I want to thank all of you for this rich and informative discussion. Before we conclude, I’d like to get final thoughts—a few minutes only, not a dissertation; don’t pretend you’re Harry Erba and go on for a half hour—from each of you on exciting new developments and emerging agents. I’m going to start with Dr. Fathi to discuss what’s exciting him about immunotherapies like ADCs [antibody-drug conjugates], and BiTEs [bispecific T-cell engagers], and things like that.

Just briefly to answer your question, Harry, I think it’s an exciting but early field. There are a whole host of targets that are potentially relevant in AML [acute myeloid leukemia]. The oldest targets are CD33 and maybe CD123. There’s a range of antibodies, antibody-drug conjugates, novel bispecific agents, novel DARTs [dual affinity retargeting molecules], CAR T [chimeric antigen receptor T cell] therapies that are now actually looking at additional targets in addition to those I mentioned, such as CD7.

CD47, a naked antibody, seems to have some legs, at least in combination with hypomethylating therapy in MDS [myelodysplastic syndrome] and in AML. We’ll see ultimately where that data is, but it seems to have some promise potentially in patients who have 

 mutated AML. The surface of the leukemic cell as opposed to the alterations within it may also potentially provide us with novel approaches to treatment.

Thank you. Mark, give us your thoughts about moving the second-generation FLT3 inhibitors in combination with initial intensive chemotherapy.

It’s nice that we have lots of choices but we really need to move into refining what we have and how to use them. Is there an advantage to using a more potent selective inhibitor up front and transitioning to a type 1? Is there a reason to be using maintenance, a FLT3 inhibitor after transplant? In whom and using MRD [minimal residual disease]? In other words, we actually have the drugs, I think. We have to learn how to use them. We have to go back to good old-fashioned discipline and do very careful large-scale trials with very specific molecularly defined questions. There’s still lots of work to do but progress is being made.

Dan, 1 of the most difficult subsets of patients to treat are those with 

-mutated AML. Anything besides 10 days of decitabine out there?

Gosh, I hope so. But you’re right: t’s been a long time, and this has proven to be resistant or refractory to most of what we’ve thrown at it. There’s some hope now. APR246 is something we’ve seen several presentations on, a specific 

-directed therapy for patients with that horrible mutation. Both in MDS and AML, there is some real promise.

The biology of the way that drug works is really exciting and different, and we’re all anxiously awaiting to hear more. Some of the stem cell–directed therapies may have the ability to neutralize, to some degree, 

. When you look at venetoclax combinations, response rates are in the 50% range—lower than non-

-positive patients probably, although statistically we still haven’t quite seen that. But certainly the response durations are not as long for sure with 

-positive patients. But venetoclax is still certainly an option to consider, and then other stem cell–directed therapies. Amir mentioned magrolimab, a CD47 antibody that has some stem cell–directed therapeutic potential, seems to maybe have some activity in the 

-positive subset. Around the corner I’m hopeful this will be something we can work on solving soon.

Thanks, Dan. Gail, you get to talk about the hottest topic that we have right now on our inpatient service, and it’s not COVID-19 [coronavirus disease 2019]. It’s CAR T cells. CAR T cells in AML—is there a future there?

Well, we want there to be a future there. With CAR T cells, the appeal has been, “Hey, wait a minute, that’s curing some people.” So we have cures with CAR T-cells. What can we do to move curative therapy to patients with myeloid disease, patients with AML? Our most powerful immunotherapy is transplant. We cure some patients with transplant, but transplant is a pain. We don’t cure a lot of patients with transplant, so what about CAR T cells? The land of the myeloid hasn’t come to full bloom yet. I’ve been particularly interested, as you know, in looking at allogeneic CAR T cells.

Definitely 1 of the glitches with CAR T cells, which is profound, is that the patient has to be able to hang out until the cells are manufactured, and they also may have some extensively beaten up cells that may or may not survive what’s needed. Neither the patient nor the T cells may be adequate to get them to CAR T. We haven’t had this available for younger patients. It’s not worked in myeloid. What about taking really robust, fantastic cells from somebody else and hitting the right target. We’re looking at CD123, but as was already mentioned, there are numerous targets in AML. What about having those cells eliminate the possibility for graft-vs-host disease? We want to overcome issues of T-cell persistence. But that would be a very hot topic for AML, with at least the allure of potential cure since CAR T cells have cured other patients with other hematologic malignancies.

Thank you again, and thanks to our viewing audience. We hope you found this 

Videos

Progress in the Treatment of AML

Gail, a couple of other things that came from the QUAZAR study that caught my attention. Unlike the last German study of 12 months of parenteral azacitidine, where those patients just got induction and then got maintenance or observation, in this study, 80% actually got consolidation. These were correctly treated patients with intensive chemotherapy. But I would also point out that in terms of that transplant option, the median survival was clearly better. But those survival curves, unfortunately, come right back together down around 2 to 4 years. It’s not a very good outcome. I personally wouldn’t replace transplant with this without a properly controlled study showing that.

My final point before I let you jump back in is that I’m really happy to hear you say that your patients tolerated it, because quality-of-life assessments make me a little anxious because I‘m not really a QoL [quality of life] kind of expert. I was worried about the grade 1 GI [gastrointestinal] toxicity, a few more infections, especially in older patients populations. But if you tell me your patients are tolerating it, and that’s New York, so they must be really tolerating it.

A couple of things. First, I definitely agree with you that we are not prepared to say do this instead of transplant. But there are lots of nuggets of interesting things. For example, we know from Chris Hourigan’s data from EHA that if you’re an older patient and you’re going into a reduced-intensity allotransplant with MRD [minimal residual disease], that’s not going to go well. You’re not going to get cured of that. We have some data, both from this study and from others looking at azacitidine that might you be able to reverse MRD. Can you go from MRD positive to MRD negative? We are working on looking at those data.

If you have a patient in their late 60s who would be eligible only for reduced-intensity transplant, who has a significant proportion of MRD, for example, by next generation or by flow, you know that that patient’s outcome in transplant is not going to be what you want. Might you think about a maintenance approach in that patient? Again, these were questions that were not definitively answered by QUAZAR, but in putting together what we know about MRD, about the reduced-intensity transplant option versus a myeloablative transplant option for some of these older patients, it’s thought provoking.

From a quality-of-life perspective, we have to present the data exactly as they are. But let’s look into the nuances of it. First, the quality-of-life measures. We really worked hard on this. There were multiple significant tests. They were done throughout the entire follow-up period, which was long. What we saw on the GI [gastrointestinal] toxicity, it maybe took a couple of cycles for patients to straighten out. If you talk to my patient who’s been on it forever, she’ll tell you, “Oh, on those days, I might take an extra Pepto-Bismol, or I might not eat cheese on those days.” People work out what they need to do to manage the GI toxicity, and then it gets better. If you look at the cycles where it happens, it’s incredibly well managed.

Similarly, and this is not completely clear unless you dig into the weeds, but yes, there is some myelosuppression. But the significant myelosuppression occurred in the study at the time when patients were relapsing from their disease where the toxicities were because there was leukemia there, not so much because of the drug. So I actually was very, very encouraged and reassured that this is a feasible long-term ongoing therapy that doesn’t wreck people’s lives.

Gail, those insights from an investigator on the trial are very important in considering maintenance where we know quality of life is going to be important for these older patients. There were a couple of follow-ups to QUAZAR. One of them was the quality-of-life data that Farhad Ravandi-Kashani presented. The second is that in the study, patients for whom the blast count started to go up between 5% and 15%, they went to a 21-day schedule from 14 days, and about 25% of those patients responded, and their survival looked to be better.

Yeah, there were 3. Ravandi presented, so I presented the quality of life. That’s actually in a talk which is I believe available for eternity on video. Then Farhad did a poster on the patients over 75 years old.

He did a great job. It’s 1 of these interactive posters that I found to be a fun aspect of the virtual meetings. Somehow, when you’re in a poster hall with 9000 people and bad beer, bagels, and stuff, you can’t focus as well as when the author is walking you through it. But he will walk you through that poster if you look at it on the EHA site. You can see that these are older patients. These are exactly the patients you would like not to do a transplant on, and they actually had a survival benefit, tolerated the drug, etc.

The third data set was regarding the salvage option of patients who had from 5% to 15% blasts who look like they were in early relapse, and actually by increasing from 14 to 21 days in those patients, we were able to get some of those patients back on track. That goes along with Uwe Platzbecker and some of the RELAZA2 data suggesting that staving off relapse is a potential concept of application of HMAs [hypomethylating agents]. They were not using CC-486. They were using regular azacitidine in that.

The benefit in survival in the MRD-positive patient subgroup—there’s a lot here. I agree with you: This is going to become a new standard of care. Imagine that, maintenance in AML after all this time. Hints along the way, but you’ve shown it now.

Implications for Treating AML With CC-486

 Gail, you’re going to lead the last section before we get to concluding remarks. I would like to preface this, maybe controversially, by saying maybe this is the next practice-changing thing that’s coming for AML [acute myeloid leukemia]. Tell us about CC-486, also known as oral AZA [azacitidine].

 I don’t think it’s controversial. I’m going to just go out there and say that this is going to be practice changing in AML because I’ve been working for so long that it has to be practice changing in AML, otherwise it will be too depressing. But we had some very gratifying data that were released at ASH [the American Society of Hematology Annual Meeting], and then we had some follow-up data that were presented at the EHA [European Hematology Association Congress] looking at CC-486 in the QUAZAR study. This is an oral formulation of azacitidine. I want to say quickly and out there in front that it’s not just the same thing. It’s not going to be a 1:1 easy substitution, and there’s lots of discussion about where CC-486 could potentially be used instead of currently available formulations of AZA [azacitidine]. But that’s not what this study was.

What this study looked at were patients who were treated with intensive induction chemotherapy, older than 55 years old, who were deemed ineligible for stem cell transplant. Patients were in remission. Patients were in remission, CR [complete response] or CRi [incomplete count recovery], and then they were randomized to CC-486 versus placebo. It’s a clean study, hundreds of patients, international, placebo controlled, all the good stuff. Basically there was a very clear overall survival benefit for the patients treated with CC-486.

Some of the additional data, which were really gratifying, showed, first of all, the patients were in pretty good shape in terms of health-related quality of life when they started the study. They were reporting solid measures of quality of life in remission. For the ongoing therapy, we’re talking 12 cycles of therapy. A lot of these patients were on it for a long time. I actually have the longest patient on it in the United States. I think we’re at 7 years at this point.

These patients were able to maintain their very good quality of life, so this was feasible to administer. It was also feasible and tolerable in older patients, in patients more than 75 years old. For those patients, yes, you can do transplant, but it’s a struggle. A lot of patients over 75 years old don’t want to do a transplant even if they can do it.

Maintenance has not shown an overall survival benefit to date in AML. This is a first. It’s with a tolerable agent. I think it’s going to change practice. It’s going to be stressful initially to figure out who should get this. If the question is, should I not do a transplant, should I get this instead, this study didn’t answer that question. This wasn’t a head-to-head study that do you do maintenance or do you do transplant. But we’re digging very hard into the data to try to answer as many questions as possible about which patients will benefit from maintenance. In combination with other studies that have suggested a maintenance benefit for hypomethylating agents [HMAs], this is suggestive that ongoing treatment with an HMA might be a good thing. After regular chemotherapy, it might be a good thing to have ongoing cycles, for example, of HMA–VEN [venetoclax]. Will you be able to substitute CC-486? I’m not sure.

Of course the post-transplant azacitidine is another interesting area to investigate, since there have been some data suggesting restoration of chimerism and staving off relapse by using azacitidine post-transplant. I find these very encouraging data, very thought provoking, and I can easily envision another decade of study to figure out what we can actually do here.

CC-486 Maintenance Therapy in AML

 Let me, for the sake of time, move on and turn it to gemtuzumab. How do you use gemtuzumab? In June the FDA approved gemtuzumab for pediatric AML [acute myeloid leukemia].

 I didn’t know that. Thank you. I should know that, but I didn’t know that.

 I can’t comment on that, but the answer to your question is that I use gemtuzumab sparingly. The scenario in which I definitively do use it is in patients who have core binding factor AML and who are younger. We know you generally combine that with chemotherapy. One of the challenges with that of course is getting those results back, so you have time enough to incorporate the treatment. But I do use gemtuzumab in that setting. I also use it in consolidation as well in an effort to cure them. In other scenarios, I have not used it in older patients as single therapy. I know it’s approved for use in that setting. I have not used it in intermediate-risk AML.

I have used it a handful of times for relapsed/refractory disease when I don’t have anything else. I have nothing against gemtuzumab, but it’s not the easiest agent to give. There are infusional reactions in patients who have relapsed disease. You have to worry about the history of transplant. You can dose at a lower dose and decrease that risk, but you’ve still got risk. And I’ll be honest with you, the rates of remission, rate of response is very modest, with a high potential for toxicity in certain patients. So my use of gemtuzumab is very low.

 The other place you haven’t mentioned, I’m sure you’ve probably used it here, would be the high-risk APL [acute promyelocytic leukemia] patient. Instead of using an anthracycline, using a single dose of gemtuzumab for a count reduction.

 It’s hard to get around the meta-analysis showing a survival benefit of gemtuzumab when added to chemotherapy. We’re talking over 3300 patients, a subset with core binding factor [CBF] leukemia is clearly doing better. Some have argued that they didn’t do so well with just chemotherapy, but actually core binding factor leukemias, the overall survival is only around 50% to 60% at best without gemtuzumab. In that study, in that meta-analysis, it was in the 75% range. I think there’s something there.

 I actually have been surprised by the seeming lack of knowledge about those particular data. I guess CBF leukemias are rare. People may not see them much, but it’s important to emphasize on a call like this that those are resoundingly positive data. Most of us view the addition of gemtuzumab in the CBF, the core binding factor leukemia scenario, as in fact standard of care. I still get consults for transplant, consults for all kinds of things for CBF patients who have not received those. In particular, when core binding factor leukemias are combined with many other cytogenetic abnormalities and/or a 

 mutation, which are scenarios that happen, those are the particularly confusing ones that then result in, well, does this patient need a transplant? Does this patient need this, that, or the other?

I think we want to keep it pretty pure. If you have a core binding factor patient with or without other cytogenetic abnormalities, with or without 

, you probably do have a good shot of curing that patient with chemotherapy and gemtuzumab without doing a transplant and without switching that patient. We’re asked very frequently, should that patient get 7+3 and midostaurin, or 7+3 and gemtuzumab, or 7+3 and both of those drugs? Mark is here, but I believe the best answer to that is to focus on the core binding factor and not on the 

 We get our FISH [fluorescence in situ hybridization] back for core binding factor within hours, so the GO, gemtuzumab ozogamicin, is already hanging before we know the 

 status. Our usual decision is they’ve had 7+3, they’ve had GO [gemtuzumab ozogamicin], and they say, “What do we do with the 

?” And we often throw it on if the patient is still talking to us.

 I actually think people are much more—it’s somehow much more prominent in their mind to be looking for the 

 FISH [fluorescence in situ hybridization].

 Our algorithm is find that core binding factor fast.

 We have to emphasize that point, that these are your curable patients. These are the ones you want to call.

 Some of your neighbors up there don’t do that, Gail.

 I moved to the Duke University School of Medicine 2 years ago and I’ve taken a little heat for giving GO [gemtuzumab ozogamicin] to the CBF patients as opposed to giving midostaurin alone to those patients if they’re 

 positive. But they’re getting used to me, and I’m hoping they’re not going to kick me out. I kind of like Durham, North Carolina. It’s beautiful here.

Gemtuzumab for AML

 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.

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.

Amir Fathi, MD, Massachusetts General Hospital

Articles

Progress in the Treatment of AML

Implications for Treating AML With CC-486

CC-486 Maintenance Therapy in AML

Gemtuzumab for AML

CPX-351 in Newly Diagnosed Secondary AML

Differentiation Syndrome and IDH Inhibitors in AML

IDH Inhibitors and MDS

IDH1/2 Inhibitors for AML

IDH Mutations in AML

FLT3 Triplet Regimens for AML

Monitoring Response to Venetoclax Regimens in AML

Venetoclax in AML: Safety Profile

Optimal Use of Venetoclax in AML

VIALE Studies in AML

Utilization of Second-Generation FLT3 Inhibitors in AML

Second-Generation FLT3 Inhibitors in AML

FLT3-Mutated AML: Midostaurin and Chemotherapy

FLT3-Mutated AML: Early Aggressive Therapy vs Transplant

FLT3 Genetic Alterations in AML