 Chemotherapy isn’t dead. It still has a role. CPX-351, the liposomal daunorubicin cytarabine, has been approved by the FDA for patients with secondary AML [acute myeloid leukemia], those who had previous MDS [myelodysplastic syndrome], CMML [chronic myelomonocytic leukemia], patients with de novo AML [acute myeloid leukemia] with complex karyotypes or MDS-like karyotypes, and treatment-related AML. And this was based on a large phase 3 study that actually validated the results of a prior randomized phase 2 study, showing improvement in overall survival but also CR [complete response] rates and not at the cost of higher induction mortality. It was actually lower.

And I think what was most impressive about that report to many of us is that the patients who went on to allogeneic transplant for this very poor-risk AML following CPX-351 had a much better survival than those who went on to allotransplant after 7 and 3. We’re not really sure of the reason for that. It could be that we induced a deeper level of minimal residual disease with CPX-351. It could be that they weren’t as beaten up from CPX-351, but we really don’t know the answer. The question is: What populations are really benefiting?

And I think at the ASH [American Society of Hematology] meeting, R. Coleman Lindsley, MD, PhD, presented data where two-thirds of the patients, we had mutational data, and what he showed was no response benefit, no survival benefit in the 

-mutated patients, which was a third of the population. But in patients with the secondary AML mutational profile, those are both those patients with a spliceosome mutation, chromatin modifiers, and that 1 cohesin mutation, 

, those patients did have an improvement in survival in that subset with CPX-351 versus 7 and 3. But, of course, we have a lot of questions about how to use CPX-351. I want to throw those open to the group. And the most obvious one to start with is: Does anyone use CPX-351, or is everyone giving AZA [azacitidine] or HMA [hypomethylating agent]/venetoclax [VEN] combinations now? Dan is smiling. You’re the first to smile, so you’re on.

 I’m giving a lot more venetoclax combinations to patients that would have probably, in most places or in a recent era, felt to be pretty decent intensive induction chemotherapy candidates. And there’s a variety of reasons for that, and there’s a lot of extrapolation that goes into that. But that has eroded my use of all chemotherapy, including CPX-351. What you brought up about the presentation at ASH [American Society of Hematology] , the Coleman Lindsley presentation, is somewhat concerning because if you are now removing the 

-positive patients from the secondary and treatment-related AML population, you’re starting to leave yourself with very little. There’s not much left if you have to subtract those patients out. And so I think it’s a commentary about disease resistance subtypes against chemotherapy approaches that maybe aren’t so dramatic or set in stone for nonintensive chemotherapy approaches.

 Harry, I would take a stab and say we use about 10:1 HMA/VEN to CPX, if I had to guess.

 So, Mark, as long as you chimed in, let me ask you. So let’s say you have somebody in the labeled indication for CPX-351, a secondary AML, someone who had received an HMA before, even, or a therapy related, and they have a 

 mutation; would that be a patient you would consider CPX-351 with a FLT3 inhibitor?

 When I’m considering CPX, I look at fitness. It’s not a drug for the unfit, even though it sort of has that reputation, uh-uh. You’ve got to be fit enough to tolerate it. And so then it’s on that paradigm, and the AZA, as you know, if they’ve had AZA before, their response to anything is less good. As far as the FLT3 inhibitor goes, no. If they’re fit enough to get CPX, I’m more likely to give 7 and 3 in a FLT3 inhibitor than CPX. We’re doing that trial investigating that.

 I would just bring up the point that we clearly are using a lot more HMA/VEN. But I actually believe in pharmacologic optimization of chemotherapy, and I think that I’m interested in knowing is this better, 7 and 3, and what about younger patients or patients. I mean, most patients with AML are cured with chemotherapy. We haven’t cured anybody yet, really, with the venetoclax-based or IDH-based therapies. We’re doing well, and we want to avoid, we wanted to avoid intensive chemotherapy in those patients to begin with, so we’re excited to be able to do that.

But the question is: Can 7 and 3 be replaced with CPX? And I still wonder about that. Because if it’s just plain better, if it’s better tolerated and sort of better pharmacologically, that’s still an open question that I think we have to study. I definitely feel that we have the over 70 or over 75 in bad shape doing well with HMA/VEN, begs the question that if you’re 62 and in great shape, why can’t that regimen be for you, too? And I think that has shifted the use of HMA/VEN to what is clearly the off-label setting of better patients who could theoretically target, be treated with CPX as an on-label.

 To that point, we would, for the 62-year-old fit enough, we would not offer HMA/VEN; we would offer CPX if they have secondary AML.

 That’s the question that how, exactly, what is the calculus in that, and I think it’s hard sometimes. That’s a little bit in the eye of the beholder.

 And those clinical trials are being designed to look at those fit patients. Because I agree with Mark: CPX-351 still is chemotherapy with a very same profile of toxicities, maybe a little bit less mucosal toxicity but for more prolonged myelosuppression. But let’s come back to the fact Jeffery Lancet, MD, our friend from Moffitt Cancer Center, presented the 5-year update of that phase 3 study. And there’s a true plateau on the survival curve with CPX-351, much above 7 and 3, or double 7 and 3 is what I should say. We have a lot of room for improvement, and I would imagine I think most of those patients out in the tail of the curve were patients who were able to undergo allogeneic transplant. So I use CPX-351 for that fit, younger patient where I know that the goal of therapy is cure, and I’m trying to get them to transplant for a secondary AML or therapy-related AML.

Videos

CPX-351 in Newly Diagnosed Secondary AML

 I’ll come back to Courtney DiNardo, MD’s presentation at the randomized phase 2 of an enasidenib and AZA [azacitidine]. The azacitidine arm probably looked good because 20% of those patients got enasidenib. But that’s the only data she had. I wonder how many of those patients actually got venetoclax combinations and did well.

So your point is well taken, Amir, that, you know, I’m not saying that the survival of patients is that AZA alone was going to be 22 months. I’m saying that we now have choices for these patients, and we can add these therapies in.

 But I think it’s worth it to note that we’re probably never going to really know the answer to that study for the exact reason that choices evolved as those studies were trying to accrue. So, I mean, if it becomes really difficult, you have to go outside of the United States to complete the accrual because there was complete contamination with venetoclax. I think it’s just been; I think that those companies have had a very tough time accruing and have definitely broadened the net with respect to participating trial centers.

 Amir, I want to come back to you. You published in 

 the enasidenib experience with differentiation syndrome [DS]. Do you want to talk about IDH DS?

 Thank you, Harry. So we published about 14%, 15% of patients who received enasidenib monotherapy on the original phase 1 dose expansion, dose escalation and dose expansion study.

And so differentiation syndrome in that setting mainly manifested as a vague constellation of symptoms–unexplained fever, respiratory symptoms, pleural effusions, pericardial effusions, rash, mild generally azotemia, boney pains, adenopathy. And we tried to go through every single case that had that constellation. And, as you can imagine in a patient who has AML [acute myeloid leukemia], who is marrow suppressed, who has potential cardiac pulmonary symptoms, it’s difficult to tease out what’s what.

Courtney DiNardo and Stéphane de Botton, MD, we came across approximately 15% or so of patients who we could easily or fairly well rule out other secondary causes of those symptoms and signs and could say that the likelihood of differentiation syndrome was possibly or likely related to the actual process, meaning that the differentiation syndrome was caused by the drug. There wasn’t a secondary cause for those symptoms, that it was related to differentiation syndrome.

Differentiation syndrome can be severe. It can be potentially life threatening, and so our recommendation has generally been that if you suspect it and worry about it, and secondary causes cannot be immediately and quickly ruled out, that you should start treatment with dexamethasone and we recommend 10 mg twice a day. And then once, and if, patients get better, and they should if it is differentiation syndrome, then there should be a tapering down dose over time.

On occasion, differentiation syndrome is accompanied, but not necessarily so, but on occasion it is accompanied by leukocytosis. So hydroxyurea may be necessary if there is concurrent leukocytosis. Similarly, other phenomena can be a co-occurrence such as DIC [disseminated intravascular coagulation], which can be potentially life threatening, as well as tumor lysis syndrome.

So, again, it is hard to necessarily tease out differentiation syndrome in the patients with AML. But when secondary causes are ruled out or cannot be ruled out in a timely manner, I think initiation of steroids is important because these conditions can escalate. And if you look at study after study after study with IDH inhibitors, either as monotherapy or a combination with induction for HMA [hypomethylating agents], you’ll see this weight of approximately 12% to 20% of patients getting differentiation syndrome. So it is a real entity, and it is to give up that closely in patients who get treatment.

One other important point I would just say is that the timing of developing differentiation syndrome can be somewhat delayed as opposed to a differentiation syndrome that you get with ATRA [all-trans retinoic acid]. The differentiation syndrome with IDH inhibitors, typically we found that the median time of loss was around 12 weeks. So anywhere between 10 days to 6 months, you can potentially get it. Especially if you stop treatment and resume it later, you can get recurrent episodes of differentiation syndrome.

 I agree, Amir. A couple other important distinctions between IDH and APL [acute promyelocytic leukemia] differentiation syndrome is 100% of patients with APL who developed differentiation syndrome, if they survive, will achieve a remission. That is not true with the IDH inhibitors. And partly because I think some of these patients really, truly had progression and pneumonia and heart failure and things like that. But also it may be that you’re differentiating 1 clone, but then a 

-positive clone grows out. And so you may see some differentiation.

 I completely agree with Harry on this one. APL is about the cleanest disease that we treat, right? Generally, it’s only 1 alteration that causes the disease. On occasion, you’ll get a 

 mutations, you get the ugly AMLs, but you know, the multiple clones, the other mutations, you might have a potential efficacy. And training the 

-susceptible clone with an IDH inhibitor, that might be why not everybody with treatment with IDH inhibitors gets to the differentiation syndrome and certainly not, well partly the reason why patients don’t respond, some patients don’t respond to IDH inhibitors.

But also keep in mind that about 100% of patients treated with ATRA arsenic trioxide these days also get remission anyway, so… we’re kind of comparing something that is good with something that is outstanding. So it’s a little bit challenging and not perhaps the comparison for IDH Inhibitors.

 Harry, I wanted to jump in with 1 comment that, you know, AML used to be an inpatient disease, right? And now there are lots of aspects of AML therapy that are being handled as an outpatient. And one of the things that we’ve actually worked on with the company is surrounding differentiation syndrome is I think a little bit of patients’ education and advocacy so that they understand, also, this toxicity. Now you have an APL patient coming into the hospital, you’re not necessarily giving them a lecture on differentiation therapy if you’re watching them in the hospital. If they get it, you’ll tell them about it—if they don’t, they don’t.

But a lot of these patients, we were thinking as we were trying to prepare some of the educational materials for patients, they might end up in a local emergency department where somebody has never heard of the medication that they’re on; they don’t know what to look for. And I think that there actually has been an explosion of articles for internal medicine doctors, for emergency department physicians, on various oncologic complications of therapy, both for solid tumor therapies and also increasingly for heme malignancies, to just alert doctors that if you have somebody coming in with leukocytosis and infiltrates, it may not be pneumonia. You might have differentiation. So I just think it was a whole other axis of this particular toxicity to kind of warn the public, if you will, warn the patients, warn so that people were not getting really down the wrong path as they presented to the emergency room.

 Yeah, early intervention with dexamethasone, 10 mg twice daily, is very important, admitting them for monitoring. I will say that I have a dot phrase, a smart phrase in my Epic notes that go into the patient instructions as to when to call, how to monitor their weight, if they’re getting more short of breath. And I talk to my nurses as well. If someone on an IDH inhibitor calls in with shortness of breath, yeah, most leukemia patients, that means you can set them up for a transfusion in a couple days. The IDH inhibitor patient, we need to see. And so there is a lot of education there.

I want to move on to our last topic. So, before I get there, I want to mention a couple of other important toxicities that we do manage with the IDH inhibitors—hyperbilirubinemia is seen with enasidenib. Quite frankly, I don’t do dose adjustments because of hyperbilirubinemia from inhibition of UGT1A1.

 Harry, it’s a measure of patient adherence.

 But it is a mistake to stop the therapy, and I think you want to get that in your smart phrase, which, by the way, please send me because that will save me a lot of work. But I think that point also, you don’t want to stop the effective therapy for an irrelevant bilirubin value.

 QT prolongation has been associated with ivosidenib, and then not many people know this, but in our phase 1 experience, you all may remember Guillain-Barre in 2 out of about 250?

Differentiation Syndrome and IDH Inhibitors in AML

Let me come back to Mark’s point about picking ivosidenib for the previously untreated patient, maybe an older patient who couldn’t tolerate a myelosuppression. How about the patients who’ve had MDS [myelodysplastic syndrome] and been on a hypomethylating agent [HMA]? Is that another niche for the IDH inhibitors?

 So, I can’t comment from trial data. There are trial data looking at, for example, IDH inhibitors, as you know, in MDS. You all have used it; I know you have, I have. And, actually, it works. It’s highly effective. It’s the same disease. Who cares if it’s 9% blasts; it’s going to be the same disease. In fact, my impression is it works a little better than if it’s AML [acute myeloid leukemia]. If you’ve got an 

-mutant high-grade MDS, I love an IDH inhibitor in that setting, to be honest.

 But that’s basically what I’m asking, though. This is an MDS patient progressed to AML, and now you find an 

 mutation. Do you add VEN [venetoclax] to an HMA? Do you add an IDH inhibitor to an HMA?

 No, I would probably go with the IDH inhibitor at that point. That’s been my practice.

 I mean, there are some data that Amer presented at ASH [American Society of Hematology], Amer Zeidan, MBBS, Yale School of Medicine, showing the utility of addition of venetoclax to a hypomethylator failure MDS situation. And they were pretty promising data. It’s still early, but I think it’s a bit of a dealer’s choice at that point.

 Yeah. AZA failed, so then add VEN. IDH failed; add AZA/VEN, it worked great. AZA/VEN failed; add IDH. I mean, we have choices.

 But I want to be very clear about Amer Zeidan’s presentation about adding venetoclax to HMAs, and HMA failure patients who still have MDS. You know what was most impressive to me about those responses that he saw was that these were mostly more marrow CRs [complete responses] with an AML we would call a morphologic leukemia-free-state. In his ECOG presentation, he tried very hard to say that there is some hematologic improvement, which he showed. But I would say that the real benefit there is if you have, these are high-risk patients, right? MDS progressing to AML. Their only long-term outcome is going to be good with, if they undergo allogeneic transplant. So it may be a way of cytoreducing the marrow and making your transplanter happy that the blasts are gone even if the disease isn’t.

 Yeah, I just did it with a guy who failed 3 cycles of AZA who was sent to me; we added VEN. Great; his counts normalized. He looked like he was in CR. The marrow still shows MDS. He’s going to transplant, but this is a high-risk transplant.

 That happens with some frequency, actually. I do think that we should make the point, because you were talking about MDS transformations, there are perhaps a surprising proportion of patients with antecedent MPN [myeloproliferative neoplasms] who also seem to pop an 

 mutation. And those patients have historically been particularly difficult to manage and get into remission. So that’s another scenario where if you have an antecedent MPN, it may be worth it to wait an extra minute and get back your mutation studies.

 Yeah, the responses to the IDH mutant inhibitors in those patients are extraordinary.

 They’re really good. So, yes, that’s a group where other therapies are particularly not good, so there’s no need to rush to launch something ineffective for those patients.

IDH Inhibitors and MDS

 Dan, you published on newly diagnosed patients getting enasidenib. Gail did publish the single-agent IVO [ivosidenib]. Tell us about enasidenib there.

 Yeah. You know, very similar experience, very similar numbers, you know, patients included in that study. Similar sort of a design, in which the patients really came out of the original phase 1 in an arm of patients who had no prior therapy. And very similar outcomes, really, with respect to response rates and survival. So no big surprise there.

 And the responses with both IDH inhibitors seemed to be higher with combination with an HMA [hypomethylating agent] than VEN [venetoclax] alone. So there seems to be something there with that combination in the relapsed/refractory setting and in the newly diagnosed patient population. But I guess 1 of the most interesting things I saw at EHA [European Hematology Association] and ASCO [American Society of Clinical Oncology] was Dr Courtney DiNardo’s presentation on the randomized phase 2 of AZA [azacitidine] alone versus AZA and enasidenib. Higher response rates, but no difference in survival. Getting at what Gail was talking about, we’re starting to see some studies there. These patients are having longer survival than what we expect and have seen before.

 I’ll tell you, though, we haven’t done a good job with the super responders also. And, I mean, Dan and I have been looking at this in some of the data sets. So I actually have on the ENA [enasidenib] single agent, I have the person who would be absolutely the longest. I mean, this is an older patient who was 71. I forget exactly how old he was when he initially went on the trial, was told, “Forget it,” you know, nothing. Five years on single-agent ENA, and now back in remission in a hot minute on AZA/VEN when he relapsed. But the question is: How do we pick; how could we pick the patients who were destined to be those super responders? And this is single-agent initial therapy. So even though you’re absolutely right that the response rates that were just presented for the ENA/AZA were higher, the question is: Some of these patients actually can get single-agent IVO or single-agent ENA and live for a long time on it, but how do we pick them out?

The statement that you were getting to that we’re just getting very good at keeping people alive for longer on azacitidine: I think that that should still be used. Keeping people alive for longer is good. But I absolutely believe that if we could ever study the quality of life, being in remission is better; it has to be better.

 I just wanted to bring up a couple of points that I think are pretty important. One is that if you’re alluding to the abstract presentation from ASCO, I think you’re talking about the one that had 100-something patients who were randomized 2:1. I think we have to keep 68 in 1 arm, 30 patients in the other. The overall survival was, I think, 20 months or 22 months, something like that.

 So, you know, this is not a randomized phase 3 study; we cannot assess, in my view, survival advantage in such a study. And so I worry a little bit about drawing conclusions from that. I do think that HMA plus IDH inhibitor is better tolerated than HMA/VEN, so I agree with Mark. And there are certain patients in whom HMA/IDH inhibitor is the better approach.

But all else being equal, I like to keep my IDH inhibitors in my back pocket because they provide a response rate that approximates relatively the response rate you get in the up-front setting. Meaning the relapsed/refractory patients had a response rate that approximates as monotherapy what you would get in the up-front setting. You know, somewhere in that overall response of 40% to 45%, complete response of around 20% to 25%. That’s what we have generally seen roughly in patients. And because HMA/VEN has such a high response rate, I generally in the younger patient, or the patient who I think can tolerate HMA/VEN, generally go with that first and keep the IDH inhibitor in my back pocket. On occasion, however, if I’m worried about whether a patient may tolerate HMA/IDH inhibitor better, I generally go with that. I’d like to have options so that I can prolong a patient’s survival with sequential therapy. There’s no data, obviously, to guide that, but that’s just been my general approach with these patients.

 When considering how to sequence these agents, HMA/venetoclax versus an IDH inhibitor, let’s remember that the labeled indication for venetoclax is in the treatment-naive older, unfit AML [acute myeloid leukemia] patient. Whereas the labeled indications for the IDH inhibitors were initially, at least, in relapsed/refractory; now, ivosidenib is approved also in the treatment-naїve patient, but if you sequence it with venetoclax first and then the IDH inhibitors, you don’t have to worry about the opposite situation, having to get approval for venetoclax. Having said that, I think most of us on this panel haven’t had much difficulty getting these medications for our patients with AML if they need it. But it is something to consider in terms of the sequencing.

IDH1/2 Inhibitors for AML

 I’m going to have Gail take the lead on this because we’re going to switch gears and talk about the 20% of AML [acute myeloid leukemia] patients that have mutations of 

 Well, the land of IDH, you know, brought us back to all of our nightmares of organic chemistry and Krebs cycle and everybody kind of got through that year where we actually learn this again. And everybody took their medications and learned about IDH inhibitors. Revolution No 1 was the identification of course of the mutations in AML. Then, all of a sudden, wait a minute: AML, relapsed AML, isn’t necessarily completely the land of the lost. We have been suffering so enormously with poor therapies for patients with relapsed AML.

We have been pummeling them for decades with lots of different chemotherapy combinations without success. And then there are data emerging for ivosidenib and for enasidenib: ivosidenib being a potent IDH1 inhibitor, an orally available drug; ivosidenib an orally local inhibitor or mutant 

, showing actually quite high response rates as well as durable remissions, a single-agent, pretty well-tolerated drug in relapsed AML; not curing the patients.

We are not talking years in remission, but we are talking durable remissions in the 6-to-12-month zone, with a single agent. Toxicity profile has been a little bit glitchy for these drugs. You have to watch for differentiation syndrome, which we’ll have to talk about a little bit. You have to look at QT. But basically these are orders of magnitude different from standard chemotherapies. They have better response rates than chemo in these groups, and we thought, “OK, this is huge; let’s move them to the up-front setting.”

So there we also had some interesting data. The ivosidenib trial allowed some patients with newly diagnosed disease to be treated. And actually, there’s the CR-CRH [complete remission and CR with partial hematologic recovery] of about 40% in those patients. Some of those, if you look at the swim plots, were actually durable remissions. These are patients who might not actually have been offered anything. They were the patients that you’re talking about, Harry, who are older and may have fallen into the nihilism trap of, “Oh, you can’t get anything.” And yet here they are in a durable remission. So the data are very similar for enasidenib. So these are single agents that can be used both up front and in relapsed disease to be remitters; they induce remission.

Of course, then the question is: What about combining them with standard chemotherapy? What about combining them with azacitidine? But actually the world was moving at the same time, and what we started seeing as this tremendous success was evolving was that actually venetoclax was particularly affective in the IDH subgroups, in IDH1 and IDH2.

So, this is confusing, and I’m sure you’re going to want to initiate some discussion about this, because now you have choices: “Well, how do I pick a combination of venetoclax and azacitidine, or which ivosidenib and azacitidine, or which 1 if we have a little bit of too many choices?” And I would say the summary that we’ve seen so far out of EHA and out of ASCO [American Society of Clinical Oncology]. So we know that IVO plus azacitidine works. That is an effective regimen for up-front disease.

We think that ENA plus azacitidine might be a little bit less resoundingly good. It didn’t look like there’s a survival benefit over azacitidine alone, but there was still high response rates. So the question on everybody’s mind is: Why would you choose one of those instead of AZA and venetoclax? And I think we can deliberate that a little bit. I think there are individual patient characteristics that may lead to 1 combination versus the other.

But I think the most interesting part is that we were telling everybody, all of us on this call, on this kind of conference—although we didn’t use to use Zoom, but we were all telling people: “Wait, calm down; wait for the mutational profile. AML [acute myeloid leukemia] doesn’t have to be treated for everybody in 5 seconds; let’s apply our inhibitors.” Certainly in the relapsed setting, this is absolutely true, and the last thing we want to be doing is pummeling on chemotherapy.

But in the newly diagnosed patients, I think the question is: Do you wait? Or do you say, “Well, venetoclax kind of fixes everything, and then we’ll see who didn’t get fixed and deal with an inhibitor later.” And I think that’s kind of where we are right now for newly diagnosed patients who are older patients who are not going for intensive induction.

 I might interject, though, that in the really unfit, or the really elderly patients, I kind of want to know if they have an 

mutation, because I honestly would rather do HMA/IDH if I have the choice. So I agree with you for the most part.

 No, I think that’s exactly what I was alluding to, though, that in certain specific situations where you really don’t think that the myelosuppression is going to fly, might you wait for the IDH? Because both the single agent and the combination with venetoclax of IDH inhibitors are less myelosuppressive. And, actually, Courtney presented some data that some of those people seemed to recover counts faster, actually.

So I think those might be the right patients. The question is, though: How quickly are people getting their data back? And I think that is a major issue. Some people are getting the IDH data back quickly from PCR in 72 hours. Others are still taking 2 weeks. And I do think that, you know, 2 weeks of hanging around and waiting—one does have to be a little bit careful when you do have AZA/VEN as a potential compelling option. If the patient absolutely can’t tolerate it, then you can wait.

IDH Mutations in AML

Mark, I’m a leukemia doctor because I’ve only had to deal with giving 2 drugs at once, and that’s all I can do—7 and 3—40 years of that. And now we’re talking about triplets over the doublets. I feel like they’re trying to make me a myeloma doctor. It’s making me very anxious. Well, what’s your thought about adding in FLT3 inhibitors to HMA/VEN [hypomethylating agents/venetoclax] or HMA with a FLT3 inhibitor alone?

So we’ve been wrestling with that a fair amount. Obviously, we know that a 

 mutation, you don’t respond as well or you don’t respond for as long with an HMA /VEN combination. However, we’ve already just spent all this time talking about the myelosuppression with AZA [azacitidine]/VEN, and it’s real.

FLT3 is important for normal marrow to reconstitute. That multipotent progenitor cell has FLT3 on it; it wants to regenerate all those nice neutrophils and start that process. You inhibit FLT3, you are going to impair that. And when we combined gilteritinib with venetoclax, we saw some profound aplasias.

So what we do at my institution is a little bit different. We will pop the patient into remission with AZA/VEN if they have a 

, and then we actually sequence with AZA/GILT, and then we go back to AZA/VEN.

I don’t like using that triplet. I recognize that’s a great interest. A lot of people say, “Of course; it’s perfectly safe to do.” That hasn’t been quite our experience with the profound aplasias. So that’s been our approach; I don’t know the answer.

 Yeah, I agree with you, Mark. The triplet can be very myelosuppressive. I’ve had experiences where I gave the AZA/VEN. I clear an 

 clone, and then give gilteritinib [GILT] and kind of had an almost bounce between the 2 regimens because it’s hard to give them at the same time.

Yeah, so we’ve been sequencing AZA, AZA/VEN, AZA/GILT.

 So, before we leave this combination, any final thoughts about HMA/VEN that you want to bring up? Any of you.

Yeah, I would think that a cautionary tale is to not sound like we want to be advocating tons of mixing and matching. Because I think that it’s really hard to do sometimes, and I think that we are a victim a bit of our own success in having a lot of these drugs available. And I worry, actually, that it’s going to be very hard to learn what to sequence and how to change these doses and schedule if everybody kind of makes it up. And I philosophize about this a little bit because I’m not sure what the right answer is.

We’re in the middle of COVID-19 [coronavirus disease 2019]. People can’t get to centers. You want to try to do things that make sense. You have a smattering of data from here and there. So I understand that, and I do that, too. But at the same point, I think it’s going to actually be very difficult for us to answer the fundamental questions of which triplet? And is it an IDH inhibitor up front, or is it venetoclax up front? Or are you doing, are you alternating between them, or is 1 of the agents to be continued indefinitely? It’s really hard to answer this if we don’t do some studies.

 Gail, I’m going to come back to you when we get to the IDH inhibitors for that very thought about the triplet combination there. Dan, I think you wanted to ask something.

Yeah. I mean, it actually dovetails nicely with what Gail said. I mean, something you said earlier. You know: Pay attention to that plateau curve at the end, and let’s dig into who those patients are and how, what had made them be able to have this very long-term remission. And let’s learn about who those people are, and if they are people whose disease is being driven by a 

 mutation, then those are patients we should probably be leaving alone. And so understanding who those patients are and leaving them alone and giving them this very potent and reasonably well-tolerated, reliable therapy is the way to go. And then learning who the patients are who aren’t staying in these long long-term remissions and doing something like what Mark said makes a lot of sense to me. You know, sort of alternating therapies and sort of pushing on 1 clone versus another to try to prolong the remission. But like Gail said, I mean, I would really discourage us throwing everything in at the beginning because you can see, there’s 20%, maybe 30%, of patients for whom that will only represent overtreatment and excess toxicity.

 Dan, that dovetails into the question I wanted to ask you. You presented previously data from the HMA/VEN combination with clearance of leukemic stem cells. So my question for you is: Is there any data from any of the venetoclax data that a deeper remission, an MRD [minimum residual disease]-negative CR [complete response], means anything more than an MRD-positive CR? Is there any difference in survival that we know of between those two qualitatively different CRs?

Not that I’m aware of at this time, but I will say that I don’t think that the appropriate studies have necessarily been done. Or to answer that specific question: In the very near future, that’s going to be an extremely relevant question that we need to answer.

FLT3 Triplet Regimens for AML

It’s hard to imagine doing a trial to decide the best regimen. We are all applying our knowledge, our art, our experience. I’m curious, how would we design a trial to answer Amir’s question, how do we give this drug? Because every patient is different.

This is a nightmare. Dan Pollyea and I have been together on a couple of meetings where the goal was to try to get an expert consensus. We were cracking up at the end. Not only can we not get an expert consensus; we can’t even decide what you want to have for lunch, let alone this. In Colorado, apparently there are no fungal organisms so they don’t use it; Dana-Farber Cancer Institute doesn’t use antifungal prophylaxis. I absolutely do. The University of Texas MD Anderson Cancer Center does. I do marrow on day 21.

You have a lot of people who treat a lot of AML [acute myeloid leukemia], and of think we know what we’re doing. We concluded at the end of this that we are all either a bunch of geniuses or a bunch of idiots, because we couldn’t agree on anything. The problem is that bespoke medicine is nice, so you want to take individual excellent care of your patients, but it’s nerve-racking not to be able to figure out some consistency.

I would say the closest we have—I mean, we’ve got the Bible and the Torah and we have Dan’s paper on leukemia. All the PAs [physician assistants], if they can’t at least tell me about that paper and exactly what it says, I don’t want to have a conversation. Because at least that gives you something to hang on to. Please check the marrow. Please don’t be on day 60 of venetoclax without a bone marrow biopsy. Please don’t add an antifungal. If you’re going to add it, you’ve got to down the dose. At least certain basic principles have to be absolutely applied, even if the subtleties of exactly what day you do the marrow and exactly which antifungal can’t be mandated.

 I think it’s in the Talmud, Gail. It’s really far in there.

I was sure of that, but I wanted it to come from you.

I need to push back on Amir’s comment about patients who are pancytopenic and maybe sick with a fungal infection. You don’t want to give them venetoclax in the first cycle. I disagree.

I didn’t say that if patients are very sick with their leukemia, I don’t want to give venetoclax.

I said there are some patients because of their functional limitations or their age, I may worry about significant marrow suppression with the combination, and I may choose to give HMA monotherapy. Again, everybody is different. Do I see 85- or 90-year-olds with AML who I want to treat? I do.

But you’ve actually seen deaths from AZA-VEN [azacitidine-venetoclax] in some of those patients.

Right. Actually, we get very arrogant with this today for everybody. You’re 100 years old and this and that, and we’re going to do AZA-VEN [azacitidine-venetoclax]. Every now and then we get bit.

This is a bad disease but having count recovery faster is better.

That’s the key here. The last patient I just treated—a 90-year-old man, hand fungal pneumonia but cleared his blast quickly—I did the marrow on day 14, and it was clear: Start him on GCSF. Counts came up. You’ve got to get the leukemia to go away for the counts to get better, and that’s the FISH [fluorescence in situ hybridization] regimen. But I think what Gail said is exactly right: There’s not 1 size fits all. And Mark said it, there’s an art to giving these regimens. The community needs to learn about how to do this with experience and reaching out to people.

I also really worry, and I just want to emphasize this point: I worry that people got on the wrong bandwagon somehow with venetoclax because of the original CLL [chronic lymphocytic leukemia] approval. Obviously in practice settings there’s an awful lot more CLL than there is AML. Everybody knows the ramp-up, the TLS [tumor lysis syndrome], watching all that. That’s the stuff that we don’t care about in AML. Ramp up? I don’t even ramp up. I just get there. I think I’ve seen 1 laboratory TLS when I had to look for it.

These are the kinds of things I can say. I’m not advocating in any way not to be careful. I am very careful. I admit these patients unless they live within 3 feet of the hospital entrance. I treat them as induction. I watch them very carefully. But I don’t even know where to find the 20-mg dose of venetoclax or 50 mg. That’s for CLL doctors. It was a wrong focus of the toxicity, and it should be regarded differently. It should be regarded as intensive induction. My final point, which Harry and I love to talk about, is the end and death hopefully of fit and unfit. There are fit people and there are unfit people. You worry about everything. You want to give everybody this high response rate. You want to get them into remission. Whether you do a transplant or send them on a cruise after they get a remission, never mind. Just get them into remission. But the point is that you can’t. You can’t call somebody unfit forever and say, well, they’re not a candidate for a transplant because they might be a candidate later.

Similarly, you can’t necessarily call somebody fit forever because if they have a terrible time with induction and they blow up and have infections, that patient may actually want to go on or need to go on to some sort of maintenance. So flexibility in thinking.

I agree. We’re coming to the end of this part of the discussion. I’m going to turn to Mark for that. But I want to quickly come to a few other end points from VIALE-A that were important. You know we talk about a CR [complete response], CRi [CR with incomplete hematologic recovery] rate of about two-thirds of patients in the study. But what’s really impressive to me is that they weren’t just CRis, almost all of them were also CRhs [CRs with partial hematologic recovery]. Mark, what’s a CRh?

The other thing was that patients who were transfusion dependent at baseline had a higher rate of transfusion independent after AZA-VEN [azacitidine-venetoclax] than after azacitidine. We’re still waiting to see the quality-of-life data. But it gets to the point that Gail just made about fitness, which is important. If you look at the hazard ratio for survival in patients under the age of 75 years old in VIALE-A, that actually is pretty close to unity. But remember, for the patients to be under 75 years old to get on the study, they had to have very poor performance status or comorbidities. So we really need to continue to take very close care of these patients, especially older patients who with infirmities who may make them at higher risk of infectious complications and bleeding.

Monitoring Response to Venetoclax Regimens in AML

Let me move on to the next part of the conversation about this regimen. I agree that this is a new standard of care. HMA [hypomethylating agent] monotherapy should not even be considered for our patients anymore. To put the benefit of this in perspective, we have to remember that in the United States almost two-thirds of patients, at least half over the age of 65 years old, don’t even get treated for AML [acute myeloid leukemia], and that’s because of the low response rates important with HMAs. That’s because we didn’t really see a survival benefit with the HMAs. But this is different. We have a median survival of 15 months with AZA-VEN [azacitidine-venetoclax] compared with 10 months with AZA [azacitidine] alone.

Not only that, the curves aren’t coming together. What I find most interesting is that the curves aren’t coming together. A few of those patients, 10%, did go on to get a transplant in the phase 1 study, but that’s very few. I think 1 or 2 in the VIALE-A study actually got a transplant.

You have to put this into context because the median survival of untreated patients with AML, based on SEER [Surveillance, Epidemiology, and End Results] registry data are 2 months. It’s not 15 months versus 9 or 10 months; it’s 15 months versus 2 months. We have to get away from the therapeutic nihilism in older AML patients. In fact, the greatest survival benefit is actually seen in those over the age of 75 years old. That’s my commentary. We won’t see those response rates—we won’t see that improvement in survival—unless we manage the toxicities of this regimen the way they were managed on the clinical trial and really educate people how to do that.

I’m going to turn to Amir to start us on a conversation of some of the toxicities associated with this regimen. What are the pitfalls? What are the caveats? How do you get patients through this? Let’s start with tumor lysis.

Well, that’s an easy 1. There isn’t much tumor lysis with HMA and venetoclax in AML. There are reports, so it’s important to watch and be careful, especially with the first few doses. But it’s much less common than what you would see with chronic lymphocytic leukemia, a scenario that is a much more challenging situation. Thank you, Harry, for allowing me to follow up with the very pleasant and uplifting data associated with HMA-VEN [venetoclax]. I’m happy to do so because I really think there are some challenges with the combination.

I worry that this combination is being pervasively used and increasingly used in the community. One of the challenges with the combination of HMA and venetoclax is the lack of consistency regarding the approach to its use. Do you give it for 4 weeks? Do you give it for 3 weeks? Do you give it for 2 weeks? I’m referring to venetoclax. Do you pause after the first cycle of treatment? Do you do a marrow after the first cycle of treatment? If you have low blasts, do you wait or do you continue?

What do you do with additional drugs that may come onboard? For example, the azoles are the most common challenging drugs. These are scenarios that require a lot of education. I do worry, based on what I have seen in my practice, that this degree of education and pervasiveness of knowledge is not yet there. When you say that there are amazing data we see in the phase 2 or the phase 3 data, I’m extremely happy about that. Like everybody else on this panel, I’m very reassured by the data that were presented 2 days ago.

But I worry that the real world is also very different when it comes to HMA–VEN [venetoclax], just as it is with HMA. I can’t tell you how many patients we’ve admitted to our ICUs [intensive care units] who have been treated in the community with HMA–VEN [venetoclax] cycle after cycle after cycle and end up with severe marrow suppression, infections, and bleeding complications.

Ultimately what we need is a consistent treatment. I can tell you what I do. I suspect it’s probably similar to what other folks do here; maybe not. There might be nuances and that’s part of the challenge. What I do is I start with the 4 weeks of treatment with venetoclax, plus azacitidine or decitabine. And after the first 3½, 4 weeks, I do a marrow biopsy to see what I have. If the marrow is empty, I allow count recovery and then resume. If it is full of blasts, I go with the second cycle.

: Amir, let me hold you up for 1 second because you left out a very important clause. If the marrow is empty of blast, you allow count recovery by holding the venetoclax, right?

Of course. That’s what I meant to say, holding the venetoclax and basically waiting until starting the next cycle until there is count recovery. If this process of marrow suppression happens over and over again, there is a need for decreasing the intensity. Generally I reduce to 3 weeks or 2 weeks. There is a variability when it comes to that.

Finally, with concurrent medications such as the azoles—isavuconazole, voriconazole, posaconazole—you have to reduce the dose. That is something else that I see in the community that doesn’t always follow, even with fluconazole. It does require adjustment of dose; otherwise, you’re going to get in trouble with marrow suppression.

I don’t think this is the equivalent of HMA when it comes to tolerability. It’s not. It’s much better tolerated than induction chemotherapy, but it is less tolerated than HMA. Because of the enhancement and the efficacy and the improvement in the overall survival, I agree with my colleagues that it should be the new standard. But I don’t agree that there are no patients in whom HMA alone is not appropriate. There are scenarios where I’m really worried about marrow suppression and patients who have AML, and I may still use decitabine or azacitidine.

Venetoclax in AML: Safety Profile

Twenty percent of the patients in the VIALE-C trial with Ara-C [cytarabine] did have prior HMA [hypomethylating agent] exposure. But I’m not convinced, Gail, that that’s enough to explain the lower response rate in the shorter median survival of the venetoclax combination with Ara-C [cytarabine] than HMA.

That’s the primary criticism: Is that number enough? I’m definitely not in any way trying to argue that the data are as robust as they were for the other studies. I’m not trying to argue that.

However, I do think the response rates—having participated strongly in that study and watched the patients—are definitively better than for low-dose cytarabine alone, which are pretty crummy. There was enough benefit in patients as you go individually that one could imagine that it really was that short follow-up that prevented us from seeing the benefit. We’re trying very hard in a very rigorous academic way. Andrew Wei and I are trying to work on this to see if there’s any way to prove that point: Is this just a matter of statistics, or is it really not worth anything? We’re trying to answer that question as best as we can.

] paper of the phase 2 study, and they actually subtract out that subset of patients and then report the response rates. This isn’t survival but just response rates for patients who were truly treated and naïve. It’s hard to compare apples and oranges. But it gets close to the 60% range.

The other thing that may be a distinguishing feature—Andrew did a very nice talk about this at ASH [American Society of Hematology Annual Meeting] last year—is that perhaps the good subsets of patients who are highly responsive to venetoclax, the NPM1s, the IDHs, do as well with the low-dose cytarabine. It’s perhaps the tougher patients, the more secondary AML [acute myeloid leukemia] or the complex cytogenetics, or the genomically defined poor high-risk patients who aren’t doing as well with the low-dose cytarabine. Time will tell, but there may be ways to tease that out.

The other thing to do, and you’re not supposed to mix and match these comparisons, but we have the availability of a low-dose cytarabine-based regimen in combination with glasdegib, which is actually proving to be a little difficult for discussion. People aren’t exactly sure what to do there. It does not look competitive with respect to these newly diagnosed patients who were treated with AZA-VEN [azacitidine-venetoclax]. But again trying to just understand: Are there some patients for whom a low-dose cytarabine-based regimen would be something you wanted to try? Trying to answer the question of which 1 you might pick. It certainly seems that the venetoclax-based with low-dose cytarabine had much higher response rate.

Optimal Use of Venetoclax in AML

Let’s move on to newly diagnosed AML [acute myeloid leukemia] patients who are unfit for chemotherapy. I’m going to have Dan start with this. Dan, you’ve been a leader from the beginning in the phase 1B studies with hypomethylating agents [HMAs] with venetoclax. Why don’t you tell us about the practice-changing abstracts that have been presented at EHA Congress and ASH [American Society of Hematology Annual Meeting] regarding those venetoclax combinations.

Thanks, Harry. Most of us who treat AML have been very excited these past couple of years to see what venetoclax can do for some of these patients. As you said, the patients we’re focusing on here are older patients, who either by reason of age or comorbidities are not fit for standard intensive induction chemotherapy at the outset. They’re receiving either venetoclax in combination with either a hypomethylating agent or low-dose cytarabine.

With respect to the hypomethylating agent, we’ve had approval for the last year and a half based on some phase 2 data, a noncomparative study showing very promising response rates and survival. We all were anxiously awaiting Courtney DiNardo’s presentation at EHA [European Hematology Association], which was going to finally show us in a randomized setting how venetoclax plus azacitidine did in this population versus azacitidine alone. Most of us who have worked in AML know this is a critical test, a time when a lot of prior therapies have not been able to surmount this challenge in a randomized setting.

Most of us who treat this disease were relieved and very happy to see the data for the venetoclax arm really performed very consistently with what we have come to expect, based on the phase 2 data. We were not disappointed, at least specifically looking at the venetoclax patient—you know, a response rate somewhere north of 60%, 65% and a survival on the order of 15 months. The critical thing was to compare it with the azacitidine arm, and we’re ready for some twist or something thrown at us, in the underperforming or overperforming arm. In fact, the patients who received azacitidine performed pretty much how we would have expected them to historically. We saw a nice statistically and clinically significant improvement with respect to response and survival.

Those of us who have been using this frequently in the last year and a half, based on the approval, are reassured that this likely the new standard of care in this setting.

Just to comment quickly, because it was raised: There’s also the question of combining this with low-dose cytarabine, which is a much more commonly used drug in many other places in the country. Similar enthusiastic data around the phase 2. We had, based on a press release and some other information, the impression that the study would not be positive when compared with low-dose cytarabine alone. Gail is, I believe, the senior author of that study that was just published in 

. I encourage people to read it, because when patients were followed for a little longer than the original planned analysis, which is definitely a weakness of the planning of the study. But the reality of it was that when you follow patients to a 12-month time point, there was a survival benefit.

There are a lot of explanations for why the low-dose cytarabine patients seemed to underperform. It is, by definition, a tougher patient population. Upward of one-third of those patients were allowed based on who had had a prior hypomethylating agent for an MDS [myelodysplastic syndrome]. Those patients were completely excluded from the azacitidine arm. There are a couple of explanations there. But it gives us more options. For us in the leukemia world, there’s no way that’s not a good thing.

 The VIALE-A study, as it’s now kind of widely known as, looks at azacitidine plus venetoclax versus azacitidine alone. That’s the new standard of care for AML. It’s actually really rough in Europe, where they don’t have it yet, and in other parts of the world where they don’t have it yet, because we’ve seen incredibly higher-than-usual response rates. We’ve been stuck in 28%-to-30% response land for older patients with AML for decades. Now that’s not true. Those over-60% response rates across cytogenetic groups, across molecular groups are real. They happen. They’re quick. You don’t have to wait for 100 cycles of azacitidine to see those responses. It’s usually 1, sometimes 2, but usually after 1. It’s just plain better. I think Dan is absolutely right that we were relieved. First of all, there were a couple of other studies presented at ASH in which the azacitidine arm was looking a lot better than we thought. There were patients hanging out for a year or 11 months of overall survival. This, of course, was a survival study. So that was a little worrisome.

Then you have the VIALE-C trial, the low-dose Ara-C [cytarabine] study, and actually I am a senior author on that paper. Andrew Wei is first. We’re putting together an editorial to explain that. Clinically, it’s a useful regimen. Those response rates were real. It is absolutely overshadowed by the azacitidine data, but sometimes we really do have patients who had 10 cycles of the hypomethylating agents before they officially convert to AML, or who are in a situation where you just don’t think that another round of HMA is going to work.

The question is, can some of those patients still benefit from the addition of venetoclax to low-dose cytarabine? I still think there are patients for whom there might be a benefit. Much like the phase 3 data we saw years ago regarding decitabine, which initially did not show a survival benefit but then with longer follow-up did, I think those studies are statistically kind of messed up. The reason you’re not seeing the benefit is more because of disadvantageous study design with respect to hazard ratio and statistics, rather than because the regimen doesn’t have any benefit.

AZA-VEN [azacitidine-venetoclax] is the new standard of care. There are probably some patients for whom I would try low-dose Ara-C [cytarabine]–venetoclax if you really don’t think more HMA is going to be of benefit. The rest of the world is scrambling to try to get the regulatory approvals so that they can apply these—the practice-changing regimen for their patients.

VIALE Studies in AML

Amir, I want to bring you into this discussion. If you have a young patient, 

 mutated, and they’ve relapsed, you know you’ve got to get them to transplant. Doesn’t it just make sense to give them chemotherapy to get them to transplant? Don’t these targeted inhibitors take longer? Do they get you in a deep remission?

 Eventually. If I may start with some principles that have been alluded to in passing, which are very important in my opinion when it comes to FLT3 inhibitors. One is tolerability. There is a marked difference between tolerability of, say, midostaurin versus gilteritinib. Or even sorafenib versus gilteritinib. Midostaurin does cause a good amount of GI [gastrointestinal] toxicity. It’s uncommon for me to give midostaurin to somebody and not have them at least have mild GI issues, whereas gilteritinib is much better tolerated. That’s an important consideration.

The other consideration is the type of leukemia patient. This is extremely important. Now that we’re doing NGS [next-generation sequencing] panels, we’re doing additional mutational testing. We’re discovering additional 

 TKD variance. Some of these FLT3 inhibitors, as Mark just alluded to, don’t really work. Sorafenib and quizartinib don’t have much activity.

What I oftentimes worry about in the community when we see patients coming in with 

 mutations or potentially treated with FLT3 inhibitors, for a mutation that would not be expected to respond to that 

 patient. Those factors are very important.

Finally, we talk about RATIFY, and this question has come up multiple times in the past. What is a dirty FLT3 inhibitor—let’s say midostaurin or sorafenib—doing to these patients? Is it mainly having a therapeutic impact through suppression of 

 mutations? Or is it potentially having other effects on other pathways that it may be inhibiting? It’s probably FLT3 because of the patterns that we have seen, but I’m not sure if that is completely the case. For example, the Soranal study showed that across-the-board patients without 

 mutations seemed to have some degree of response.

The quizartinib study showed that patients with FLT3 wild-type disease have a response. These are all questions that are yet to be fully answered. The answer to your question giving to a younger patient who may relapse, I’m not sure if it’s clear-cut that I would necessarily combine it with chemotherapy. That question has not been sufficiently answered, so I could not fault somebody if they want to put their patient on monotherapy with gilteritinib, with remission rates at approximately 50%; you could potentially decide to combine it with an HMA [hypomethylating agent]. The data on that is emerging. It’s not final, and we don’t know if it’s substantially better. I suspect it will be better, but I don’t have any definitive evidence to suggest that. I do not know if I personally would beat up a patient, even a younger one, with intensive chemotherapy, and gilteritinib, in the relaxed refractory setting.

Relapsed/refractory treatment for AML [acute myeloid leukemia], the remission rate, as we all know, hovers around 40% to 50%. Even in younger patients, it is not, depending on the timing of relapse, necessarily more impressive. My own personal preference is either give it as monotherapy or combine it with a hypomethylating therapy.

I would just jump in with 1 comment that sorafenib gets a little bit of a bad rap. It’s not the easiest drug to take. Mark Levis always says that all his patients’ hands and feet fall off. Mine don’t do that. But that drug has magical powers in the post-transplant setting that are different. I think both the scientific and clinical data suggest that the post-transplant relapses really might be something important to consider for some patients with sorafenib that could actually be different from the others. I just put that plug in that sometimes it’s just dismissed as old and not easy to take. But I think that specific setting is 1 that is actually quite interesting, especially in combination with HMAs.

Utilization of Second-Generation FLT3 Inhibitors in AML

Let’s move on to second-generation drugs. Mark, I’m going to come back to you on the development of both quizartinib and gilteritinib in this setting. And of course there’s crenolanib if we have time.

Yes. Actually the 3 drugs were coming in parallel. Quizartinib is the older 1 that was designed specifically as a FLT3 inhibitor. It had a bumpy regulatory course, but they finally emerged with this QuANTUM-R trial. Now quizartinib is a type 2 inhibitor. It does not inhibit the TKD mutations. It’s very specific for the ITD. Very potent, very selective, and for patients with relapsed or refractory 

 ITD AML [acute myeloid leukemia], they were randomized to receive a conventional MEK, or vicious salvage, or LoDAC [low-dose cytarabine], or quizartinib as a single agent.

Although that trial met its primary end point of overall survival, and the Japan regulatory agency thought that was fine, Europe and the United States did not. There was wringing of the hands over well QT prolongation. This drug seems very toxic, which is startling to any of us who have given it compared with MEK chemotherapy. Their point was it wasn’t a big enough improvement and that it really looked good only against LoDAC. But we can quibble over that. The simple truth is that it is an effective agent, and it is being studied in a different context as well in something called QuANTUM-First trial, which is up front. We can talk about that. But that’s quizartinib for the relapsed-refractory setting, a type 2 selective potent inhibitor.

The other 1 is gilteritinib. Gilteritinib was developed, and I like to use this analogy: The developers of gilteritinib watched all the mistakes made, and we were doing quizartinib. They just waited for us to spring all the traps and find all the screwups. Then they just marched on through very rapidly and casually developed a very good drug, which is in fact a type 1 inhibitor. Less potent but still very potent in vivo and really quite selective for the 

 mutation. It got approval for the relapsed-refractory setting and is now being studied up front versus midostaurin in combination with chemotherapy.

Finally, crenolanib is still back there. We’re waiting to see what’s happening with that drug. Really the focus, for practical purposes, is still on quizartinib and gilteritinib.

Second-Generation FLT3 Inhibitors in AML

Gail, let’s move on. Your patient comes in with newly diagnosed AML [acute myeloid leukemia] and has a 

 mutation. It doesn’t matter. I just heard that if it’s an ITD or TKD, we now add midostaurin. Why?

 The RATIFY trial, which took a long time and a lot of centers to accomplish, really demonstrated a survival benefit with the addition of midostaurin to standard conventional chemotherapy. That very quickly was adopted as the standard of care. We were really ripe for looking for something to add to 7+3, first of all. Secondly, that was ushering in an era of targeted therapy. So you have these patients who are coming in. Many of them are actually quite proliferative and scary. There is a proliferative signal conferred by the mutation. These are the ones who come in sometimes with a 150,000 per mm3 white blood cell count, a head bleed, or other things that are going on.

I think Mark’s point about get the testing done is to figure out what you’re dealing with. We’re going to add midostaurin. Midostaurin is generally quite well tolerated. I would say there are some rashes, some GI [gastrointestinal] toxicities that can happen. There’s nausea. There’s vomiting that can happen, so you have to look out for that. But in general, the overall aggregate data showed that there was a clear win with respect to survival. Actually the suggestion, which was alluded to a few minutes ago, is to use FLT3 inhibition throughout. FLT3 as part of the initial therapy.

If you’re not getting a transplant, they’re getting FLT3 inhibitors as part of consolidation and even subsequent maintenance therapy. After transplant there is a strong trend for maintenance too. Transplant is not, unfortunately, actually the end. It’s “one and done,” but there’s ongoing FLT3 inhibition that is largely practiced, not completely with data but with strong suggestive data that ongoing FLT3 inhibitors might be a good idea for some patients even post-transplant. We’re waiting for the confirmatory data for that.

Right. For now we just have to be careful. Obviously, the news is whether gilteritinib, for example, would be moved forward. And for induction consolidation, post-transplant maintenance, what would be the role there? Is it better to use the midostaurin? Is it better tolerated? For right now midostaurin in combination with chemotherapy is the up-front standard, but there is the anticipation that gilteritinib may have a lot to offer in both up-front and post-transplant maintenance.

 Dan, let me ask you a question here to pull you into this conversation. You know midostaurin maintenance is approved in Europe but not in the United States based on the same RATIFY trial data. This is because of the peculiarities of the design, where patients were rerandomized at the time of maintenance assignment. Do you use midostaurin maintenance? Is there anything that makes you think it’s really valuable here?

You know in our setting, much like what has been referred to before, the eligibility criteria there for a chemotherapy 

-positive patient who gets induction plus midostaurin—if they get into a remission, there is going to be a transplant. There are all kinds of contingencies and bumps in the road that could take you off that path. But for the most part, it’s not a big issue for us for that reason. If a plan from the get-go was for them to go through reduction, then they’re going to go to a transplant and move on from there.

Whether there should be a roll for post-transplant maintenance with an agent, Amir and Mark already kind of referred to that. For us it’s a sphere without a real treatment for that reason. What I will say is that we’re in that situation where if a patient gets induction, gets into a remission, and does not go to transplant, we would consolidate that patient with midostaurin. We would follow essentially what the RATIFY study randomized patients to with a maintenance program with midostaurin, even if that’s not the labeled indication.

My objection to that approach, which is a reasonable approach, is that we know more than half the patients who are on this bad-smelling pill come off it because they can’t tolerate it. In fact, the mLSG-15 study basically had 57% of patients leave the treatment because they couldn’t tolerate the stuff. So it isn’t that we don’t like it.

 Well, you opened that can of worms. The FDA doesn’t want to be ageist, but the RATIFY trial was in patients up to the age of 59 years, not 60 years or over. The data that was used to support the use in older patients is the German AML1610 study that took older patients.

But only based on historical comparator was it better than chemotherapy without midostaurin. You’re right, they saw more toxicity in those older patients as well. But at the end of the day I’m still impressed by Dick Larson’s presentation of what happened to patients who were known to be on midostaurin for maintenance. When they stopped it, there was this rapid 25% decrease in the disease-free survival, which suggests maybe in some patients it’s doing some.

And patients relapse. We’re going to see this data soon. Patients relapse from that treatment with the ITD mutation still present. About half of them will relapse. When they relapse they still have the ITD there. So there’s potentially a role for it.

FLT3-Mutated AML: Midostaurin and Chemotherapy

Let me ask a question of Amir about this allelic ratio and how it impacts further discussion. Amir, are there patients who have such a low allelic burden of 

 and maybe other mutations where allotransplants should not be recommended?

If you ask multiple leukemia doctors you’re going to get a series of potentially different answers. Having trained with Mark, I have learned to have a very healthy fear of 

 ITD disease. It tends to relapse; it tends to come back very quickly. Relapsed disease is oftentimes difficult to overcome, although we do have much better targeted drugs right now to manage them.

 ITD disease—I recommend midostaurin with induction chemotherapy followed by transplantation. Thereafter, institutionally we recommend post-transplant maintenance therapy. And Mark Levis and a colleague of mine at Mass General Hospital have led a very large phase 3 study, which we hope to have results from in the fairly near future, to let us know ultimately if a very effective potent selective FLT3 inhibitor is going to be helpful in that regard. So I have a healthy fear of it. I recommend aggressive therapy, and I recommend a good amount of FLT3 inhibitor therapy spread throughout the course of treatment.

Looking more closely into the methodologies of transplant, which is what I’ve been spending my time doing, not all transplants are the same, and not all transplants offer the same low toxicity. The donor matters, the prep matters, and the circumstance matters. Harry’s point is a good one. Low allelic ratio, 

Gail J. Roboz, MD, Weill Cornell Medical Center

Articles