Molecular diagnosis and treatment for AML – Educational session at the 22nd congress of the EHA

At the 22^nd Congress of the European Hematology Association (EHA), Madrid, Spain, the AGP were delighted to attend an educational session on Acute Myeloid Leukemia (AML). During this session, the evolving molecular heterogeneity of AML and how diagnosis can advance therapy were discussed as well as current and targeted therapy for AML. The session was chaired by the chair of our Global and European Steering Committee, Professor Gert Ossenkoppele, VU University Medical Center, Amsterdam, the Netherlands.

Molecular diagnostics in AML

The first talk during this session was given by Professor Lars Bullinger from the University Hospital of Ulm, Ulm, Germany, who discussed how AML patients could be diagnosed molecularly and  used for therapeutic decisions.

The speaker began the talk by discussing the heterogeneity of AML. AML is a clonal disorder that is characterized by multiple somatic driver mutations that affect genes of different functional categories, a complex clonal architecture, and disease evolution over time.

In recent years, there have been a great advancement in the interpretation of the molecular heterogeneity of AML by genomics discovery studies that use novel high-throughput sequencing technique including microarray and Next Generation Sequencing (NGS) based omics technology. This has led to a multitude of new mutations and gene expression signatures resulting in distinctive genomic make up of every individual AML. Additionally, these cytogenetic aberrations have become well recognized diagnostic and prognostic markers. The speaker noted that despite these developments, improving AML patient outcomes remains a major challenge.

Then, the speaker spoke about some current unmet needs and how they could be resolved and translated into clinical routine (see slide below).

The speaker briefly discussed the 2017 European LeukemiaNet (ELN) recommendations for AML which was updated to include three group classification (favorable, intermediate and adverse risk group). Mutations in three genes, RUNX1, ASXL1, and TP53—have also been added in the adverse risk group. Additionally, AML patients with NPM1 mutation and low Fms Like Tyrosine Kinase 3 – Internal Tandem duplication (FLT3-ITD) ratio have been included in the favorable group. Conversely, AML with wild-type NPM1 and high FLT3-ITD ratio is now assigned to the adverse risk group.

The focus of the talk was then turned to the use of genomics-based knowledge data banks and its use in outcome prediction. These genomics based knowledge banks can be created using NGS-based routine AML diagnostics and can aid in individualized risk prediction and therapeutic decision making. In a study using data sets from 1,540 AML patients enrolled in three German-Austrian study group Acute Myelogenous Leukemia (AMLSG) trials, a large knowledge banks of matched genomic and clinical data was created which was explored for clinical decision making. A multistage statistical model was devised to accurately predict risk of relapse, remission and mortality in AML patients.

Then Professor Lars Bullinger suggested how this knowledge database banks can be translated into clinical treatment. He described a scenario, whereby a newly diagnosed AML patient would be molecularly screened using a web-based system to detect mutations. The information gathered can then govern the clinical decision for the patient.

Currently, there are a range of ongoing studies that are incorporating the use of knowledge bank database in therapeutic decisions including the phase III randomized study (NCT02298166) of crenolanib (Fms-Like Tyrosine Kinase 3 [FLT3]) inhibitor, in combination with chemotherapy in patients with Relapsed or Refractory (R/R) AML and activating FLT3 mutations.

The speaker then discussed several techniques including targeted resequencing strategy, molecular barcode-based NGS, omics NGS approaches, RNA-sequencing based fusion gene detection, Oxford Nanopore sequencing technology and single-cell transcriptomes which could be used for molecular diagnosis of patients and potentially identify novel disease relevant genes that could be used therapeutically for AML patients.

Professor Lars Bullinger concluded his talk by stating that a “new era in leukemia genomics” will provide novel insights. He further added that large knowledge database is now currently being used in and are translated into therapeutic benefit. However, he highlighted that a major challenge remains the “identification of predictive biomarkers” that could aid personalized therapy for AML patients.

Targeting mutated FLT3 in AML

The next talk was given by Associate Professor Mark Levis from the Sidney Kimmel Comprehensive Cancer Center John Hopkins Medicine, Baltimore, Maryland.

Mutations in the FLT3 gene is one of the most commonly encountered, and clinically challenging, classes of AML mutations and it is expressed in approximately 30% of patients.

During this talk, the speaker spoke about studies using highly specific and potent FLT3 inhibitors such as quizartinib and gliteritinb in advanced AML and also first generation FLT3 inhibitors including sorafenib, midostaurin and lestaurtinib in AML patients.  A detailed coverage discussing targeted therapy for FLT3 mutated AML is reported here.

The speaker concluded his talk by stating that “targeting mutated FLT3 is now an established but evolving, part of AML treatment”.

3 + 7 and beyond

The final talk in this session was given by a member of our European Steering Committee, Professor Norbert Vey from the Institut Paoli-Calmettes, France. He discussed new agents beyond the classical 3 + 7 induction therapy for frontline therapy for AML.

The speaker began his talk by highlighting that optimization of the 3 + 7 dose-schedule comprising of daunorubicin and a consensual anthracycline- cytarabine backbone along with risk stratification, supportive care has improved the survival of AML patients. However, there are new agents that have been developed for the treatment of AML. These classes of drugs include new cytotoxics, immunotherapies and targeted therapies.  The speaker moved on to discuss these agents and how they can be incorporated with the standard 3 + 7 backbone.

New Cytotoxics

CPX-351, a liposomal formulation of cytarabine plus daunorubicin co-encapsulated at a molar ratio of 5:1, has been shown to be accumulated in the bone marrow and preferentially taken by leukemic blasts.  In a prospective randomized phase III study (NCT01696084), 309 newly diagnosed AML patients (60–75 years) were randomized to receive either CPX-351 (n = 153) or 7+3 (cytarabine and daunorubicin, n = 156) induction therapy. Results from this study showed that CPX-351 significantly improved the median Overall Survival (OS) of patients compared to standard 7+3 (9.56 vs 5.95 months, HR = 0.69, P = 0.005). Conversely, CPX-351 significantly improved the CR + CRi response compared to patients treated with standard 7+3 (47.7% vs 33.3%, P = 0.016).

Vosaroxin is an anticancer quinolone derivative, that intercalates DNA and inhibit topoisomerase resulting in apoptosis. In a phase III randomized VALOR (NCT01191801) study, the efficacy and safety of vosaroxin and cytarabine  versus placebo and cytarabine was investigated in patients in first Relapsed or Refractory (R/R) AML. Findings from this study demonstrated that in comparison with cytarabine plus placebo, vosaroxin/cytarabine did not have any significant advantage in survival of patients. However, there were some benefits observed with elderly R/R AML patients.

Vosaroxin is currently being evaluated in a phase 2 VITAL study (NCT02658487), which is assessing vosaroxin in combination with infusional cytarabine in patients with newly diagnosed AML.

Targeted therapy

The focus was turned to the use of targeted therapies in AML particularly agents targeting Isocitrate Dehydrogenase (IDH) 1 or 2 mutations. Approximately 6–10% and 15% of AML patients carry mutations in IDH1 or IDH2 respectively.

IDH enzymes catalyze the conversion of isocitrate to α-Ketoglutarate (αKG). However, mutations in IDH1/2 lead to a reverse reaction of αKG to the oncometabolite D-2-Hydroxglutarate (D-2HG). The accumulation of 2HG competitively inhibits αKG, thus leading to alterations in TET2-dependent hydroxymetlylation, chromatin modification, activation of the hypoxic response, and increased dependence on BCL2.

The new class of targeted IDH inhibitors bind with the mutant IDH1 or IDH2 catalytic active site thus preventing the oncogenic reduction of αKG to D-2HG. These agents include AG-221, a mutant IDH2 inhibitor, and AG-120, a mutant IDH1 inhibitor.

Interim results from the phase I/II study (NCT01915498) of AG-221 monotherapy in patients with R/R AML showed that enasidenib was well tolerated with an ORR of 41% and a CR rate of 18% in 128 evaluable R/R AML patients.

AG-120, an IDH1 inhibitor, has also shown promising results in patients with R/R AML patients. In a phase I dose escalation study, AG-120 was shown to be effective at suppressing D-2HG. Additionally, 36% of patients that achieved CR attained IDH1 mutational clearance as detected by NGS.

The speaker noted that majority of AML patients lack mutations that could be targeted and hence different strategies for treating patients are required. Innovative strategies include targeting AML-specific cell surface antibody using monoclonal antibodies, bispecific antibodies or CART- T cell, targeting critical molecular pathways including apoptotic cycle, cell cycle and epigenetic regulation and finally by targeting immune effector cells such as T-cells and Natural Killer (NK)- cells.

Vadastuximab talirine (33A), a CD33-directed antibody drug conjugate, that has demonstrated an enhanced cytotoxic effect through upregulated CD33 expression has been investigated in multiple studies including a phase 1b study, which combines 33A with chemotherapy. The results of the study were reported here.

RG7356, a recombinant anti-CD44 immunoglobulin GI humanized monoclonal antibody that can inhibit cell adhesion. Results from a phase 1 dose-escalation study (NCT01641250), showed that RG7356 was generally safe and well tolerated in R/R AML patients.

The spotlight was then turned to agents targeting CD123, the Interleukin 3 α-chain receptor (IL3RA). CD123 is differentially and significantly over expressed in majority of patients with AML. Binding of Interleukin 3 (IL-3) to CD123 can induce hematopoietic progenitor cell differentiation, proliferation and also upregulation of anti-apoptotic proteins. Additionally, CD123 has been identified as a marker for Leukemic Stem Cells (LSCs), which are a small population of stem cells that have properties of differentiation, self-renewal and homeostatic control and they contribute to the maintenance and propagation of AML. In AML, LSC reservoir can lead to disease resistance, relapse and often death in patients.

Talacotuzumab (anti-CD123 monoclonal antibody), is an agent that targets CD123, which has showed limited clinical efficacy in R/R AML as a single agent. Currently, talacotuzumab is being investigated in a phase III study (NCT02472145) in combination with decitabine.

MGD006 is a Dual- Affinity Retargeting (DART) protein that was generated from antibodies to CD3 and CD123. MGD006 acts to redirect T cells via CD3, to target AML blasts cells expressing CD123. This interaction can mediate, target-effector cell association, T- cell activation, proliferation and receptor diversification.

MGD006 is being explored in a phase 1 dose-escalation study (NCT02152956), which is aiming to assess the safety and tolerability of this agent in R/R AML patients.

Induction of apoptosis

Then agents targeting apoptosis via either the inhibition of BCL-2 and other anti-apoptotic proteins or upregulation of wild type p53 were discussed next.

Idasanutlin, a MDM2 inhibitor, acts to prevent the p53-MDM2 interaction, thus leading to the activation of p53, hence apoptosis. Idasanutlin was investigated in combination with cytarabine in a phase 1b study in R/R AML patients. Results from this study showed that idasanutlin in combination with cytarabine resulted in durable CRs in AML patients.

BCL2 inhibitors including venetoclax have also been investigated in AML patients. In a phase 1 study, venetoclax in combination with low dose cytarabine in R/R AML patients was tolerable and induced durable remissions.

Immune checkpoints inhibitors in AML

Immune checkpoints are regulatory pathways induced in activated T-cells that regulate antigen responsiveness. These immune checkpoints are hijacked by tumors to promote dysfunction of anti-tumor effector cells and consequently of tumor escape from the host immune system. In AML, the PD1/PDL-1, a checkpoint pathway, is upregulated and this has a negative impact on disease control.

In a phase Ib/II study, a novel combination of two agents; azacytidine (AZA), Hypomethylating Agent (HMA), and nivolumab, a PD-1 inhibitor, was investigated in patients with relapsed AML. The rationale behind this study was to evaluate a mode of augmenting the anti-tumor and immunogenic effects of PD-1 inhibitors whilst combatting the problem of resistance in HMA therapy. Results from this study demonstrated that nivolumab/AZA combination resulted in a clinically significant increase in the median OS.

Currently, there are a range of immune checkpoint inhibitors being investigated in AML patients either as monotherapy or in combination with HMAs, 7+3 or other targeted agents. These include atezolizumab, pembrolizumab, and durvalumab.

Targeting effector immune cells

There are a range of agents that have been developed to activate NK cells, which can lead to the destruction of tumor cells, including IPH2101 and lirilumab.

Lirilumab, is a humanized anti-KIR monoclonal antibody that can block the interaction between KIRDL-1, -2, -3 inhibitory receptors and their ligands. Lirilumab is being investigated in multiple studies including the randomized phase II EFFIKIR study (NCT01687387) of lirilumab as maintenance treatment in elderly patients with AML in first complete remission.

The speaker then focused on the incorporation of these new agents with the 3 + 7 back bone as a “third drug” strategy. He spoke about prior successful studies including the phase III RATIFY trial (NCT00651261), which is a randomized trial in 717 newly diagnosed FLT3+ AML patients. In this study, patients were randomly assigned to receive either placebo or midostaurin 50mg orally twice daily on days 8–21 of each cycle of induction and consolidation chemotherapy followed by continuous daily midostaurin for up to 12 cycles. Midostaurin in combination with chemotherapy significantly improved the median OS of patients compared to placebo (74.4 vs 25.6 months, HR = 77, P = 0.0074). Midostaurin was recently approved by the U.S Food and Drug Administration (FDA)  for the treatment of newly diagnosed AML who are FLT3 mutated-positive in combination with standard cytarabine and daunorubicin and cytarabine consolidation.

Finally, the speaker spoke about novel-novel combination treatment in AML patients. He suggested that this chemo-free strategy represents a promising alternative for treatment of AML and discussed several ongoing studies using this combination including the phase I/II study (NCT02530476) of selinexor (nuclear export inhibitor) in combination with sorafenib (FLT3 inhibitor) in FLT3 mutated positive AML patients.

Professor Norbert Vey spoke to the AGP about his presentation at this educational session. He highlighted that 3 + 7 is “still the mainstay” of treatment of AML in frontline for fit AML patients. He spoke about how therapy for AML could go beyond the classical 3 + 7 regimen and added that “there is a great variety of new agents with different mechanisms of actions”. Some of these agents include agents targeting specific mutated genes, targeting specific pathways including apoptotic pathways. He emphasized that agents targeting apoptotic pathways are “interesting” as they have shown clinical efficacy and can form a backbone by which new regimens can be built most particularly, venetoclax and MDM2 inhibitors. He further added that addition of these new agents to the standard 3 + 7 “should lead to improvements of patient outcomes”.