Treatment Approaches in Acute Myeloid Leukemia

David Henry, MD; James M. Foran, MD


April 22, 2020

This interview was originally published as part of MDEdge's Blood & Cancer Podcast series. In this episode, podcast host David Henry, MD, of Pennsylvania Hospital in Philadelphia, spoke with James M. Foran, MD, chair of the Acute Leukemia & Myeloid Neoplasm Disease Group at the Mayo Clinic Cancer Center in Jacksonville, Florida, about his preferred algorithm for treating acute myeloid leukemia (AML).

David Henry, MD: I attended the American Society of Hematology (ASH) Highlights meeting this January in New York City, where you led a popular session discussing your approach to treating AML. I was happy to see that we've moved beyond the era of the 7 + 3 regimen, with 7 days of cytarabine and 3 days of an anthracycline. Let's discuss how things have changed by posing a hypothetical patient. I'm presented with a 50-year-old man with comorbid disease, whose white blood cell count is 30,000/µL. His smear reveals 90% blast and a couple Auer rods. He has a hemoglobin of 10 g/dL, platelets of 20 x 109/L. I refer him to you. What happens next?

James M. Foran, MD: When I trained in the 1990s, we would have said that this patient needs to start treatment within 24-48 hours because it's a medical emergency. It's still an urgent situation, especially with a high white cell count. But we've learned from experience—and this is recognized in the National Comprehensive Cancer Network guidelines—that if it's appropriate for this patient to wait, and for most it is, it's better to first get some results back.

You want to ask, is this therapy-related AML (t-AML) or AML with myelodysplasia-related changes (AML-MRC)? Do they have an FLT3 internal tandem duplication (FLT3-ITD) mutation? You also should do a workup for comorbidities and cytogenetics.

A diagnosis should be available in 24-48 hours. Fluorescence in situ hybridization (FISH) studies should come back in 2-3 days, cytogenetics within a week with a little bit of pressure on your lab, and a FLT3-ITD test in 3 working days.

Generally, we recommend launching next-generation sequencing to look for mutations, but particularly for FLT3. There are survival advantages to adding the FLT3 inhibitor midostaurin in the first-line setting if somebody has an FLT3 mutation. It looks like there is a survival advantage for using liposomal formulations of daunorubicin-cytarabine, CPX-351, if they have t-AML or AML-MRC in their cytogenetics.

If they have a core binding factor AML (CBF-AML), there is pretty compelling meta-analysis data showing a survival advantage to adding gemtuzumab.

So 7+3 is still in the mix for all of those scenarios, except for CPX-351, which is its own formulation. But you want to be rational about picking whether they're appropriate for other treatments. It's a question of what are the best additives to try to get you anywhere from a 7% to a 20% improvement in overall survival.

Mutations and Maintenance

Henry: The liposomal combination formulation CPX-351 seems so attractive, but its use depends on these other features, correct?

Foran: The pivotal study was restricted to patients 60-75 years of age with newly diagnosed secondary AML. That's the group where we really believe there is a higher complete remission rate, a survival advantage. Honestly, some of that survival advantage came after subsequent allogeneic transplant, so you want to use it in a patient where you think that might also be an option. The safety profile was similar to conventional 7+3. There are now studies to see if CPX-351 has less cardiotoxicity in the pediatric setting. You can't say it's wrong to give, but you want to use it on-label whenever possible. I'm going to get behind the US Food and Drug Administration (FDA) on that one. It's competing a little bit with gemtuzumab as an additive, or midostaurin if you have an FLT3 mutation. These are also expensive drugs. You want to take the 3 or 5 days to get the details right, as long as that's safe, because it turns out it matters what you start with.

Henry: What about patients with isocitrate dehydrogenase 1 (IDH1) or 2 (IDH2) mutations?

Foran: We're getting those results back more often, and we have approved inhibitors for IDH1 and IDH2. The FDA approved the IDH1 inhibitor ivosidenib as a single agent for frail elderly with AML in the first-line setting. It has some activity there. The question is, how do you incorporate that?

At ASH 2018, there was a study where IDH1 or IDH2 inhibitors, respectively, were added to 7+3 if you had an IDH1 or IDH2 mutation in induction, consolidation, and maintenance. It showed very good complete remission rates. It was quite favorable in these patients who tended to be older, but it wasn't a comparative study.

We know you can do it; we just don't know that it's better yet. It's still an open question as to whether you target that by adding the IDH1 or IDH2 inhibitor or just treating generically for AML in that setting.

Henry: In your presentation from January, you discussed maintenance therapy with a hypomethylating agent. Could you elaborate on that?

Foran: That was based on a late-breaking abstract at ASH, which was very compelling and a little surprising, actually. We were offered the chance to participate in that study and didn't because I was a bit dubious, and I was wrong.

Investigators randomized patients in remission following induction therapy with a 7+3 based regimen to either placebo or an oral formulation of azacitidine, CC-486, given 2 weeks on and 2 weeks off indefinitely. Around 80% of patients received consolidation therapy. There was some gastrointestinal and hematologic toxicity associated with CC-486, and they had to lower the dose to get it in, but it was associated with a significant improvement in remission duration and a significant improvement in overall survival.

The drug's not yet available, but I think it's going to change the landscape for us once it gets regulatory approval.

What to Do in Older Patients With Secondary AML

Henry: Let's move to another hypothetical patient. This is an 84-year-old woman who appears to have gotten to AML status through myelodysplastic syndrome (MDS). She has some hypertension, atrial fibrillation, and has been on and off anticoagulation. She has a little anemia, a lowish platelet count, and a white blood cell count of 4000/µL, and a bone marrow shows 60% blast. What does your algorithm suggest in this older, not-so-fit patient?

Foran: That's a more common scenario. As you know, the median age for AML in the Western hemisphere is about 70 or 72 years. It's a little more complicated landscape; however, as a general rule, we think it's worth treating as long as the patient has an Eastern Cooperative Oncology Group performance status of 2 or better or a Karnofsky Performance Score of 50, 60, or better.

In the United States, the standard has been to give azacitidine or decitabine, and the complete remission rates are in the range of 20%, give or take.

The FDA approved venetoclax almost a year and a half ago based on a large phase 2 study showing a remarkable improvement in complete remission rate, close to 70%. The addition of venetoclax seemed to have that impact regardless of your mutation status, although it was actually even better if you had an IDH mutation. The problem is that venetoclax makes treatment harder. You get a lot more myelosuppression, a lot more supportive care. It's hard to do the dose adjustments. The remissions happen in 1 month versus 4 months with azacitidine. But it's becoming a rapid standard.

At the same time that venetoclax was approved, a small randomized phase 2 study looked at low-dose ara-C with an oral Hedgehog inhibitor called glasdegib. The complete remission rates weren't so impressive, but it did show a survival advantage.

Between these two treatments, it looks like combination therapy is the new standard. At the moment, most of us give a venetoclax-based combination, although it's certainly not wrong to give glasdegib, and frankly the data might even be stronger for it.

One of the compelling abstracts from the most recent ASH was a study of azacitidine with or without the IDH2 inhibitor, enasidenib, in patients with an IDH2 mutation. The combination led to a much higher complete remission rate than azacitidine alone, but it did not translate into a survival advantage. It kind of left us all scratching our heads a little bit. We've all adopted venetoclax because of the complete remission rate being so much higher. Now with enasidenib for those with IDH2 mutations, you see the same thing: high complete remission rates without the survival advantage. We think it's because patients are now getting crossed over to other therapies, as they did on that enasidenib study.

That's a different landscape for older patients with AML. We're not just giving one line of treatment and then saying we're done. That's going to make it cloudier for us to see straightforward survival advantages. Patients have more options—not as many as we'd like, but more than we had. It's going to take a little bit of reading between the lines to really know where the benefit is. We might even end up going to an event-free survival type of outcome in the future in that population.

Quicker Testing, More Informed Choices

Henry: How long does the laboratory testing take if you're not at a university?

Foran: It's a great question. The workup can take 1 or 2 weeks. Maybe it's a patient you suspected of having MDS and you're transfusing them and giving them supportive care. Or they're an outpatient coming back 2 weeks later, and their bone marrow biopsy returns a little surprisingly, showing AML.

A lot of the reference labs are now doing next-generation sequencing to look for mutations. That's being packaged with normal pathology vendors, even outside of university hospitals or academic centers. We're blessed at Mayo Clinic, where we have all of our own infrastructure, and it's outstanding. In the community, that takes a little longer, but I think you should expect that within 2 weeks now.

Findings were presented at ASH from the Beat AML study, which was sponsored by the Leukemia Lymphoma Society. It shows that it's feasible to get rapid sequencing and assign a treatment within 7 days based on the mutation.

I think we should all be raising our expectations to try to get those results certainly within 2 weeks. The problem is that it means extra phone calls and following up on things, which doesn't yet fit into the easy flow of practice. That's where we have to learn how to adapt it, because those patients are now outpatients. It's going to require different outpatient infrastructure to support them. We all have to be prepared for how that's going to evolve.

Henry: I couldn't agree more. It takes a lot of time and effort because the patients don't always call you. If they have a fever, they may think that's normal and not reach out to you.

To wrap up the scenario with this hypothetical older patient, how long are we extending her treatment?

Foran: That's the problem; so far, we haven't learned when you can stop it. So there are treatment delays and dose reductions, especially with the risk of cumulative myelosuppression. We need better guidelines for how and when to reduce the dose of venetoclax, to increase the frequency between cycles, sometimes reduced to 5 days of azacitidine. It's better if we fly by instruments than by instinct on this one, and we need those instruments.

We're all still waiting for better guidelines from the randomized studies as to how the dose reduction should happen, because you have to continue it or you lose the benefit, and yet not all patients can stay right on schedule.

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.
Post as: