Acute Myeloid Leukemia Podcast

Treating Older AML Patients and Managing Unique Toxicities

Gail J. Roboz, MD; Courtney D. DiNardo, MD, MSCE


July 20, 2022

This transcript has been edited for clarity.

Gail J. Roboz, MD: My name is Dr Gail Roboz. I'm a professor of medicine and director of the leukemia program at Weill Cornell Medicine and the NewYork-Presbyterian Hospital in New York City. It is my total pleasure to present to you my colleague and friend, Dr Courtney DiNardo, from the M.D. Anderson Cancer Center in Houston. Hi, Courtney.

Courtney DiNardo, MD, MSCE: Hi. Good to be here, and thank you.

Roboz: We're going to talk today about some difficult topics in the management of acute myeloid leukemia (AML). We're going to be discussing some unique toxicities related to some of our new and exciting drugs in AML and also specifically discuss management of elderly patients. The median age of diagnosis of AML is 67, 68, 69 years old, depending on where you are. I want to just open with the discussion of forgetting the definition of elderly. But at M.D. Anderson, is there an age cutoff where you would say that intensive induction chemotherapy just can't be done?

DiNardo: No, there is no strict age cutoff. We'll take that a step further and say that things have changed and we're going to get there in terms of treatment options. And so one of the most important things is that a decade ago, when we had standard intensive chemotherapy as the only option for a patient to derive a durable, ongoing benefit, we wanted everyone to be fit for intensive chemotherapy. And so we were willing to take people who maybe had some underlying organ dysfunction or were of older age because that was their chance for a cure. The discussion about age and fitness is still very relevant, and it is still very important to help us choose what the treatment is. But it's not quite as black-and-white as it used to be because we have different options now where patients can derive benefit that is not a standard cytotoxic therapy. We have patients who have core binding factor leukemias. And we know they do the best with cytarabine and Mylotarg (gemtuzumab ozogamicin)-based therapy. And so those patients, even if they are in their 70s, if they are fit for intensive chemotherapy, we are going to optimize that. Whereas for other patients and intensive chemotherapy, the likelihood of response to that isn't the same. That decision tree is a little bit different.

Roboz: That launches an interesting discussion of when we say "elderly," maybe what we really mean isn't actually an age. Maybe what we're trying to talk about is somebody who has some problem with a standard intensive chemotherapy. And we're happier now because we're not relegating everybody who can't get standard intensive chemotherapy to a prognosis of under 6 months. We have other things. We have hypomethylating agents (HMAs) combined with venetoclax, which is showing lots of promising results much better than HMAs or low-dose cytarabine alone. But maybe what we're trying to say in this talk is actually that, sure, there are patients in their 80s and 90s where you're going to use the term "elderly" and where you're going to probably have a not hard stop but a somewhat age-based cutoff. But what we probably mean is that there's something else going on with the patient that is making you hesitate about hanging intensive daunorubicin-, cytarabine-based therapies or, at M.D. Anderson, even the intermediate dose or even more intensive induction strategies.

What do we mean by that? And this is what I really want to dig into in the episode — that AML patients are complicated. The first thing is to note that HMA plus venetoclax isn't only for the elderly. I mean, if you look at the eligibility for VIALE-A, it's not an old-people-only trial, and I want you to talk about that for a minute, because again, there's this concept that older than X needs X therapy, younger needs Y. But similarly, there might be frailties or infirmities or organ system comorbidities that make a young person seem like an elderly person. What are those, and what are the Ferrara criteria getting at?

DiNardo: That is exactly right. And it is important to get away from just a specific age cutoff. There are other wordings in some of the trials or other medical problems that, in the opinion of the investigator, are going to impact the tolerability of your patient. So, for me, that would be a patient who has a prolonged history of autoimmune diseases, who's been on biologics and steroids and various things forever. And so the risk of infection, for instance, with the standard intensive chemotherapy would be much, much higher.

Roboz: One of the misconceptions that is still fairly pervasive is that let's say you've got some problem and you're not getting intensive chemotherapy, whether it's advanced age or comorbidities. What happens to the patient who isn't going to be getting an intensive induction but who's presenting with an elevated white blood cell count? How do we get that patient on to a lower-intensity regimen right out of the gate? It's a 45,000 white count. You've got cardiac issues or whatever it is that preclude intensive therapy, but you want to give them AML induction. You want to get them into remission. What do you do?

DiNardo: There are patients that we see who have that elevated white count of 30,000, 40,000, 50,000. And in that case, you cannot, and should not, just discharge the patient and say, "Come back in a week when I have all of this information." So we need to do something. Typically, what is done is Hydrea. Hydroxyurea is typically initiated and there's a bit of a lag time; 1-2 days after you start the Hydrea, you can see this nice white count blast reduction — really, a reduction in everything. It just kind of dampens the bone marrow from making anything. So it's a very nice temporizing measure. If you're using a lot of Hydrea, more than 2 g or 3 g a day for a prolonged period of time, you will see mucositis. So that's not something that you want to be continuing forever and ever, but it can be a very nice strategy to stabilize your patient while you're waiting for treatment information. Here at M.D. Anderson, we are very comfortable giving a one-time low dose of cytarabine to stabilize our patient. So it's like a flat dose: 1 g of cytarabine; you're not giving the anthracycline. Patients with underlying cardiomyopathies are not impacted and so that it works similarly. Same idea as Hydrea. It's not in any way curative, but it is helping to temporize while you're waiting for the actual initiation of your planned treatment.

Roboz: And how do we deal with the fact that docs are worried about using venetoclax in patients who have an elevated white blood cell count? I want to use a hypomethylator, either azacitidine or decitabine, and I want to give that in combination with venetoclax. I read a paper that said that it's okay for me to treat patients with high white blood cell counts because we're going to get them into remission. But wait a minute — venetoclax is associated with tumor lysis syndrome and with other problems. So do we say yes, I'm going to give them that regimen, but I'm only going to start it on day X when the white count is Y? Or how do we do that?

DiNardo: There is a concern for tumor lysis with venetoclax combination. So that is known. You see tumor lysis with CLL with bulky lymphoid diseases frequently; you see it much less frequently in AML. With myeloid malignancies in general, you will not have the same degree of tumor lysis syndrome. That being said, you can when you have very proliferative disease. And so yes, the recommendation is to get that white count down under 25,000 before you are starting venetoclax combinations. And I do think that that is a appropriate rule of thumb. We often are very concerned about the risk of tumor lysis. We still err on the side of admitting our patients to start venetoclax in combinations, and so we are monitoring chemistries every day during that initial 3- to 4-day venetoclax ramp-up period to identify any sort of unusual bump in the potassium or the creatinine function so that we can hold venetoclax that day, manage it, and then start the next day. That happens incredibly infrequently. But those patients that it does happen to tend to be patients with particularly proliferative disease.

Roboz: There are a couple of really, really important points that deserve extra attention in what you just said. One of them was, you used the scary leukemia word, "admission." This is something that I think is very important to address with respect to the discussion of the elderly or more frail patients. Historically, I would say that it was absolutely the case that we were so unconfident about many of those patients achieving a durable remission that there was actually a lot of attention paid to whether or not it was worth it to spend any significant time in the hospital for something that was really not going to be, in any likelihood, a durable response or really significant chance of remission. But now that we're doing better, we are actually very likely to admit the patient absolutely — not only the hyperleukocytosis patients, but just the elderlyness or the frailness or whatever it is that is driving you to not use intensive chemotherapy. I don't think that that should be the reason to keep people out of the hospital. Bringing people in the hospital and watching them very carefully, whether it's for tumor lysis monitoring or whether it's for general neutropenia management to prevent them from having to come in in the middle of the night, is fair game as part of the management strategy for older and more frail patients, just because they are likely to have trouble with the regimen, even though the regimen could theoretically be given as an outpatient. Do you agree?

DiNardo: I do. I think it very much depends on the patient and the family structure and how far they are from a medical center, because I don't think it is required that a patient needs to stay in the hospital for a venetoclax combination. However, it is far more convenient. That first week, they're coming every day for their HMA anyway. And you're monitoring the labs and making sure your patient doesn't have tumor lysis. And then for week 2, week 3, if they are at high risk of infection — this therapy does cause neutropenia — in that first month or so we'll see infections happen and we'll see patients need more transfusions than they did, probably before they started that first month as we're clearing everything out. So if you have a patient who lives very close and has super-supportive friends and family that are fine with taking the patient back and forth because they want to sleep in their own bed, then sure. It just takes a little bit more organization to make that happen.

Roboz: I want to get back to one other point that you mentioned. We're talking about older patients, we're talking about more frail patients. Hospital time should be in the discussion and in the calculus, and it's worth it because if they do well and they get in remission and they get out, that is actually now a worthwhile goal. I want to focus on that high white blood cell count patient for just one minute to address a question: If you are using cytarabine or Hydrea to bring down the counts, how important is the upfront synergy, for example, of the azacitidine or the decitabine in combination with the venetoclax?

DiNardo: We don't have a ton of great data to answer this question in a fully vetted manner. But the most important time in this cycle of your patient getting a venetoclax-with-HMA combination is that week that there is overlap. That is where you see the clear synergy of this therapy killing leukemia blasts, preclinical in vivo, in vitro. That's what we see. And so my preference is to always make sure that you are starting and giving those things together. There are also some data that suggest that giving Hydrea with azacitidine is not a great idea, that actually it can be a little bit antagonistic. And so I am very comfortable, I will say, holding and waiting to start those two agents, the HMA and the venetoclax together, because I think that's the optimal approach.

Roboz: And then for your patients who are coming in to the hospital: So this is an AML patient who lives 4000 miles outside of M.D. Anderson or they need to be in the hospital. Do you tend to, in a very neutropenic patient, just avoid the combination issues with azole antifungals and with venetoclax and just use an echinocandin, or use something else for your antifungal prophylaxis and avoid the azole altogether? Do you skip antifungal prophylaxis because there are no fungal organisms in Texas? Do you use azoles? Walk us through a patient who's on a venetoclax-based regimen; what am I doing with the antifungals?

DiNardo: I wish there were no fungal infections in Houston, but that is definitely not the case, so not doing antifungal prophylaxis for us here in the Deep South is not appropriate. We definitely are giving antifungal prophylaxis for our patients if they are one of those patients like we just talked about — that is, in the hospital for that first month. We may just have them on an echinocandin like caspofungin or anidulafungin and avoid the azole issue. But actually, that's the minority of patients. With most of our patients, we are fully comfortable using azoles with the appropriate dose reductions now. It's one of the most common questions I still get from providers: "I want to start venetoclax, but my patient's on voriconazole; what is the right dose and do you really, truly dose-reduce?" Because it was a little bit uncomfortable at first, right? Trying to say that 400 of venetoclax is the right dose, but yet you're adding voriconazole and that dose is now 100. So it's 25% and even less with posaconazole 50 mg or 70 mg. It sounded crazy and I was also hesitant to believe it. But we have seen with increasing time and years of doing this that those are the appropriate recommendations. And so I am very comfortable keeping my patient on an azole at the recommended dose reduction of venetoclax. So 100 mg with voriconazole 70 mg with posaconazole and then 200 mg of venetoclax with any of the moderates. So that would be isavuconazole, fluconazole, or even ciprofloxacin; some of the fluoroquinolones are moderate.

Roboz: That's a lot of management of older and more frail patients; whether they have a high white blood cell count or a low white blood cell count, we're trying to get them on an HMA-venetoclax–based regimen. We're trying to see whether we want them in or out of the hospital. At any point, we're dropping the white cells so that we can try to optimize the concomitant therapy up front, at least for some days of overlap between the HMA, plus the venetoclax. We are watching antifungal therapy, making sure to have dose reductions for the antifungals, and then we're going to watch and we're going to probably do a bone marrow biopsy 3 to 4 weeks in, either day 21 or day 28, or getting close to somewhere between there, to see if the patient is wiped out at that point. And if they're wiped out, maybe some growth factors support to help them recover. And if they're not, maybe launch into a second cycle. Would you say that's the cycle 1 and cycle 2 plan — wipe out the marrow on day 21, day 28, something like that, and then use some growth factors to get these patients back to a safe absolute neutrophil count?

DiNardo: Yes. I now talk to my patients and say, "Expect that your counts are going to drop very low. You will need frequent transfusions that first cycle. We will do a bone marrow biopsy at the end of the first cycle." So sometime in that last week from day 21 to day 28, just depending on how things fall into place, I'm often doing it day 23, day 24 or something along those lines. And then many of your patients will already be in a leukemia-free bone marrow state by the time you do that first bone marrow biopsy, so that's a really important one to do so. Then I'm going to stop the venetoclax and I'm going to wait a week or two for the counts to recover before I start the next cycle. And so that is probably the most critical management piece of how to do venetoclax combinations. I do frequently then have patients' ongoing cycles that are shorter than continuous 28-day venetoclax cycles. Very, very few of my patients are on continuous 28 days of venetoclax because almost all of my patients will have some degree of cytopenias that require a venetoclax duration reduction. So, 21 days, sometimes 14 days, and then I do have patients — probably about a third of my patients — who, even with those durations, with subsequent cycles, you're still seeing on-target neutropenia. Even though the hemoglobin is stable, the platelets are recovered, they're not needing transfusions; they're doing great. But you still see this neutropenia. And so in those patients, I will give granulocyte colony stimulating factor with cycles to continue to keep that neutrophil count above 500 goal, above 1000, of course.

Roboz: Sometimes we have more information than others on our AML patients, both in the newly diagnosed setting and in relapse. And every once in a while, we do have somebody again, either newly diagnosed or relapsed, who comes up with an IDH mutation — IDH1 or an IDH2, or, to be super-confusing, both mutations. And so what do we do now? Because we have all seen and heard data suggesting that venetoclax-based regimens are actually quite powerful for these patients. But there are also targeted inhibitors of the IDH1 and IDH2 which have some unique toxicity profiles. So how do we sort out which agent to pick for an IDH-mutated patient?

DiNardo: If I have a particularly frail or particularly elderly patient — you will have that adorable 92-year-old woman who still wants therapy and you're like, Oh my goodness, please have an IDH mutation. And sometimes they do. And so that is a situation where I would definitely prioritize just a single-agent IDH inhibitor over any HMA combination. When you're using the IDH inhibitors the safety profile to be aware of with the IDH2 inhibitor enasidenib, there's a pretty high risk of an indirect rise in bilirubin. It's an indirect hyperbilirubinemia very similar to unmasking a Gilbert syndrome, and that is typically not clinically significant. We often treat through that unless the bilirubin gets really high and the patient turns a little bit highlighter-yellow, and then we hold it and start a reduced dose. And then with ivosidenib, you can see QT prolongation in a minority of patients, which, as you know, rarely is clinically significant. But it is something that you will see if you're doing a lot of EKGs. And if we have patients on azoles and fluoroquinolones and antiemetics and a whole bunch of things that are QT prolonging, you will see elevated QT prolongation when you're adding ivosidenib to it. So be aware of that and monitor QT in that situation. The big thing with either IDH inhibitor is differentiation syndrome. About 12%-15% of the time you're using one of these agents, you will see very nonspecific symptoms. So if you're not thinking about differentiation syndrome, you will never see it because you're not going to realize that's what it is. You see pulmonary infiltrates, effusions edema, culture-negative neutropenic fevers, and these are very similar to what happens in our acute promyelocytic leukemia patients that we give ATRA or arsenic to. So just be aware; be thinking about this if you're seeing someone with effusions and infiltrates and then start them on steroids. One of the things we hear sometimes is that we shouldn't be thinking about venetoclax combinations in people who have a high white count. And we need to be cautious about tumor lysis in patients that we are starting venetoclax on who are more proliferative. But I think it's really important to make sure that people realize that treatment is still appropriate with the venetoclax combination in that setting; you just need to make sure that your patient has been appropriately "debulked" — meaning, use Hydrea, consider a 1 g flat dose of cytarabine, and do something that is going to bring that white count down to a more safe level, which in general we think is about 25,000 or so. Then when we are starting our HMA and venetoclax together, there's just not that same bulk of disease. So the risk of tumor lysis is less. We talk about the risk of tumor lysis a lot. We don't actually see it that often in AML, but those patients that you might see it in are going to be those patients with proliferative disease or particularly venetoclax-sensitive genomics like NPM1- or IDH2-mutated patients. And so those are the patients to be thinking of and being very cautious about as you initiate therapy.

Roboz: I want to thank you so much for a great discussion about how to try to safely get some of our older or more frail patients into remission using what I think much of the world is calling now the standard of care, which would be an HMA combined with venetoclax. We talked about some of those unique situations and toxicities to look out for in our patients. Thank you so much for joining me today.


Acute Myeloid Leukemia


Mylotarg (gemtuzumab ozogamicin)

A Phase 3 Trial of Azacitidine Versus a Semi-Intensive Fludarabine and Cytarabine Schedule in Older Patients With Untreated Acute Myeloid Leukemia

Azacitidine and Venetoclax in Previously Untreated Acute Myeloid Leukemia

Ferrara Criteria






Hydroxycarbamide in Combination With Azacitidine or Decitabine Is Antagonistic on DNA Methylation Inhibition

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.