic cells [427]. Galinpepimut-S and OCV-501 are examples of WT-1 vaccines which showed potency at inducing immunological responses and improving survival in AML patients in CR1/2 in phase I or II trials [492,493,494]. The combination of galinpepimut-S with pembrolizumab is currently under evaluation (NCT03761914). Ombipepimut-S (DSP-7888) is another WT-1 peptide vaccine evaluated in AML (NCT04747002).

Owing to evidence suggesting that allogeneic dendritic cells (DCs) induce stronger immune responses compared to autologous ones, an allogeneic DC vaccine, DCP-001, was manufactured and examined in a phase I trial with optimistic outcomes and tolerability [495]. This vaccine is under clinical investigation in a phase II trial among AML patients in remission (NCT03697707).

In addition to the above agents, NY-SEO-1 vaccination is currently studied in a phase I trial in combination with decitabine and nivolumab (NCT03358719). This vaccine formulation comprises three components: (1) a mAb against DEC-2015 (CD205), a dendritic cell surface receptor which promotes antigen presentation; (2) NY-SEO-1, a leukaemic cell surface antigen; and (3) polyinosinic-polycytidylic acid complexed with poly-L-lysine and carboxymethylcellulose (Poly-ICLC), a double-stranded mRNA complex which serves as an immune stimulant [496, 497]. No preliminary results are available at the moment.

16.16 Conclusion

Given the plethora of aberrant pathways in AML, the aforementioned novel strategies only provide a glimpse of the endless therapeutic options against this aggressive haematological malignancy. Although the current prognosis of AML remains suboptimal, intensive efforts on the development of novel agents may soon bring about unprecedented pharmacological breakthroughs.