Uproleselan Combinations Contribute to Robust Investigational Profile in AML

The E-selectin inhibitor uproleselan (GMI-1271) has shown preliminary efficacy in patients with acute myeloid leukemia (AML), and E-selectin could represent one of several novel targets for treatment of this patient population, according to Michael Andreeff, MD, PhD.

Previously, a phase 1/2 trial (NCT02306291) investigating uproleselan plus mitoxantrone, etoposide, and cytarabine (MEC) showed that the combination elicited a 41% remission rate and a median overall survival of 8.8 months in patients with relapsed/refractory AML.¹ The findings from this study provided the rationale for an ongoing phase 3 trial (NCT03616470) evaluating uproleselan plus MEC or fludarabine, cytarabine, and idarubicin vs chemotherapy alone in patients with relapsed/refractory disease.²

In addition, an ongoing phase 1 trial (NCT04964505) is evaluating the safety and preliminary efficacy of uproleselan plus azacitidine and venetoclax (Venclexta) in patients with treatment-naïve AML.³

“We have a long way to go, but we shouldn’t forget that we have many new drugs available now that we didn’t have 10 or even 5 years ago. We are finally on the path to cure AML,” Andreeff said in an interview with OncLive^®.

In the interview, Andreeff discussed the potential benefits of E-selectin inhibition in the AML tumor microenvironment, data that have contributed to the further development of uproleselan in this population, and additional AML targets under investigation that may broaden the treatment landscape.

Andreeff is a professor of medicine in the Department of Leukemia in the Division of Cancer Medicine, director of the Flow Cytometry and Cellular Imaging Facility, the Paul and Mary Haas Chair in Genetics, director of the Bone Marrow Aspiration Clinic in the Department of Hematology in the Division of Medicine, and chief of the Section of Molecular Hematology and Therapy in the Division of Cancer Medicine at The University of Texas MD Anderson Cancer Center in Houston.

OncLive^®: What role does E-selectin play in the tumor microenvironment in AML?

Andreeff: Most leukemia research focuses on the leukemic cell. Many leukemic stem cells are defined by several markers. The role of the microenvironment has been under-investigated. [AML cells] have a microenvironment, which in this case is the bone marrow, a complicated organ with many cells.

In the past few years, people have paid much attention to the immune microenvironment, which is important, but other cell types are equally important. One cell type is the mesenchymal stem cell. A major target there is CXCR4, and the ligand is SDF-1. SDF-1 is produced by stromal cells, and CXCR4 is on the leukemic cells and leukemic stem cells.

The vascular niche has not been well investigated. There is no clear distinction between the endosteal niche and the vascular niche. It is 1 composite niche. Functionally, it’s 1 organ. One factor of the vascular niche is E-selectin. There was no drug targeting this, but GlycoMimetics made a compound targeting E-selectin that is in clinical trials.

The phase 3 trial is ongoing. Results will probably take another 9 months, but the phase 2 study looked good. The responses were better, [indicating] a deeper response to induction therapy and longer remission durations because the interaction between the leukemic cells and the vascular niche is disrupted. Additionally, [we saw] a myeloprotective effect. In the [phase 2] trial, neutrophils recovered faster in the presence of E-selectin inhibition.

When we started working on this years ago, we observed many healthy megakaryocytes in the marrow. That [indicated] that the leukemic cells are more susceptible to chemotherapy and perhaps venetoclax, but also that the normal hematopoietic cells are, to a degree, protected, which would shorten the time to platelet and neutrophil recovery.

How does uproleselan work in combination with chemotherapy to amplify responses in patients with AML?

Uproleselan is similar to SDF-1 and CXCR4 that link the stromal cells to the leukemic cells. GlycoMimetics has a combination drug, which is not yet in clinical trials, that inhibits both CXCR4 and E-selectin. [Preclinical] studies have shown that [CXCR4/E-selectin] combination [inhibitor] is much more effective than the E-selectin inhibitor. But you need to take small steps before you can run. Currently, the important event will be the outcome of the randomized phase 3 trial [of uproleselan]. If that’s positive, they will [likely] put the combined inhibitor into clinical trials.

We have investigated the stromal cells in the microenvironment for 18 years. CXCR4 signals into leukemic cells via the transcription factor Yin Yang 1 [YY1], which is immediately downstream of CXCR4. The signaling downstream of E-selectin is not YY1, so these are alternative pathways. However, blocking E-selectin detaches the cells from the vascular niche and affects intracellular signaling pathways. Ideally, we want to cut off both supply structures in the marrow: the mesenchymal stem cells and the vascular niche. That can be achieved with the dual inhibitor.

What is the significance of minimal residual disease (MRD) negativity in AML?

A myriad of studies show that the presence of MRD is a bad prognostic sign. MRD supersedes all other prognostic factors. It means there is leukemia left [in the patient]. It’s established that at the end of cycle 1 or cycle 2, [MRD] is a better measure of leukemic cell sensitivity. However, [around] 40% of patients who are MRD negative by flow cytometry [will] relapse. [We need to improve] how we determine MRD.

Now, we routinely look for mutations by polymerase chain reaction [PCR], [such as] FLT3 and NPM1. That increased sensitivity [can] detect MRD down to [about] 10^–4 [leukocytes], whereas flow cytometry [detects MRD] down to 10^–3 [leukocytes]. We are now picking up positivity that we would miss by flow cytometry. In the future, this may be done by bidirectional, error-corrected, deep DNA sequencing. That is probably the best method. Other methods are droplet digital PCR, which is also sensitive, maybe down to 10^–5 [leukocytes]. But for that, we need to know what to look for.

Flow cytometry can probably be improved, and 2 major projects are ongoing. One is from the National Cancer Institute, which [has] a consortium to standardize MRD detection. There is also an organization called Break Through Cancer, a foundation that has brought together the Massachusetts Institute of Technology [in Cambridge], Dana-Farber Cancer Institute [in Boston, Massachusetts], Memorial Sloan Kettering Cancer Center [in New York, New York], Johns Hopkins University [in Baltimore, Maryland], and MD Anderson. We are comparing different techniques, [such as] single-cell sequencing, improved flow cytometry, and soluble DNA mutations that are not cell restricted but are in the plasma. In solid tumors, this is being established as a sensitive way to look for residual cells. Mutations are seen in ways they are not seen when we sequence individual cells. That means there may be a metastasis somewhere with a different mutation that is spilling DNA into circulation that can be detected. [Additionally], single-cell sequencing [can be done] after the stem cells [have been sorted].

The other aspect is to molecularly characterize the MRD cells. What is overexpressed in these cells? [These findings will differ] between treatment protocols. The quality of MRD cells in [a patient receiving] cytarabine-based treatment [may be] different. Different genes [may be] overexpressed compared [with those in a patient] treated with venetoclax-based therapies.

We have worked on the Break Through Cancer project for a few years. One exciting possibility is to detect neoantigens on these cells that are not present when leukemia is diagnosed. Immunotherapy in AML has been limited in its efficacy. CAR T cells, engager T cells, natural killer cells, and antibody-drug conjugates each have about a 20% response rate. [Those response rates differ from those in] lymphoid malignancies. We need better targets than CD123, CLL-1, and CD33, [which are] not specific for leukemic cells. Once we map the surface of residual cells, we will likely discover new epitopes that immune cells [could be] directed against. [This may] take another decade to come to fruition. Induction therapy kills [around] 3 or 4 logs of cells, [and afterward], it will be easier for immune therapy to mop up the residual cells.

What other novel targets under investigation in AML could lead to more targeted therapies for this population?

We have learned how to target a few mutations [over the past 15 to 20 years]. FLT3 inhibitors have had an influence, but they are not curative. Even after a transplant, patients need a FLT3 inhibitor to prevent relapse. IDH1/2 are effective [targets]. In those cases, venetoclax is effective. IDH2 mutations treated with venetoclax plus a hypomethylating agent [HMA] can be curative to a degree because these cells are highly sensitive to BLC-2 inhibition.

We don’t have a good NRAS inhibitor yet. PTPN11 [mutations] are a problem, and we don’t have anything specific [to treat] that. SHIP2 inhibitors are under investigation. We are running out of targets.

p53 is the highest-hanging fruit, and no drugs target p53. Eprenetapopt [APR-246] did not change the conformation of p53. This drug induced reactive oxygen species [activity], which killed leukemic cells. It has activity, but does not preferentially kill p53-mutant cells, and clinical trials have been negative.

One drug, developed by PMV Pharma, targets a specific p53 mutation, Y220C. One percent of [patients with] solid tumors, and probably the same [proportion] of [those with] leukemias, have this mutation. This mutation lacks a portion in its DNA binding domain, so it doesn’t allow the transcription activation that p53 wild-type does. It’s a small region filled by a small molecule. This agent induced massive apoptosis preclinically. It is in clinical trials now in solid tumors and was presented at the [2022 ASCO Annual Meeting]. [Five of 21 patients achieved a partial response]. There were no complete responses, but it was a good first step. We need to develop combination therapies. That was the first time a p53 mutation was converted into a wild-type conformation, which then signaled apoptosis.

Other effective but nonspecific drugs for p53 mutations are magrolimab and other CD47/SIRPα inhibitors. Patients with p53 mutations have a high response rate of around 50% [with these agents], but they relapse steadily. We need new approaches for MRD following CD47 antibody treatment, which causes the “eat me” signal. The macrophages are agnostic and do not [specifically] eat p53[-mutated] cells.

Those are the best therapies we have for [patients with] p53 mutations, but they are not the end of the story. The relapse rate is high. We now have combinations in clinical trials with venetoclax and HMAs. HMAs upregulate CD47, and venetoclax lowers the apoptotic threshold. [These studies] will be a step forward, but we still have a way to go with p53 mutations.

Splice site mutations are being targeted. Omar Abdel-Wahab, MD, of Memorial Sloan Kettering Cancer Center, has interesting drugs under development. There are many other new targeted agents, but they’re all early.

[Targeting] RAS [alterations] would be nice. We are working on targeting NRAS [mutations] in addition to KRAS [mutations]. KRAS [mutations] are rare in AML. We need an NRAS- or a pan-RAS–targeted agent.

MYC is an interesting target that has not been successfully targeted. We have presented data at [the 2022 ASH Annual Meeting] for a MYC degrader we are developing. That [work] is early and preclinical, and it will take time to bring into the clinic. About 20% of the transcriptome depends on MYC. Once we can target it, we will have effective treatments.

The other big drug development is menin inhibition. MLL fusion leukemias had been horrible. When patients relapse, there was no effective treatment. Now, [with menin inhibitors], we have a response rate of 30%, and 90% of patients are MRD negative. These are drugs developed by Syndax Pharmaceuticals and other companies.

Interestingly, downstream are NPM1-mutant cells, which we couldn’t specifically target until now. NPM1 mutations may be targeted by inhibiting menin. A Nature paper was published on this that defined the resistance mechanism as a mutation in the menin gene. We know what to look for now when we see resistance to these menin inhibitors. The future is moving fast, and we still have a long way to go to cure all patients.

How might uproleselan and other targeted therapies in development move the AML treatment paradigm forward?

We need more patients in clinical trials. In the United States, few patients are in clinical trials, and the number is going down. The best way to develop these drugs is to [enroll patients in] phase 1/2 trials with good molecular end points. Patients in phase 1/2 trials have longer life expectancies than patients [not enrolled in these trials], because they get better supportive care, they may benefit from these drugs at least temporarily, and they get undivided attention. [In contrast, a patient not enrolled in a trial] is being sent home and is maybe on low-dose cytarabine or an HMA only, which has no effect on survival. I encourage everybody with patients to refer at least their patients with refractory leukemia to centers with clinical trials.

References

1. DeAngelo DJ, Jonas BA, Liesveld JL, et al. Phase 1/2 study of uproleselan added to chemotherapy in patients with relapsed or refractory acute myeloid leukemia. Blood. 2022;139(8):1135-1146. doi:10.1182/blood.2021010721
2. Study to determine the efficacy of uproleselan (GMI-1271) in combination with chemotherapy to treat relapsed/refractory acute myeloid leukemia. ClinicalTrials.gov. Updated April 6, 2022. Accessed April 26, 2023. https://clinicaltrials.gov/ct2/show/NCT03616470
3. Uproleselan, azacitidine, and venetoclax for the treatment of treatment naïve acute myeloid leukemia. ClinicalTrials.gov. Updated August 15, 2022. Accessed April 26, 2023. https://clinicaltrials.gov/ct2/show/NCT04964505