The treatment landscape for acute myeloid leukemia (AML), once considered a graveyard for drug development, has undergone a rapid evolution in the last three years, signaled by the approval of eight new drugs  and the start of nearly 200 industry-sponsored clinical trials, many of them for combination therapies.* While I believe these developments will ultimately lead to a brighter future for the treatment of AML, they have created a period of uncertainty in which physicians and industry leaders are figuring out how best to develop and use these therapies alone and in combination, and in which patients.
The uncertainty is especially notable in the setting of newly diagnosed unfit AML. Although three of the drugs approved since 2017 are indicated for this population, they are not curing the disease and each comes with pros and cons, leaving a still-unmet need. The situation is starker in relapsed or refractory AML (RR-AML): the four drugs approved since 2017 for this population are indicated for a disease that harbors specific markers or mutations. While these are beneficial advances for the corresponding subsets of patients, RR-AML patients who lack these markers still have few options and median overall survival remains low at less than six months.
I believe this period of uncertainty will ultimately lead to a treatment landscape where available drugs – current and future – are used in various combinations and sequences for specific subsets of patients, thereby transforming AML from a deadly disease into a more chronic and manageable one. To this end, a key feature for the success of any new drug entering the AML landscape will be “combinability,” a feature determined by having a unique mechanism of action (MOA), a favorable tolerability profile, and marker(s) that define the patients likely to respond.
Here, I’ll examine the need in newly diagnosed unfit AML, and RR-AML for well-tolerated targeted therapies that improve outcomes and offer Quality of Life (QoL), and discuss why new therapies will become increasingly important as the trend towards combination therapies and the ability to predict patient responses reshape the treatment landscape.
The unmet need in two AML settings
In newly diagnosed AML patients, high age (≥75) has traditionally been a determinant of “unfitness” for intensive induction chemotherapy, but newer definitions incorporate additional criteria, including severe comorbidities.  Nevertheless, the basic concept remains unchanged: unfit patients are sicker than fit patients. Thus, intensive induction chemotherapy that is the mainstay of AML treatment for fit patients may do more harm than good in unfit patients.
More importantly, the therapeutic objectives for unfit patients are different than for fit patients. The goal is not to get unfit patients to the point of a cure but instead to control their disease so they can live longer with a better QoL. To that end, the tolerability of a therapy is crucial.
Since 2017, FDA has approved the BCL2 inhibitor venetoclax (Venclexta®; AbbVie/Genentech) the hedgehog pathway inhibitor glasdegib (Daurismo®; Pfizer), and the IDH1 inhibitor ivosidenib (Tibsovo®; Agios) for newly diagnosed unfit patients. The first two are indicated regardless of mutation status; the last is only for disease harboring IDH1 mutations.
These emerging therapies are exciting and offer a greater benefit than earlier therapies, but they come with certain trade-offs. For example, composite complete response (CR/CRi) rates of 66.4% have been reported for venetoclax/azacitidine (Vidaza®; Celgene), meaning approximately one-third of patients do not achieve this response.  A majority of patients (82%) have high grade (grade 3 or 4) hematologic AEs, including various cytopenias and febrile neutropenia; this level of combination toxicity may not be easily tolerated by newly diagnosed unfit patients. Lastly, median overall survival among patients who are refractory to or relapse after venetoclax/HMA therapy appears to be worse at 2.4 months than in the overall RR-AML patient population.
Additionally, although venetoclax is for “all-comers” in the newly diagnosed unfit setting, studies have begun to uncover markers of response and resistance to it. For example, high response to venetoclax has been associated with NPM1, RUNX1, and SRSF2 mutations. Poor response has been associated with FLT3 internal tandem duplications (FLT-ITD), mutations in TP53, NRAS, and DNMT3A, and an AML monocytic phenotype.  Primary and acquired resistance have been linked to loss of BCL2 expression and concomitant overexpression of MCL1.  Further research is needed to better define which patients are the best candidates for these drugs.
Together, these factors underscore the ongoing need to find new approaches that can offer durable, well-tolerated responses, with unique MOAs so we can target treatment of patients with combinations that are most likely to work for them.
The unmet need in RR-AML is perhaps clearer to see because the drugs approved since 2017 are indicated for disease-carrying certain markers or mutations: gemtuzumab ozogamicin (Mylotarg®; Wyeth/Pfizer). for CD33-positive AML; gilterinib for FLT3-mutant AML; and ivosidenib and enasidenib (Idhifa®; Celgene**/Agios) for IDH1– or IDH2-mutant disease. Relapsing patients whose disease lacks these markers have few options.
CR/CRi rates for these four targeted therapies are in the 20-35% range with a duration of 4-8 months. Emerging data also speaks to the importance of sequencing therapies. For example, the majority (~90%) of patients relapse after treatment with an IDH inhibitor, and the emergence of the opposite mutation – swapping mutant IDH1 for IDH2 or vice versa – or other mutations for which there are no targeted therapies is becoming increasingly recognized. 
It should also be noted that in the RR-AML setting, patients come from both the fit and unfit groups, further complicating treatment decisions.
The solution for addressing the unmet needs of both newly diagnosed unfit and relapsing AML patients is two-fold. First, it involves identifying the optimum combination of therapies for each subset of patients. Second, it requires the development of drugs that have the “combinability” needed to make them flexible tools in the AML armamentarium.
The growing importance of combinations
There is clearly an emerging paradigm supporting the use of combination therapies in AML, consistent with therapies in most other malignancies. Many of the newer targeted agents are being tested in clinical trials in combination with each other and/or with chemotherapies in newly diagnosed fit and unfit populations and in RR-AML.
Combinations are critical to preventing the emergence of drug-resistant clones. It is thought that most patients’ disease harbors multiple mutations, even if only one is detected prior to treatment. For example, about 20% of patients with IDH1 or IDH2 mutations also have FLT3 mutations. And, as noted above, relapsing patients can exhibit new mutations they did not have, or which weren’t detected, prior to treatment.
This underscores the importance of attacking cancer simultaneously and early on multiple fronts. By deploying combinations as the first line of treatment, we could potentially get the greatest impact for the patient by eradicating cancer when there are fewer cells.
However, to be “combinable” with existing agents and provide the maximum benefit in the emerging combination landscape, any new AML drug should ideally have three key features: (1) an MOA that doesn’t overlap with other drugs; (2) a good safety profile, preferably distinct from the combination partner(s), so combinations do not increase the toxicities seen with the single agents used in combination; and (3) markers to define which patients are likely to respond.
These are the principles underpinning our AML program at Syros, which is currently in a Phase II trial in newly diagnosed unfit AML and RR-AML patients. Based on preclinical and clinical data, we believe that SY-1425, an oral potent, and selective retinoic acid receptor alpha (RARα) agonist, has broad potential as a combination agent for genomically defined patients.
First, SY-1425 has an MOA that is distinct from other drugs on the market or in development. RARα, a transcription factor encoded by the RARA gene, is important in myeloid differentiation. In a previously unidentified subset of AML patients, a super-enhancer drives abnormally high expression of RARA, contributing to the disease. By binding to RARα, SY-1425 promotes differentiation and decreases proliferation of AML cells. 
Second, SY-1425 has demonstrated a favorable tolerability profile that supports combinability. Data to date from the ongoing trial show SY-1425 combined with azacitidine is highly active in newly diagnosed unfit AML patients who are RARA-positive and that the combination is generally well-tolerated with no evidence of increased toxicities beyond what is seen with either agent alone. In preclinical studies, SY-1425 has also shown synergy or additivity with other commonly used AML agents.
Third, preclinical and clinical data support that our biomarker for the RARA super-enhancer can be used to select patients for treatment with SY-1425. Based on our screens of about 350 patients in connection with our clinical trial, we believe that about 30% of AML patients are RARA-positive. Notably, the presence of the RARA biomarker overlaps with, but appears to be independent of, other AML mutations, pointing to potential combinations with other targeted agents.
As our industry looks to transform AML from a deadly disease into a chronic one, we will need an arsenal of therapies, both broad-acting and targeted, that can be used in multiple combinations at multiple stages of the disease. Meeting the needs of each and every AML patient depends on bringing more drugs to market that are combinable, as well as conducting more research to understand which patients are likely to respond to which drugs and how best to sequence therapies for maximum benefit.
* Based on updated www.clinicaltrials.gov search on July 20, 2020. Trial starts since January 1, 2017 (3 yrs 7 mos): for all sponsors, 518 trials, 234 for combinations; for industry sponsors, 260 trials, 121 for combos. Trial starts for the preceding period (June 1, 30, 2012 to Dec. 31, 2016; 3 yrs 7 mos): for all sponsors, 420 trials, 175 for combos; for industry sponsors, 212 trials, 100 for combos. Search parameters: interventional, Phase I to Phase III trials, all recruitment statuses, using indicated date ranges for Start Date.
** Celegene is now part of Bristol Meyers Squibb
A Biomarker-Directed Phase 2 Trial of SY-1425 in Patients With Acute Myeloid Leukemia or Myelodysplastic Syndrome – NCT02807558
 Richard-Carpentier G, DiNardo CD. Single-agent and combination biologics in acute myeloid leukemia. Hematology Am Soc Hematol Educ Program. 2019;2019(1):548-556. doi:10.1182/hematology.2019000059 [Pubmed][Article]
 Ferrara F, Barosi G, Venditti A, et al. Consensus-based definition of unfitness to intensive and non-intensive chemotherapy in acute myeloid leukemia: a project of SIE, SIES and GITMO group on a new tool for therapy decision making. Leukemia. 2013;27(5):997-999. doi:10.1038/leu.2012.303 [Pubmed][Article]
 DiNardo C., Jonas B, Pullarkat V., Thirman M., Garcia J., Wei A., Döhner H., Fenaux P., Recher C., et al. A randomized, double-blind, placebo-controlled study of venetoclax with azacitidine vs azacitidine in treatment-naïve patients with acute myeloid leukemia ineligible for intensive therapy-VIALE-A. Presented at the 25th EHA Annual Congress (2020 [Abstract]
 Maiti A, Rausch CR, Cortes JE, et al. Outcomes of relapsed or refractory acute myeloid leukemia after frontline hypomethylating agent and venetoclax regimens [published online ahead of print, 2020 Jun 4]. Haematologica. 2020;haematol.2020.252569. doi:10.3324/haematol.2020.252569 [PubMed][Article]
 Wang YW, Tsai CH, Lin CC, et al. Cytogenetics and mutations could predict outcome in relapsed and refractory acute myeloid leukemia patients receiving BCL-2 inhibitor venetoclax. Ann Hematol. 2020;99(3):501-511. doi:10.1007/s00277-020-03911-z [Pubmed][Article]
 DiNardo CD, Tiong IS, Quaglieri A, et al. Molecular patterns of response and treatment failure after frontline venetoclax combinations in older patients with AML. Blood. 2020;135(11):791-803. doi:10.1182/blood.2019003988 [Pubmed][Article]
 Pei S, Pollyea DA, Gustafson A, et al. Monocytic Subclones Confer Resistance to Venetoclax-Based Therapy in Patients with Acute Myeloid Leukemia. Cancer Discov. 2020;10(4):536-551. doi:10.1158/2159-8290.CD-19-0710 [Pubmed][Article]
 Kuusanmäki H, Leppä AM, Pölönen P, et al. Phenotype-based drug screening reveals association between venetoclax response and differentiation stage in acute myeloid leukemia. Haematologica. 2020;105(3):708-720. doi:10.3324/haematol.2018.214882 [Pubmed][Article]
 Zhang, H., Wilmot B., Bottomly D., Kurtz S.E., Eide C.A., Damnernsawad A., Romine K., Patel S., et al. Biomarkers Predicting Venetoclax Sensitivity and Strategies for Venetoclax Combination Treatment. Blood 132 (Supplement 1):175 (2018) doi.org/10.1182/blood-2018-175 [Article]
 Knight T, Edwards H, Taub JW, Ge Y. Evaluating venetoclax and its potential in treatment-naïve acute myeloid leukemia. Cancer Manag Res. 2019;11:3197-3213. Published 2019 Apr 23. doi:10.2147/CMAR.S180724 [Pubmed][Article]
 Campos EDV, Pinto R. Targeted therapy with a selective BCL-2 inhibitor in older patients with acute myeloid leukemia. Hematol Transfus Cell Ther. 2019;41(2):169-177. doi:10.1016/j.htct.2018.09.001 [Pubmed][Article]
 Stein, E. Targeting IDH1/IDH2.Presented at the European School of Haematology (ESH) Conference on AML (Oct. 26, 2019)
 Harding JJ, Lowery MA, Shih AH, et al. Isoform Switching as a Mechanism of Acquired Resistance to Mutant Isocitrate Dehydrogenase Inhibition. Cancer Discov. 2018;8(12):1540-1547. doi:10.1158/2159-8290.CD-18-0877 [Pubmed][Article]
 Boddu, P., et al. Influence of IDH on FLT3-ITD status in newly diagnosed AML.Leukemia 2017; 31:2526-2529 as cited in: Cortes, J., et al. Activity of Multiple Targetable Therapies in FLT3-Mutated (mu) Acute Myeloid Leukemia (AML) Patients (pts) with Concurrent Isocitrate Dehydrogenase Mutation (IDHm).Presented at the 61st American Society of Hematology (ASH) Annual Meeting (2019)
 McKeown MR, Corces MR, Eaton ML, et al. Superenhancer Analysis Defines Novel Epigenomic Subtypes of Non-APL AML, Including an RARα Dependency Targetable by SY-1425, a Potent and Selective RARα Agonist. Cancer Discov. 2017;7(10):1136-1153. doi:10.1158/2159-8290.CD-17-0399 [Pubmed][Article]
 Wang E, Aifantis I. Targeting the Noncoding Genome: Superenhancers Meet Their Kryptonite. Cancer Discov. 2017;7(10):1065-1066. doi:10.1158/2159-8290.CD-17-0860 [Pubmed][Article]
 McKeown MR, Johannessen L, Lee E, Fiore C, di Tomaso E. Antitumor synergy with SY-1425, a selective RARα agonist, and hypomethylating agents in retinoic acid receptor pathway activated models of acute myeloid leukemia. Haematologica. 2019;104(4):e138-e142. doi:10.3324/haematol.2018.192807 [Pubmed][Article]