Reach Us +44 7460312890

Acute myeloid leukaemia - therapy - past, present and future

Paul Faduola1*, Alan Hakim, Juli Mansnérus1, Atsuko Imai1, Rob O’Neill2

1University of Edinburgh

2Edinburgh Cancer Research Centre.

*Corresponding Author:
Paul Faduola
E-mail: [email protected]
Visit for more related articles at Translational Biomedicine


Acute myeloid leukemia (AML) is characterized by genetic aberrations and a variable response to therapy which has made treatment of AML challenging. The objective of this paper is to review conventional treatments and their development, phase I-III clinical trials of new agents, novel pathways where future interventions may have therapeutic potential, and clinical trial assessment in AML. This study showed that a detailed understanding of the molecular changes associated with chromosomal and genetic abnormalities is necessary to pilot new therapy design. Although several deregulated proteins and genes have been identified, their diversity among AML patients have made it difficult to identify a single substance that can hit these diverse targets . New agents have shown promise but there remains a huge need to be met for effective and targeted therapies to be successful.


Chemotherapy, irradiation and haematopoietic stem cell transplantation (HSCT) are now standard therapies in acute myeloid leukaemia (AML) [1]. Chemotherapies, anti-metabolites blocking DNA/RNA synthesis, have been a corner stone for three decades. However they have numerous side effects and limitations in efficacy [2]. HSCT is similarly challenged by toxicity and efficacy [3]. The overall 5-year survival from AML remains poor, particularly in adults and in the elderly [4]. Genomic and proteomic technologies have provided opportunities for development of targeted therapies through improved understanding of molecular biology, and better characterization of AML subgroups [5-7]. New agents are in phase I-III clinical trials. Several have been rejected as ineffective or unsafe but a small number have demonstrated potential [8]. However, none stand alone as mono-therapies or more effective than standard chemotherapy in low- and intermediate-risk patient groups, and continue to be assessed for adjuvant properties. This paper will review conventional treatments and their development, phase I-III clinical trials of new agents, novel pathways where future interventions may have therapeutic potential, and clinical trial assessment in AML.

Drug therapies


Chemotherapy protocols are divided in to two stages. The first, ‘Induction’, aims to reduce diseased cells to undetectable levels (complete remission (CR)). The second, ‘Consolidation’ (or post-remission), is the elimination of residual undetectable disease to achieve a cure [9-10]. In relapse treatment may revert to Induction, though there may be need to lower dosage depending on individual circumstance e.g., toxicity of previous therapy and level of morbidity [11]. HSCT might be undertaken if Induction chemotherapy fails or a patient relapses despite Consolidation therapy[12]. It may also be undertaken as first-line therapy alongside chemotherapy for patients with high-risk disease e.g., cytogenetic group, underlying myelodysplasia (MDS), or secondary and therapyrelated AML [13].

Successful intervention is hampered by cytogenetic heterogeneity, toxicity (Box 1), multi-drug resistance (MDR), and age [14]. Older patients (age >60) respond less well. In part this is due to the presence of co-morbidities but also greater association with MDS and secondary AML. In addition, clinicians have had a tendency to view therapy as palliative in older age groups, preferring to avoid toxicity, and studies have been biased by exclusion of these groups. In many respects it has been this poor outlook for elderly and secondary AML that has driven new agents through phase I and II trials recently. Although the majority of patients under 60 years reach CR after intensive chemotherapy[15-16], relapse free survival (RFS) is uncommon. The 10-year overall (OS) and event free (EFS) survival for children/adolescents after Induction therapy is 55- 65%. However, in adulthood only 20-40% of all patients gain disease-free survival (DFS) of >5 years from chemotherapy alone.


Box 1. Common side effects / consequences of chemothera.

Radiation therapy for AML is generally used only if there is central nervous system involvement and no response to systemic and/or intrathecal chemotherapy [17]. It may also be used in preparation for HSCT.

‘Old dogs, new tricks’

The focus of chemotherapy clinical trials research over the last 15 years has been to identify:

i. The most appropriate dose and frequency of administration,

ii. The efficacy of combination therapies,

iii. Whether certain agents are better than others by direct comparison,

iv. New formulations of established drugs (e.g., Liposomal drugs), and

iv. The value of adjunct therapies - facilitating effectiveness of chemotherapy and/or combating MDR.

Tables 1 (Induction) and Table 2 (Consolidation) are a synthesis of trial literature demonstrating the development of chemotherapies over the last decade and a half, with reference to differences between age groups, and to CR and OS. Various Consolidation strategies have been evaluated including intensive conventional chemotherapy, prolonged maintenance treatment, and high-dose therapy followed by autologous or allogeneic HSCT.


Table 1. Induction Protocols with Conventional Chemotherapies


Table 2. Consolidation Chemotherapy.

The encapsulation of drugs in liposomes has led to new ways of more effectively delivering chemotherapy with a reduction in toxic side effects. Liposomal Daunorubicin has been shown in a Phase III trial to be as effective as normal Daunorubicin but better tolerated [60]. The agent CPX-351 is a liposmal fixed combination of Daunorubicin and Cytarabine. Recent Phase I/II trials suggest CPX-351 to have an acceptible safety profile for use in older and previously untreated patients [61,62]. Similarly, Elacytarabine, a derivative of Cytarabine but one that inhibits both DNA and RNA synthesis, has been demonstrated to be efficacious at least to the same degree as other agents, but with less toxicity in recent Phase II trials of patients in relapse requiring salvage therapy [63].

Given poor outcomes despite major advances in understanding best use of conventional chemotherapeutic agents, the need to develop novel therapies with different anti-leukaemic mechanisms is paramount. New agents entering the clinical arena include:

New Molecular Targeted Therapies

i. Monoclonal antibodies,

ii. Tyrosine kinase, and farnesyltransferase inhibitors,

iii. Cell growth blockers,

iv. Immunotherapies,

v. MDR1 inhibitors, and

vi. Peptide vaccines.

Table 3 and 4 outline trials in this area. In the majority of cases these trials have small numbers of patients and/or have selected for more complex disease and non-responders.


Table 3. Molecularly Targeted Therapy in Clinical Trials and Practice.


Table 4. Peptide Vaccination in AML.

On the whole the inhibition of deregulated transcriptional activity consequent on gene mutations has not led to therapeutic innovation; the exception to this is all-transretinoic acid and arsenic trioxide in APL.Inhibition of tyrosine kinase activity, nucleoside analogues, and monoclonal targeting of the antigen CD33 have demonstrated some success but none of these agents are superior to the combination Induction and Consolidation highlighted in Table 1 and 2.

Combinations of novel and current therapies are currently being explored in multi-faceted/multi-centred trials. The AML-17 (recruitment from 2002 to 2014) is one such phase III study looking at:

i. The best dose of Daunorubicin,

ii. CEP-701, a new FTL3 inhibitor,

iii. Everolimus (Afinitor) a signal transduction inhibitor that blocks the signalling protein mTOR,

iv. The comparison of 2 chemotherapy treatments before HSCT,

v. The comparison of 1 with 2 or more cycles of chemotherapy, and

vi. The role of Arsenic Trioxide in non-APL AML.

And in addition:

i. Clofarabine, a nucleoside analogue (see above), is being compared with other chemotherapies having been shown to have fewer side effects but similar efficacy to Fludarabine in the treatment for older patients considered unsuitable for induction chemotherapy, and

ii. Studies will continue to look at the role of Gemtuzumab.

There does not appear to be a revolutionary step change in drug therapy on the horizon in the management of AML. Attention is focused on synergistic effects of combining conventional with novel targeted agents. Though targeting of leukaemic stem cells (LSC) by, for example, receptor specific small molecules and peptide vaccines would appear a reasonable approach, the similarities between LSC and normal stem cells is also a challenge. Currently the targeting of LSCs remains relatively non-selective and requires simultaneous interventions.

Haematopoeitic stem cell transplantation

The developmental milestones in stem cell therapy (SCT) from the 1950s from preclinical trials to the successful application in human transplantation in the late 1970s are shown in Table 5. These laid the foundation for many areas of stem cell research, as well as current HSCT practices.


Table 5. Developmental milestones in HSCT.

Four sources of HSCT are available and each has its pros and cons (Table 6). The vast majority of clinical trials have been in allogenic and autologous stem cell transplants (SCT). Despite our learning of how best to use these therapies the challenges remain better techniques for cryopreservation, identification and classification of genetic markers, and understanding the influence of SNPs [3], non-HLA genetics, and cytokine genes [110-112].


Table 6. Pros, Cons and Challenges of the different types of HSCT.

In patients with favourable- and intermediate-risk cytogenetics, autologous HSCT is an alternative Consolidation option to chemotherapy. It is not recommended in cases with highrisk cytogenetics [113-115]. There is no evidence that this approach gives a better outcome in general, however it may be of advantage in cytogenetically normal and tandem repeat subsets of AML [116]. The lowest relapse rates are observed following Consolidation with allogeneic HSCT. The benefit is in part attributable to a potent graft-versus-leukaemia (GVL) effect [117]. Meta-analyses of clinical trials comparing allogeneic HSCT versus Consolidation chemotherapies after first CR show a significant improvement in OS in intermediate- and high-risk AML [118-120]. Table 7 shows data from several studies in the 1990s. The DFS following Consolidation is in general superior for allogenic vs autologous HSCT, and for HSCT vs chemotherapy. However OS rates were not significantly different in a number of these studies. Data from the 2000s in childhood disease is more compelling for favourable outcome of DFS and OS after allogenic HSCT (Table 8).


Table 7. Comparative Disease Free Survival following Allogeneic HSCT, Autologous HSCT, and Chemotherapy for AML patients in first remission


Table 8. Comparative outcome data in Childhood Trials of Allogeneic HSCT, Autologous HSCT, or Chemotherapy for AML

It is important to consider the risks of treatment-related mortality (TRM). These range between 15-50% and may out-weigh the benefits. These risks have been improved for older patients in particular by using reduced-intensity chemotherapy (RIC) regimes that are non-myeloablative. The outcomes are much more promising and comparable with younger cases [134,35],although relapse rates remain a challenge [136]. Data available from the European Group for Blood and Marrow Transplantation (EBMT) and the Centre for International Blood and Marrow Transplantation Research (CIBMTR) demonstrates that RIC regimens result in comparable outcomes across the adult age range (Table 9) [137]. As raised above for chemotherapy, data is difficult to interpret for older patients due to small patient numbers, heterogeneity, and selection bias. For these reasons prospective comparison of allogeneic HSCT from matched related and unrelated donors using RIC with conventional Consolidation therapy was launched in 2008 and as of December 2011 continues to recruit to this important clinical trial (ClinicalTrials. gov Identifier: NCT00766779). The use of RIC led to concern over graft versus host effects (Table 6) and has prompted research into immune-suppression studies to modulate GVL and GVH reactions [138]. Finally, research continues in to the benefit and most appropriate regimens using umbilical cord blood (Table 6) [139].


Table 9. Comparison of EBMT and CIBMTR data in Elderly Patients

Novel pathways and potential future agents

Several pathways and novel mechanisms of intervention are the focus of attention in current AML research.

AKT Inhibitors

As raised in Table 3 mTOR is a kinase involved in regulation of cell growth and proliferation. Signalling depends on its interaction through the PI3/Akt pathway [140]. Both PI3K and Akt are considered to be protooncogenes. Increased membrane expression of Akt is important in intiating malignancy. It also appears to confer resistance to apoptosis through the mitogen-activated protein kinase (MAPK) pathway [141]. As well as targeting mTOR (Table 3), Akt and MAPK inhibitors may represent new classes of drug. Perifosine is an Akt inhibitor and has shown preclinical activity against haematologic malignancies [142]. Phase I and II trials have been conducted in patients with solid tumours but not in leukaemias [143]. Alkylphosphocholines are lipophyillic drugs that have also been shown to modulate signal transduction by their interaction with c-myc, PI3-Akt, and MAPK pathways. Erufosine, an alkylphosphocholine, has anti-leukaemic properties that warrant further exploration [144].


Studies have demonstrated that AML blasts exhibit significant lower levels of Histone H3 acetylation (H3Ac) compared to CD34+ progenitor cells. As a consequence it is suggested that a number of genes are epigenetically silenced or diminished in AML. Agrawal Singh et al [145] recently showed that Peroxiredoxin 2 (PRDX2) is a novel potential tumour suppressor gene in AML. H3Ac was decreased at the PRDX2 gene promoter in AML, and correlated with low mRNA and protein expression. Low protein expression of the antioxidantPRDX2gene was clinically associated with poor prognosis in AML. They identified PRDX2 acts as an inhibitor of myeloid cell growth by reducing levels of reactive oxygen species (ROS) generated in response to cytokines. Taken together, epigenome-wide analyses of H3Ac in AML, led to the identification of PRDX2 as an epigenetically silenced growth suppressor suggesting a possible role of ROS in the malignant phenotype in AML. This may be a pathway to explore in the application of Histone Deacetylator agents (Table 3).

TET2 mutation

Over the last 2-3 years mutations of Ten-Eleven Translocation 2 (TET2), have been found in various myeloid malignancies. The gene is associated with DNA methylation, mutations leading to inhibition or reduction in appropriate myeloid cell differentiation (Chapter 2) and appears to be a prognostic biomarker in AML associated with intermediate-risk cytogenetics [146-153].

In a study last year by Weissmann et al [154] 131 somatic TET2 mutations were identified in 87/318 (27.4%) patients, and in 30% of cases of normal karyotype AML versus 19% of abnormal karyotype. Mutations of TET2 were concomitantly observed with mutations in NPM1, FLT3-ITD, FLT3-TKD, JAK2, RUNX1, CEBPA, CBL and KRAS (Chapter 2). Patients tended to be of older age, with higher haemoglobin level, higher neutrophil and monocyte counts, and lower platelet count. Similar mutational associations were identified by Chou et al [155]. Survival analyses (restricted to the normal karyotype population (n=165)) in Weissmann’ study showed inferior EFS in the presence of TET2 mutations.

In two other studies, one retrospective [156] the other prospective [157] the presence of TET2 mutations did not appear to influence CR or OS after standard therapy. There has also been one clinical study in higher risk MDS and AML with low blast count, where TET2 status was observed to be a genetic predictor of response to Azacitidine, independently of karyotype [141]. Further clinical studies with such hypomethylating agents are warranted.

DNMT3A mutation

DNMT3A mutations are observed in up to 22% of AML patients and appear more prevalent in the intermediate-risk groups, and especially of normal karyotype [158]. The mutations are strongly associated with poor prognosis [159-161], and like TET2 are associated with decreased DNA methylation and promotion of cell differentiation. DNMT3A forms a complex with transcription factors like histone methyltransferase and histone deacetylase [162,163]. Novel DNA methyltransferase and histone deacetylase inhibitors can reverse the methylomic phenotype of myeloid blasts (Table 3 Nucleosidase and Histone Deacetylase Drugs).During therapy, early platelet response and demethylation of the FZD9, ALOX12, HPN, and CALCA genes were associated with clinical response. Epigenetic modulation deserves prospective comparisons with conventional care in patients with high-risk AML, at least in those presenting previously untreated disease and low blast count.

Trial methodology


In vitro studies of cultured native AML cell lines and blasts have remarkably contributed to our current understanding on the pathogenesis of AML (Table 10). Well-characterised serum-free in vitro conditions are now used in experimental studies of AML, facilitating comparisons between different experiments. Assays for characterisation of AML progenitor subsets such as suspension cultures, colony assays, long-term in vitro culture, xenotransplantation in immunocompromised mice, as well as, AML cell lines as experimental models have been used to increase our knowledge on pathogenesis of AML [164]. Furthermore, biomarker studies suggest that the in vitro growth characteristics of AML blasts assayed by shortterm culture of the total native populations can be used as a predictor of prognosis after intensive chemotherapy. In vitro assays may be used for more accurate identification of prognostic parameters and for creation of a basis for the development of simplified laboratory techniques suitable for routine evaluation of patients undergoing risk-adapted therapy [164].


Table 10. Experimental models for the study of AML cell proliferation.


Drug development processes are lengthy and costly. While the phase I-III sequence of clinical drug testing has remained intact for decades, it appears inherently inefficient and the high frequency of false-positive results obtained in phase II studies constitutes a significant scientific concern [175- 180]. The sequential trial scheme puts major emphasis on such studies because they typically inform the decision to proceed to a phase III evaluation [175]. Strategies to mitigate shortcomings caused by lack of control groups, patient heterogeneity, selection bias, and choice of end points and strategies for streamlining trial design have been suggested. Such enhancements would among others encompass larger phase II studies, inclusion of (preferably randomised) controls, consideration of integrated phase 2/3 studies, accounting for patient heterogeneity even in small randomised studies, provision of information about the number of patients available for study vs. those actually treated, and avoidance of unvalidated alternate endpoints and premature publication (Table 11) [175].


Table 11. Suggestions for improvements of clinical trial designs in AML.

Phase I trials often provide novel agents to patients with relapsed and refractory disease [181]. It has been argued that noncytotoxic, molecularly targeted agents have not been very successful in this setting. Thus, signals of their true biologic efficacy may be missed and consequently potentially useful agents seem fail to demonstrate a signal of efficacy in the phase I setting. It has been proposed that at least some of these compounds should be considered ins tead in trials to prolong response duration [181].

Phase II trials in AML are usually small-scale and may give misleading efficacy signs [175, 166]. Due to the heterogeneity of the disease, subset analyses based at least on age, performance status, cytogenetics, and molecular features are necessary [175]. However, these are meaningless when the total group includes a small number of patients [181]. Consequently, there is increasing support for randomised phase II trials strategies planned to quickly compare new treatments with existing standards using as few patients as possible and to proceed only with those that meet predetermined efficacy benchmarks [181].

Phase III trials in AML are often slow, expensive to complete and, regrettably, often resulting in minor improvements. It has also become increasingly difficult to determine a feasible control group for new phase 3 trials due to the large number of molecularly and clinically defined subgroups. Further enhancements in the molecular characterisation of AML could allow the identification of more homogeneous treatment cohorts and tailored therapeutics [181].

Translational strategies to accelerate drug development

• Focusing on development of existing drugs in addition to searching for new ones. Due to the heterogeneous pathogenesis and molecular genetics of AML, tailored, personalised treatment based on the specific biologic features of the leukemic cells should be the objective. This also indicates that combination therapy is likely to remain superior to any single compound [181]. One interesting suggestion is to use the gene-expression signature generated from drugs that effectively ablate LSCs to study publicly available databases for other similar signatures [182]. In the case of “off-patent agents” , this could possibly al-low existing agents with well-identified clinical profiles and easy availability, to rapidly lead into AML treatments [181].

• Increasing participation in clinical trials. Less than 5% of adult cancer patients in US participate in clinical trials, contrary to 60% of paediatric cancer patients [181,183]. Recently, a French survey reported that 25% of AML patients among 1066 adults with AML were enrolled in clinical trials [181,184]. Physician and patient education about clinical trials should be enhanced and collaboration between academic centers and cooperative groups should be improved [181]. Increasing accrual in clinical trials is vital, as there the traditional phase I-III drug-development paradigm seems ineffective in this disease [175,181].

• Improving of safety and efficiency. Implementing biomarkers in clinical trials may improve decision-making in drug development process [185]. Biomarkers predicting therapeutic response enable the selection of patients most likely to have positive treatment outcomes with a particular oncologic therapy. Predictive pharmacogenomic biomarkers enabling selective treatment, are likely to become increasingly common in future therapies [186]. Biomarkers predicting the safety of a compound are highly valuable for preclinical testing, or early clinical studies. Microdosing studies could be used for improving safety when evaluating drug candidates at early stage development. Adaptive trial designs in AML studies could improve safety and efficacy by providing opportunities to make changes to a study in response to accumulating data whilst maintaining the trial’s integrity and validity.


AML is characterised by a multitude of chromosomal abnormalities and gene mutations, which translate to marked differences in responses and survival following chemotherapy, radiotherapy and HSCT. These chromosomal and genetic abnormalities make the treatment of AML challenging. The limit of acceptable toxicity for standard chemotherapeutic drugs used in AML therapy has been reached. A detailed understanding of the molecular changes associated with chromosomal and genetic abnormalities is necessary to pilot new therapy design. Although several deregulated proteins and genes have been identified, their diversity among AML patients have made it difficult to identify a single substance that can hit these diverse targets . New agents have shown promise but there remains a huge need to be met for effective and targeted therapies to be successful.


Select your language of interest to view the total content in your interested language

Viewing options

Post your comment

Share This Article

Flyer image
journal indexing image