Erba on Identifying the Optimal Dose of Ziftomenib in Relapsed/Refractory AML


In an interview, Harry Erba, MD, PhD, compared ziftomenib with other Menin inhibitors and discussed data from the phase 1/2 trial evaluating the menin-KMT2A inhibitor in patients with AML.

Harry Erba, MD, PhD

Harry Erba, MD, PhD

At 600 mg, ziftomenib (KO-539) monotherapy led to antileukemic activity and a manageable toxicity profile among heavily pretreated patients with relapsed/refractory acute myeloid leukemia (AML), according to findings from a phase 1/2 trial (NCT04067336).

Ziftomenib is an oral investigational drug designed as a highly potent and selective inhibitor of the Menin/KMT2A interaction. According to Erba, this leads to the downregulation of HOXA9 and MEIS1, and increased differentiation.

In the phase 1/2 study, ziftomenib 600 mg was administered to patients with NPM1-mutated disease who were enrolled in phase 1a and 1b of the trial (n = 20). Treatment with the menin-KMT2A inhibitor produced a complete remission (CR) rate of 30.0%, the CR/CR with partial hematologic recovery (CRh) rate was 30.0%, and the composite CR (CRc) rate was 35.0%. Ziftomenib at 600 mg also generated a minimal residual disease (MRD) negativity rate of 42.9%, and an overall response rate (ORR) of 40.0%.

Patients with KMT2A-rearranged disease enrolled in the phase 1a and 1b portions of the trial (n = 18) who were given the agent at 600 mg had a CR/CRh rate of 5.6%, and the CRc rate was 11.1%. Moreover, the MRD-negativity rate with the ziftomenib was 100.0%, and the ORR was 16.7% in this population.

Based on these findings, the FDA has selected 600 mg as the recommended phase 2 dose for patients with NPM1-mutated AML.

“What we're excited about is that we believe there is single agent activity and NPM1-mutated disease that may be useful to some of our older patients who get HMA, then relapse and don't have any other options. A single agent Menin inhibitor with a 30% to 40% response rate with a tolerable toxicity profile would be an advantage,” Harry Erba, MD, PhD, told Targeted OncologyTM, in an interview.

In the interview, Erba, professor of medicine at Duke Cancer Institute, compared ziftomenib with other Menin inhibitors and discussed data from the phase 1/2 trial evaluating the menin-KMT2A inhibitor in patients with AML.

Targeted Oncology: Can you discuss the mechanism of action of ziftomenib?

Erba: Ziftomenib is an oral investigational agent that inhibits the interaction between Menin and the protein KMT2A, used to be known as MLL. This is an important interaction because that interaction leads to upregulation of HOXA9 and MEIS1 and in so doing, it causes proliferative signal blocking differentiation in the cells. Ziftomenib inhibits that interaction, and by inhibiting that interaction, leads to downregulation, decreases the expression of HOXA9 and MEIS1, leading to differentiation of the cells. It can be seen as a differentiating agent.

What can you tell us about the use of Menin inhibition in AML?

Based on the mechanism of action, we would expect a Menin inhibitor to be effective in any of the AML where the pathogenesis of leukemia is related to signaling through that Menin KMT2A fusion or interaction. There are at least 2, and probably more, subtypes of AML, where this is a critically important interaction that underlies the leukemogenic properties of the mutation. One is the nucleophosmin or NPM1 mutations, and the second are KMT2A rearrangements or fusion proteins. Those 2 will lead to activation of the Menin KMT2A complex and upregulation of HOXA9 and MEIS1. An inhibitor there would be important in breaking that pathogenic signal.

There are others that may be important besides NPM1 and KMT2A. Other fusions are known to work through the Menin and proteins such as NUP98, so that could be a rare target in AML. Then, it's speculated that in AMLs with upregulation or overexpression of HOXA9 and MEIS1, presumably through this Menin and KMT2A interaction, a Menin inhibitor may also be important there, but that's yet to be determined.

How does ziftomenib compare with other Menin inhibitors?

There are a few Menin inhibitors already in clinical development. None are FDA approved or commercially available at this point. The first is revumenib [previously known as SNDX-5613], there is ziftomenib, and then there are a number of other investigational agents. What's different about ziftomenib from revumenib is [what] I could see in terms of toxicity profile. We don't see the same QTC prolongation signal that was seen with revumenib. However, on the other hand, in our phase 1 study, we saw activity of ziftomenib in the KMT2A-rearranged patients. However, there was an early increase in the blast count and an atypical differentiation syndrome. We did not see any objective responses in the KMT2A-rearranged patients, unlike what was seen in the revumenib studies.

Now we believe and we see incredible activity in this on-target effect. We believe that we'll still be able to develop ziftomenib in KMT2A-rearranged patients, but we still have to look in a little bit more detail at the dose and combination strategies there. In terms of the NPM1-mutated, we saw a 30% complete remission rate in 20 patients treated the recommended phase 2 dose, so quite a good dataset and s a single agent, a rate of inhibition or rate of complete remission that rivals what we see with IDH inhibitors that have already been approved based on single agent activity.

Can you discuss the methods and design of the phase 1 study evaluating ziftomenib?

There was a phase 1a portion and this was for all comers. About half the patients had either a KMT2A or an NPM1 mutation, but this was looking for the dose to bring forward. We treated patients in the phase 1a portion with doses from 50 to 1000 mg a day, and we did see activity. We saw 2 patients with an NPM1 mutation achieve either a complete remission or morphologic leukemia-free state. The patient who achieved a complete remission remains in a complete remission for over 2 years and single agent ziftomenib. We also saw a response in a patient who did not have NPM1 or KMT2A rearrangement. The patient had a SETD2 RUNX1 mutational profile, and there are some who believe that Menin inhibitors may be active in RUNX1-mutated AML, so more to be seen in that case.

Based on that, what we decided what to do for the phase 1b portion was to take to heart and listen to the guidance that the FDA has recently given out on the development of these targeted inhibitors. The FDA initiative is called Project Optimus, and what it acknowledges is something that we've all known. With a targeted therapy and an inhibitor, it may reach its maximal activity based on PK or PD [pharmacokinetic or pharmacodynamic] studies before it actually gets to a maximum tolerated dose. There are plenty of examples of the development of drugs in AML that missed that signal and went too far, maybe in the development of the drug in phase 1, so we didn't want to do that. What we did was we went back to phase 1a, and we developed the 2 lowest dose levels at which we saw responses, and those were 200 mg, and 600 mg. That was the phase 1b portion of the study, and that was in about 57 patients. There we enriched the population of NPM1-mutated and KMT2A-rearranged patients to look for on-target activity.

What are the main findings from this part of the trial?

We looked at 200 vs 600, so how are we going to determine from this dose optimization and what should be the recommended phase 2 dose? Well, clinical activity. We had 20 patients with NPM1-mutated disease that received 600 milligrams daily. In those 20 patients, we saw 6 complete remissions, and then another CRi [complete remission with incomplete hematologic recovery], and then another morphologic leukemia-free state, so an overall response rate of 40%. On top of that, in the NPM1-mutated patients, we saw no grade 3 or higher toxicities occurring at more than 10% of the patients, but we did see some grade 3 or higher toxicity. Specifically, of the 20 patients treated at 600 mg, we saw 4 patients have a differentiation syndrome, but only 1 of the 4, 25%, had a grade 3 event. None of the patients died of differentiation syndrome in the NPM1-mutated group, and 75% of the patients who had differentiation achieved a response, suggesting that it is the mechanism of action.

I presented the data for 1 patient showing how her bone marrow did develop into a quick morphologic leukemia-free state and then a couple of cycles later, complete remission. At that 600 mg dose compared to 200, the PK profile favored 600, but importantly, also the PD. We saw downregulation of the target genes HOXA9 and MEIS1 at the 600 mg dose more so than at the 200. Importantly, at higher doses than 600, we didn't see any further downregulation. Based on the response rate, based on the toxicity profile, based on the PK and the PD, we feel secure in our data that 600 mg once daily is the right dose to take ahead in patients with acute myeloid leukemia and the FDA has agreed to that and so the study has been amended now to get us into the expansion phase at 600 mg once daily.

However, in the KMT2A-rearranged patients, we saw something different. In those patients, we saw a rapid rise in the white count, often without signs of differentiation in the blood, often with rapidly rising white counts, and then unusual forms of differentiation syndrome. For example, in some of my patients, we saw patients develop Gingival Hyperplasia in the first cycle that went away, and that was from the disease responding in that cycle. It flared up and then went away. A patient developed pleuritis that wasn't explained by anything else and went away with steroids. Another patient had a serositis of the knee, and then the right wrist, and then in muscle that came and went with steroids, finding no other infectious etiology for it.

The reason we think this is happening is that we know that KMT2A-rearranged AML often has a monocytic morphology and has a high incidence of extramedullary disease. The other thing is that ziftomenib accumulates in tissues, and we showed that in the phase 1 portion of the preclinical studies. What we believe is happening is that with a single agent, we are seeing abrupt rises in the blast count and this differentiation. It caused some investigators to take their patients off study, before even finishing 1 cycle of KMT2A. We clearly see activity. We've clearly seen morphologic leukemia-free state and marrow clearance a blast but have not been able to keep patients on long enough to see complete clearance of extramedullary disease and reconstitution blood counts. So, [there were] no complete remissions. But we feel secure that we will be able to find the right dose, and especially the right combination of ziftomenib with chemotherapy to help ameliorate that signal of differentiation syndrome and KMT2A-rearranged patients. But in NPM1, those are the patients in the expansion that we will be enrolling at the 600 mg dose.

How do you foresee this agent being used in clinical practice?

Some would look at this and say, KMT2A-rearranged AML is pretty atypical and unusual, we don't see much of it. However, when we see it, we know that they can get into a remission. But remissions are brief, and relapses are the hallmark of this disease, unless a patient is able to go through allogeneic stem cell transplant. The other place I see the KMT2A-rearranged AMLs are in patients who have received prior topoisomerase II inhibitors, such as the anthracyclines, which developed within 1 or 2 years, a therapy related AML with the KMT2A rearrangement. Every year, I see 2-3 women who have gone through adjuvant therapy for breast cancer and then ended up developing this disease. We see these patients, but it is a smaller population. In pediatrics, however, it's a larger population, especially in acute lymphoblastic leukemia.

On the other hand, people have also said, in NPM1-mutated, isn't that a favorable risk AML? Favorable risk is only in relationship to the rest of the populations of AML, so it is favorable, but it is not. We're not hitting a homerun yet. Although NPM1-mutated disease is considered favorable risk and ELN risk stratification and NCCN risk stratification, we know these patients can still relapse, and it appears to be related to the co mutations. For example, half of the patients with an NPM1 mutation will have a FLT3-ITD mutation. In those patients with a FLT3 ITD, NPM1 mutation and a DNA methyltransferase 3 alpha mutation, the prognosis is poor with a 20% survival as shown by a retrospective analysis of a German database. We know these patients are relapsing. A testimony to that is we were able to find and put on a study of 20 patients with NPM1-mutated disease in the relapsed/refractory setting on the 600 mg recommended phase 2 dose, so those patients are there.

What are the next steps for this research?

What we're excited about is that we believe there is single agent activity and NPM1-mutated disease that may be useful to some of our older patients who get HMA, then relapse and don't have any other options. A single agent Menin inhibitor with a 30% to 40% response rate with a tolerable toxicity profile would be an advantage. But clearly where we expect to go is where we expect to go with all these targeted therapies, in combinations with chemotherapy with other targeted therapies and maybe even immunotherapy at some point.

The sponsor of this trial has already written a phase 2 study that has ziftomenib in combination with intensive chemotherapy, HMA venetoclax [Venclexta], and in FLT3-mutated patients, gilteritinib [Xospata] in the relapsed/refractory setting, and then also arms of the study of HMA venetoclax or intensive chemotherapy with ziftomenib in previously untreated patients in the appropriate patient population for each either fit or unfit for intensive chemotherapy.

Erba HP, Fathi AT, Issa GC, et al. Update on a phase 1/2 first-in-human study of the menin-KMT2A (MLL) inhibitor ziftomenib (KO-539) in patients with relapsed or refractory acute myeloid leukemia. Blood. 2022;140(suppl 1):153-156. doi:10.1182/blood-2022-167412
Related Videos
Related Content