Harry P. Erba, MD, PhD: What’s so difficult about designing clinical trials with these combinations is our lack of true understanding in the individual patient of the biology of the disease. So, for example, in the ziftomenib study, the patients that we had responses in had other mutations, as you might expect. In fact, CR [complete response rates] that were durable were seen in patients with IDH-mutated disease and FLT3-mutated disease. And so, is there something about the clones that those are involved in, or are they subclones, or it’s just NPM1 driving the disease and it doesn’t matter that the other mutations are there? There’s so much we have to learn as we start targeting the disease with these different kinase inhibitors and small molecule inhibitors.

Eunice S. Wang, MD: I think that that’s an excellent point. We saw that when patients treated with the revumenib compound that had NMP1 and FLT3 [mutations], when they responded to the menin inhibitor, the FLT3 VAF [variant allele frequency] also went down, suggesting that that’s downstream of the NMP1 and, therefore, there might be a broader scope. And then there are trials being designed, because of the concordance of those mutations, FLT3 and NMP1, IDH and NMP1, combining targeted inhibitors. But one question is, well, if you just hit the NMP1 and you get the downstream FLT3, there might be no role to adding a FLT3 inhibitor, because you’re already eradicating the clone. It’s just like in patients that have inversion 16 … One of the most common questions I get is, “If you have somebody who has a FLT3 mutation, and they have an inversion 16, do I need a FLT inhibitor?” A lot of times I’ll say no, because the conventional chemo[therapy] is so good at getting rid of the inversion 16 clone [that] you don’t have to worry about the FLT3 [mutation]. That’s one thing. I also wanted to mention that early on in the first-in-human portion of the ziftomenib trial, when we were dose-escalating, we enrolled patients regardless of their molecular profile. And in the first few dose cohorts, going up from 50, 100, 200, now the phase 1B dose is 600, we actually got responses. The 2 best responses during the dose escalation [phase], 1 had an NMP1 mutation, but the other one had no NMP1, no KMT2A rearrangement. [That patient] had an AXL1 or a RUNX1 mutation. And that patient had one of the longest responses, achieved a CR, and maintained disease control with ziftomenib monotherapy for a prolonged period of time, which brings up the question of whether this drug might be active in other biological subsets of an AML other than these ones that we’re looking at. And we’ve seen that data with the QUIWI trial (NCT04107727) at [the] EHA [(European Hematology Association) meeting], how quizartinib appears to improve survival in patients who have FLT3 wild-type disease. So, there’s a lot we don’t know about how targeted these therapies are and what they’re really doing in humans.

Naval G. Daver, MD: This is not the first rodeo we’ve had with this. [With] quizartinib, 8 years ago, [in] the initial paper from Jorge Cortes, [MD, Georgia Cancer Center at Augusta University, Augusta, GA] in JCO [Journal of Clinical Oncology], there were 40 patients with a 33% CR in the FLT3 wild-type disease. And then, of course, we moved on and I thought we had completely moved on, but then Pau Montesinos, [MD, PhD, University Hospital La Fe, Valencia, Spain] threw in a nice curveball. And, that data is hard to ignore, as you mentioned, [Dr Wang], the QUIWI data: 280 patients randomized 2 to 1, quiz[artinib] vs placebo with intensive chemo[therapy] in FLT3 wide-type [disease]. And they’re showing a big—you can put your thumb and a couple of fingers in between that survival curve. It’s actually a bigger difference in the QuANTUM-First [trial (NCT02668653)]. So, I’m still scratching my head.

Eunice S. Wang, MD: A little bit scary.

Naval G. Daver, MD: I’m trying to figure out what happened there, but I think we’re learning. I still do think, though, the numbers both with Syndax and Kura and the FLT3 are still small for us to know where this will go. I think a combination approach will be very exciting. Also, I we think we have to look at the FLT3 VAF co-mutations, but for example, with FLT3 and NPM1 co-mutated, which is very common. 50% will have a co-mutation. I think a combo could be very exciting approach. So, that’ll be good. [Dr Erba], go ahead.

Harry P. Erba, MD, PhD: It could be novel, but getting back to the initial point of where we saw the responses, and I agree with [Dr Wang], the VAFs of all of the concomitant mutations went down. We may just be adding toxicity in these combinations. And so, the report that [Dr Wang] brought up of other pathways being activated may be the smarter way for us to go if we had something that targets that pathway. But at this point these are the tools we have and, of course, I agree with doing these combination studies, but obviously we need more understanding of the biology in these patients.

Naval G. Daver, MD: I know from the Syndax, we had about 13 patients, and I think 3 or 4 of them with FLT3 [mutations] responded. It was not in the 60% and 70% range. It was similar to what you saw with others. Is that similar with Kura? Or is it higher for the relapse?

Harry P. Erba, MD, PhD: That’s the interesting thing. We don’t quite have that data yet, unless [Dr Wang] has seen it. We have the individual responders, but we don’t have the data from all patients yet.

Naval G. Daver, MD: Right. We’ll have to wait. And with gilt[eritinib] and combos of gilt[eritinib], we’re getting good response. I think the question is, can we do shorter durations of combination [therapies] and maybe stop treatment? There [are] going to be a lot of things we will learn over time.

Harry P. Erba, MD, PhD: It’s felt that these mutations, NPM1 and the KMT2A, signal through menin by causing upregulation of HOXA9 and Meis1 and other proliferative signals. And so instead of looking at the mutational pattern, looking at gene expression profiling may be a way of identifying those rare patients who may benefit. For example, I think [Dr Wang] mentioned the zift[omenib] patient [who] had a very long response. So, something that we have to consider in designing further studies. But again, I think that’s the kind of thing that’s going to happen once the drug is in the hands of talented investigators.

Naval G. Daver, MD: And in fact to that extent, I think Ghayas Issa, [MD, MD Anderson Cancer Center, Houston, Texas] and [Eytan M.] Stein, [MD, Memorial Sloan Kettering Cancer Center, New York] are already starting this kind of basket of MIS1, HOXA9 overexpressing populations that includes the NUP98, the NUP214, the UBTF-FLT3. Every few weeks we see a new one, so it may end up becoming more than 20% to 25% once we know how to identify them, whether it’s by sequencing or whether it’s by using molecular surrogate. And I think that would be quite interesting. I think we’re kind of getting close to time. I will just mention there are other menin inhibitors that are [under] ongoing [investigation]. I’m sure you’ll see some of these data at [the] ASH [(American Society of Hematology] Annual Meeting]. There is one with J&J [Johnson & Johnson], which initially had issues with differentiation leukocytosis, but now I understand has been optimized, and, hopefully, we’re going to see some good clinical data there. And then there’s one with Sumitomo, DSP-5336, which also is in early stages but is looking encouraging. So look out for those data. And again, I think just like FLT3 [inhibitors], just like TKIs [tyrosine kinase inhibitors] and CML [chronic myeloid leukemia], just like IDH [inhibitors], there may be room for multiple of these drugs based on their safety. Some have QT effects, some have differentiation, some may have other [adverse effects]. And also their ability potentially to be combined safely may be quite different. And then as many of you mentioned, the resistant pathway that we initially thought was going to be a converging pathway for all menin inhibitors, based on some of the data shown at ASH and published, that may not be true. This may be very similar to TKIs and CMLs, where you have to look at the particular mutation, do a kinase sequencing, and then use the other menin inhibitor, which may or may not have that same resistant pathway. I think that’s going to be quite an interesting area.

Transcript is edited for clarity and readability.