Commitments to AML Research Uncover Numerous Targets in Relapsed/Refractory Disease

Naval G. Daver, MD

A host of novel targetable pathways is emerging in the acute myeloid leukemia (AML) research field on the heels of success seen with standard FLT3 and IHD1/2 inhibitors, signaling a new era of prolonged responses and improved quality of life (QOL) for patients with relapsed/refractory disease, according to Naval G. Daver, MD.

“There’s much more optimism and hope that we will be able to prolong life and improve or maintain a good QOL in [patients with] relapsed AML, which 5 to 10 years ago was not the case,” Daver said in an interview with OncLive^®.

In the interview, Daver highlighted ongoing research efforts in relapsed/refractory AML, including combination regimens with standard-of-care (SOC) therapies, pathways targeted by novel agents in early stages of development, and the evolving role of minimal residual disease (MRD) negativity as an end point in AML clinical trials.

Daver noted the prevalence of available FLT3 inhibitors in AML, including gilteritinib (Xospata), which was FDA approved in 2018 for adult patients with FLT3-mutated relapsed/refractory disease.¹ He also emphasized the importance of conducting continued research with novel agents such as the E-selectin inhibitor uproleselan (GMI-1271), which was investigated in a phase 3 trial (NCT03616470). Although the trial failed to meet its primary end point of overall survival (OS), with a median OS of 13 months uproleselan plus chemotherapy vs 12.3 months with chemotherapy alone, Daver explained that E-selectin inhibitors may still play a role in AML management, particularly in combination regimens.² 

Daver is a professor in the Department of Leukemia in the Division of Cancer Medicine, as well as director of the Department of Leukemia Division of Leukemia Research Alliance Program at The University of Texas MD Anderson Cancer Center in Houston.

OncLive: What are the greatest unmet needs for patients with relapsed/refractory AML?

Daver: The greatest need at this time [for patients] in the relapsed/refractory setting in AML is identifying novel targets for further improvement and expansion of available targeted therapies. A secondary [need] is developing combination treatments, which will likely lead to deeper and more durable remissions, hopefully improving survival. The third [need] is to identify novel immune pathways and immune mechanisms that can be leveraged toward targeting AML, which has been a challenge compared with other leukemias. With more gene modifications and the availability of double- and triple-targeting CAR T-cell therapy and CAR natural killer–cell therapy approaches, [these developments are] emerging.

The fourth [need is] to try to reduce the toxicities of therapies. The goal for most [patients with] relapsed/refractory AML is to move them to an allogeneic stem cell transplant, especially if they’ve not had a prior transplant. In some cases, even [patients] who have had a prior transplant [can undergo a] second transplant. Any treatment that has good efficacy but also good tolerability, preserving the ability of the patient to be a transplant candidate and eventually transition to transplant, would be preferred. These are the key [features] we’re looking for in new drugs and new agents in relapsed AML.

What data in AML need to be seen to push the needle forward in a recognizable way?

In relapsed/refractory [AML, there are] 3 key criteria we’re going to be looking for to see an improvement on currently existing SOC. One is the response rate and, specifically, the complete response [CR] rate or the CR/CR with partial hematologic recovery [CRh] rate, which has been established as a validated end point, even leading to FDA approvals both in acute lymphoblastic leukemia [ALL] and AML, and has been clearly associated with improved survival. That would be the first benchmark we look for: whether a treatment is giving CR/CRh rates [that are] encouraging or better than expected with current SOC.

The second would be to look at depth of response, which is usually assessed now using MRD. There are various ways to do this; the most common that has been used in the United States [US] has been multiparametric flow cytometry; however, we now have emerging molecular next-generation sequencing [NGS]–based approaches that may be more sensitive than flow cytometry, but [these] are only available for certain molecular markers, such as NPM1 mutations, FLT3 mutations, and KMT2A rearrangements. [We want to see whether] we can achieve deeper remissions in addition to higher CR/CRh rates because, historically, the most important factor that has correlated with improved event-free survival [EFS] and OS has been the achievement of deeper or [increased] MRD-negative responses.

The third factor is to look at improvement in OS. This is the gold standard and what most approvals have been based on, both in the relapsed and frontline settings; however, [Os] also takes the longest time [to evaluate]. We usually use CR, CR/CRh, and MRD as surrogate markers for us to know what to expect or to predict whether the OS will be improved. However, for final registration and approval, the FDA, as well as most expert investigators, want to see a delta improvement in OS at the end of the day to change the SOC.

In April 2024, the FDA’s Oncologic Drugs Advisory Committee voted in favor of the use of MRD as an accepted end point for accelerated approval in multiple myeloma. Will other hematologic cancers potentially follow suit in that regard?

As in multiple myeloma, I anticipate that MRD-based assessments may become a future regulatory pathway for AML [agents]. However, we need several prospective, randomized studies with longitudinal MRD collection analyses to establish MRD as a tight marker associated with EFS and OS. In AML, one of the challenges has been the type of MRD assessment [we use] compared with what we do in chronic lymphocytic leukemia, myeloma, and ALL. [In AML], we have historically used flow cytometry, which is effective, predictive, and sensitive when done in experienced labs with experienced hematopathologists. However, in the US, the number of such highly experienced labs and pathologists is limited. Additionally, flow cytometry has to be done on fresh samples, making it difficult for [MRD assessment] to be implemented in multicenter, international, randomized studies. Flow cytometry also requires a baseline bone marrow assessment so we can compare the subsequent longitudinal samples to the baseline to identify differences in immunophenotype.

In the past 5 to 6 years, we have started moving more toward using molecular NGS. Especially for FLT3 mutations, NPM1 mutations, and KMT2A arrangements, as well as a few others, there are now emerging data with multiple retrospective and prospective datasets showing that these markers can be strongly associated with EFS and OS. These [assessments] can be done on stored samples, so they can be implemented in multinational, international, randomized studies. Additionally, [these assessments] do not need baseline bone marrow [samples] for comparison, which makes them much more attractive [vs flow cytometry], especially when we think about using them in future randomized registrational studies.

In the future, we will see, like in myeloma, the potential use of MRD for academic exercises, as well as for regulatory approvals and maybe for clinical treatment decision-making, such as [determining] patients who need maintenance post-transplant vs [those who do] not, as well as the patients who may need a transplant in the first remission vs those who may not need it. There are a lot of efforts in this direction, but it’s still probably a few years away. With several ongoing prospective studies implementing this, we may have a lot more data in the next few years to be able to push toward using [MRD negativity] as a regulatory marker [in AML].

Based on the mode of disease in AML, what are the most relevant pathways for targeted therapy development?

At this point, when we talk about targeted therapies in AML, one [mutation] we’re focusing on is FLT3. Multiple FLT3 inhibitors have been approved: midostaurin [Rydapt] and quizartinib [Vanflyta] in the frontline setting and gilteritinib in the salvage setting. Beyond that, numerous combinations with FLT3 inhibitors, such as FLT3 inhibitors plus venetoclax [Venclexta] or FLT3 inhibitors plus chemotherapy, have promising activity and higher response rates than what we saw with single-agent [FLT3 inhibitors], as well as more durability and the ability to get more patients to transplant. FLT3 is probably the most important molecular marker for targeted therapies in AML, and [FLT3 inhibitors] happen to be the most common mutation seen in AML, [present in approximately] 30% to 35% of patients, making FLT3 an important, targetable pathway.

We also have IDH1 and IDH2. The IDH1 inhibitors olutasidenib [Rezlidhia] and ivosidenib [Tibsovo], as well as the IDH2 inhibitor enasidenib [(Idhifa), have been] approved in the US as single agents in the relapsed setting, giving approximately 30% to 40% response rates. At MD Anderson, we have been developing combinations with [IDH inhibitors and] venetoclax, azacitidine [Vidaza]/venetoclax, and chemotherapy, showing that with combinations, we can approximately double the CR/CRh rates [observed with IDH inhibitors alone] and more importantly and impressively, improve the MRD depth of response. We think this will lead to better outcomes than just single-agent [treatment].

Most recently, there has been an exciting development with the emergence of a group of targeted therapies called menin inhibitors that target KMT2A/MLL rearrangements, as well as NPM1 mutations. In the past 2 years, 4 menin inhibitors have gone into clinical development, [which are made by] Syndax Pharmaceuticals, Kura Oncology, Sumitomo Pharma, and Johnson & Johnson. All 4 of these [agents] are inducing clinical activity. They have differences in their drug-drug interactions, dosing, kinetics, frequency of differentiation syndrome, and safety. However, all of them are active, and there may be a role for all these drugs in different specific patient populations and with different combinations.

At this point, for approximately 50% to 55% of [patients with] AML, we may be able to identify an [actionable] aberration [and use] these targeted therapies. However, there is a need to develop more. We are looking at other targeted therapies that will target the RAS/MAPK pathway, which is becoming a major mechanism of resistance to the standard therapies we’re using in AML today, including venetoclax, FLT3 inhibitors, and IDH inhibitors. For the most difficult subset [of patients to treat—those with] TP53 mutations—there are [several research] efforts [underway]. In the near future, there may be TP53-specific targeted agents, which would be important. In the next 5 to 10 years, we’re going to continue to see the development of targeted therapies that will hopefully cover most, if not all, subsets of AML.

What might agents targeting the E-selectin, PKL1, and MDM2 pathways add to the AML treatment paradigm?

[Agents targeting] E-selectin are probably the most advanced [of the drugs targeting those 3 pathways] in clinical development. Uproleselan was evaluated in combination with intensive chemotherapy in relapsed/refractory AML in a randomized phase 3 study. This study evaluated a combination of FAI [fludarabine, cytarabine, and idarubicin] or MEC [mitoxantrone, etoposide, and cytarabine], which are intensive chemotherapy combinations, with or without uproleselan in relapsed/refractory AML. The primary endpoint was OS.

[We had hoped that this trial, if positive, would] lead to the approval of uproleselan in combination with intensive chemotherapy as an effective [agent] in relapsed AML. From there, [uproleselan] could be moved into various combinations in the frontline setting with hypomethylating agents [HMAs] plus venetoclax, intensive chemotherapy plus FLT3 inhibitors, intensive chemotherapy plus gemtuzumab ozogamicin [Mylotarg], etc. [Although this phase 3 trial did not meet its OS end point], the nice aspect of E-selectin is that it is ubiquitously expressed in AML and does not require a specific biomarker or molecular cytogenetic aberration. [E-selectin inhibitors] could be effective add-on therapies across the board, and interestingly, seem to reduce rates of mucositis and toxicity, making these agents even more attractive to add to existing SOC therapies. I’m excited about this mechanism and approach and improvement in safety.

MDM2 is an interesting pathway that I’ve personally been involved with. We have done a lot of preclinical research at MD Anderson under Michael Andreeff, MD, PhD. Marina Konopleva, MD, PhD, [of the Montefiore Einstein Comprehensive Cancer Center in Bronx, New York], and I have been involved with clinical trials with MDM2 inhibitors. We’ve evaluated several of these initially with the nutlins, then idasanutlin [RG7388], as well as milademetan [RAIN-32] from Daiichi Sankyo, and others. Unfortunately, the main issue that has halted the development of MDM2 inhibitors has been toxicity, specifically gastrointestinal [GI] toxicity in the form of diarrhea, nausea, vomiting, and poor appetite. Despite good clinical activity, single-agent responses, and good responses [seen when] combining MDM2 inhibitors with the BCL2 inhibitor venetoclax, this is a field that has not yet come into its prime time.

A new group of drugs, MDM2 degraders, has entered the field. We’re working with 1 of them from Chimera Therapeutics. We hope to present some of those data at meetings in the summer [of 2024]. We are seeing interesting, promising activity with this agent both in myeloid leukemia, as well as in NPM-mutated myelofibrosis and solid tumors. The MDM2 degraders tend to be 80- to 100-fold more potent than the MDM2 inhibitors, and often, because they have a rapid on-off switch with a short half-life of approximately 6 to 8 hours, do not seem to bind to the GI MDM2 [receptors] at a continuous rate, potentially resulting in less GI toxicity. We are seeing this in the clinical setting. These degraders may eventually become the way the MDM2 pathway finds its role in the treatment of myeloid diseases, and even potentially beyond that in solid tumors.

The PLK inhibitors are early [in development]; we don’t [yet] have clear clinical data [with these agents]. The PLK pathway has been shown by many different labs and groups to be preclinically important in AML. [Investigations with several] combinations of PLK inhibitors with venetoclax and with HMAs are ongoing. At this time, I do not have a clear sense of where these studies are and the future development path [of these agents], but it is definitely an important pathway in AML development.

With the emergence of additional targeted agents and more personalized chemotherapeutic regimens, how has the discussion around survivorship and AML changed or evolved?

At this time, [we are asking about] survivorship in AML, which is progress in itself because 10 years ago, if a patient had progressed on frontline standard therapy, such as intensive chemotherapy [with an] anthracycline plus cytarabine, the treatment options were extremely limited. We did not have targeted therapies, FLT3 inhibitors, IDH inhibitors, menin inhibitors, venetoclax, or HMAs plus venetoclax. The options were to use HMAs alone with an expected response rate of 15% to 18% in the salvage setting, or low-dose cytarabine, with an even lower expected response rate of 10% to 15%. Approximately 90% of patients had palliative outcomes, and more than half of patients were transitioned to or considered to be transitioned to palliative care in the first salvage setting. This has now changed dramatically.

Today, for relapsed AML, especially for [first and second] salvage, at MD Anderson, we offer treatment to 99% of patients. We look for molecular, targetable aberrations, such as FLT3, IDH1, IDH2, NPM1, MLL, new fusions and translocations, and other rare but targetable fusions and rearrangements. We can find these in approximately 55% to 60% of patients, giving us targeted therapy options, which can give us response rates anywhere from 40% to 80%. [This is] encouraging.

[These developments open] the path for potential transition to allogeneic stem cell transplant. Even for patients who do not have an identified targetable mutation, aberration, rearrangement, or fusion, we are considering venetoclax-based treatments, such as HMAs plus venetoclax, or potentially intensive chemotherapy plus venetoclax if patients did not have venetoclax in the frontline setting. With this approach, we get approximately 40% to 70% response rates, [which is] encouraging.

The overall perception of and approach to [managing] relapsed/refractory AML has changed from one of nihilism and supportive, palliative care 10 or 15 years ago, to a more encouraging outlook with targeted therapies and especially moving target therapy combinations, with the hope that a lot of these patients still have a potential to be cured. That being said, cures in relapsed/refractory AML are predominantly in younger patients with targeted mutations who can be transitioned to transplant, so we still have a lot of work to do. Immunotherapies are going to be critical in relapsed/refractory AML, just as they have been in ALL, multiple myeloma, and diffuse large B-cell lymphoma with the CAR T-cell therapy approaches. [Immunotherapy options in AML are] lacking, but there is an intense effort and research ongoing to develop these in the relapsed setting.

References

1. FDA approves gilteritinib for relapsed or refractory acute myeloid leukemia (AML) with a FLT3 mutation. FDA. November 28, 2018. Accessed May 8, 2024. https://www.fda.gov/drugs/fda-approves-gilteritinib-relapsed-or-refractory-acute-myeloid-leukemia-aml-flt3-mutation
2. GlycoMimetics announces results of pivotal phase 3 study of uproleselan in relapsed/refractory (R/R) acute myeloid leukemia (AML). News Release. GlycoMimetics, Inc. May 6, 2024. Accessed May 6, 2024. https://ir.glycomimetics.com/news-releases/news-release-details/glycomimetics-announces-results-pivotal-phase-3-study