Video
Novel Targets and Biomarkers in AML
Author(s):
Harry P. Erba, MD, PhD: I want to move on; I want to get to some big investigational agents that are important. Before getting there, I will mention that there were a number of studies looking at BCL2 inhibition with venetoclax in combination with intensive chemotherapies like CPX-351 [cytarabine and daunorubicin] and idarubicin-FLAG [fludarabine, cytarabine, granulocyte-colony stimulating factor], and the results of a study from The University of Texas MD Anderson Cancer Center of cladribine and low-dose cytarabine. These were safety studies. It is important to point out that myelosuppression appeared to be the biggest safety signal, so outside of clinical trials, these things are not yet recommended, but more study is required. We will hopefully see that BCL2 inhibition can improve outcomes for patients getting intensive chemotherapy as well.
I want to turn to a couple of important investigational agents, and one that I am really excited about, and I think others are too, has data that Eunice presented at ASH [the American Society of Hematology 2020 annual meeting] on the menin inhibitors that may or should have activity in MLL-rearranged AMLs [acute myeloid leukemias], as well as nucleophosmin [NPM1]-mutated AML. Would you tell us about the menin inhibitor that you were studying with your colleagues?
Eunice S. Wang, MD: Sure. On behalf of my colleagues, at Roswell Park Comprehensive Cancer Center, I presented some preliminary data from a first-in-human phase 1b study of KO-539, which is a novel menin inhibitor being developed by Kura Oncology, Inc. As you know, about 5% to 10% of acute leukemias with AML and ALL [acute lymphocytic leukemia] are characterized by rearrangement in the MLL gene, which has now been retermed the KMT2A gene. Those particular leukemias are largely characterized by secondary AMLs, therapy-related AMLs, and pediatric ALLs. There also is a subset, about a third of AMLs as you know, that are characterized by NPM1 mutations, and those can confer a favorable prognosis in some settings. In other settings, it is something that, to date, we have not been able to target successfully, and for that reason, it is not considered one of our actionable mutations per se.
The menin inhibitors work by targeting the menin interaction with the MLL epigenome complex, and that complex is responsible for regulation and transcription of promoters that lead to inhibition of differentiation, and preservation or halting of the myeloid blasts in an undifferentiated state. By targeting menin and targeting that interaction, we may be able to restore differentiation to myeloid blasts, thereby leading to improved outcomes.
This was a first-in-human study. Cohorts were studied anywhere from 50, 100, 200, to 400 mg. Based on preclinical data and some in-class data, there was some initial concern about EKG [electrocardiogram] prolongation, QTc [corrected QT interval] specifically, as well as interactions with the azole antifungal drugs, the CYP3A4 inhibitors, as we saw with venetoclax. To date, in 12 patients examined from 50 mg all the way to 400 mg, there have been no treatment-related adverse events of consequence. No patients have stopped therapy, there has been no QTc prolongation, and there appears to be evidence of early clinical activity in 6 out of 8 patients treated up to 400 mg of drug.
Interestingly, on the phase 1 dose escalation portion, as is common with these studies, we enrolled relapsed and refractory patients as a very-poor subset. These were patients that had between 2 to 7 prior rounds of therapy, and they were agnostic of the mutation. There were 2 CRs [complete responses] that were obtained in those early patients: 1 was MRD [minimal residual disease] positive, 1 was MRD negative, 1 of those patients had an NPM1 mutation, and the other did not. There also was activity stabilization of disease, a morphologic leukemia-free state, and decreased Hydrea [hydroxyurea] requirements in other patients, one of whom had a KMT2A rearrangement. These data, though early, suggest tolerability and early signs of some activity even in a patient who had 7 prior lines of therapy.
We are enthusiastic, cautiously. This is a small number of patients, and it is in the early days. As we all know, a few responses in a few patients does not a blockbuster drug make, but it is an interesting mechanism of action. The fact that we are seeing it across mutations suggests that there may be efficacy outside of these 2 molecular subsets. However, if we were to…on molecular subsets, that would be exciting because there is not yet a precision medicine approach for those particular disease subtypes.
Harry P. Erba, MD, PhD: What I thought was interesting about your presentation was that the 2 deepest responses were in patients who got azole antifungal agents and CYP3A4 inhibitors with the study drug, but you did not see increased toxicity either.
Eunice S. Wang, MD: Right. We do not know whether the azole antifungal agents, as we have seen with gilteritinib, would enhance the levels. We saw activity regardless. We are certainly not excluding the use of azole antifungals, but that is certainly encouraging, and is helpful because a lot of us use azole antifungals in our routine practice.
Harry P. Erba, MD, PhD: Thank you. There were 2 other targeted therapies that were discussed at ASH. I will mention one of them briefly, and that is the retinoic acid receptor-alpha agonist, which is from Syros Pharmaceuticals. The generic name that I know it by is tamibarotene, which I believe is approved in Japan and has been evaluated in acute promyelocytic leukemia. In patients with AML who have retinoic acid receptor present on a biomarker assay, there was a reasonable and significant response rate, I would say, with the combination of tamibarotene with azacitidine in AML.
The update I want to leave this section with is with magrolimab. Magrolimab is the first-in-class anti-CD47 monoclonal antibody. CD47 is known as the “don’t eat me” signal for the innate immune system, and the idea here is that, by blocking CD47, you block the signal that tells the macrophage not to eat a cell. On top of that, the Fc portion of magrolimab is active, so by binding to Fc receptors on the macrophage, you get an “eat me” signal, so you would anticipate that there would be activity of an anti-CD47 antibody. You would also anticipate some on-target toxicity. Since CD47 is expressed on red cells, you get expected anemia, and the investigators figured out how to dose magrolimab to ameliorate this expected anemia.
The presentation here by Dave Sallman, MD, was on 52 patients with AML with a median age of 73. They included two-thirds who had poor-risk cytogenetics and two-thirds who had TP53 mutations. Let me focus on those patients with TP53-mutated AML. There were 21 patients with an objective response rate of 71%, including 10 of the 21, 48%, having a complete remission, and a median duration of response was 9.9 months, which is longer than might be expected otherwise. Of course, you cannot compare between different data sets, but there was a median overall survival of 13 months in those patients, which is pretty remarkable. They also looked at the leukemic stem cell, which also expresses CD47, and they found that the leukemic stem cell was eliminated in 71% of the patients treated with magrolimab/azacitidine. The development of magrolimab, as well as other anti-CD47 antibodies, is ongoing, and we look forward to more data coming from phase 3 studies of magrolimab as well as phase 1 studies of second-generation anti-CD47 antibodies.
Transcript Edited for Clarity
