Harry Erba, MD, PhD: Acute myeloid leukemia [AML] is a disease associated with older age: a group that rarely achieves long-term leukemia-free remissions with traditional chemotherapy. In addition, older adults with AML who are unfit for standard chemotherapy have poor outcomes. Fortunately, we’ve had several recent FDA approvals of novel agents for AML, providing opportunities to individualize therapy for each patient.

In this OncLive® Peer Exchange® discussion, we will discuss current and emerging strategies for improving remission rates for patients with AML.

I am Dr Harry Erba, a professor in the Department of Medicine at the Duke University School of Medicine and a member of the Duke Cancer Institute in Durham, North Carolina.

Today I am joined by Dr Naval Daver, an associate professor in the Department of Leukemia at The University of Texas MD Anderson Cancer Center in Houston, Texas; Dr Rami Komrokji, a senior member at the H. Lee Moffitt Cancer Center [in Tampa, Florida], specializing in malignant hematology; Dr Mark Levis, a professor of oncology and a program leader of the Hematologic Malignancies and Bone Marrow Transplant Program at Johns Hopkins Sidney Kimmel Comprehensive Cancer Center in Baltimore, Maryland; and Dr Dan Pollyea, the clinical director of leukemia services and an associate professor of medicine at the University of Colorado Cancer Center in Aurora, Colorado.

Thank you. Let’s begin. Mark, I’m going to have you take us through some of your considerations of what kind of molecular testing needs to be done at the time of diagnosis. What are you doing at your institution?

Mark Levis, MD, PhD: We are getting over the panic that came over us a few years ago when we introduced next-generation sequencing and targeted agents. When a new leukemia patient would show up on the wards, the other attendants or I would freeze and say, “OK, I’ve got to do this carefully now.” We have evolved a sequential testing strategy based on what we need to know sooner and what we can wait to know a bit later. We’ll haphazardly say all of them, but the fact is that there is a sequence. The first thing I want to know, after having ruled out APL [acute promyelocytic leukemia], is if the patient has a core binding factor. The quickest test is the FISH [fluorescence in situ hybridization] test. You can get RT-PCR [reverse transcription-polymerase chain reaction] and so forth, but in fact, we use the overnight FISH test to rule out core binding factor because we might introduce gemtuzumab [ozogamicin] into our induction strategy. The next 2 things we want to know require a next-generation-based sequencing approach, which is TP53.

I don’t want to start intensive induction therapy on a TP53-mutant patient, but I’m going to get a clue if I sent AML FISH because I can get a chromosome 5 or 7 deletion back within 48 hours, and that’s going to give me a clue that this patient may be TP53 mutated. If they have a high white blood cell count, I might be forced to start treatment, although we have data that we can delay. I want to know about FLT3, of course. Do they have a FLT3-TKD or a FLT3-ITD mutation? Then I’m going to go on a normal paradigm: starting 7 and 3, with the thought of adding midostaurin on day 8. We have a rapid turnaround next-generation-sequencing panel that we can get back in time to get those kinds of results. The FLT3 is a PCR [polymerase chain reaction] assay, and we can get that within 2 to 3 days typically. We’re not in that much of a hurry, so that’s a more rapid turnaround.

It’s a series of things. Certainly, I want conventional karyotyping, but that’s going to come back a week or 2 later and isn’t necessarily going to impact induction. I’m going to want to know FLT3 and core binding factor. I’m going to want to know TP53 as well as the other broader mutations. I want to know about IDH, particularly if it’s an older patient. That might influence starting out therapy with azacitidine combined with venetoclax. I might even consider azacitidine combined with an IDH inhibitor or IDH inhibitor as a monotherapy. We will get next-generation sequencing back with the rapid panel within the week, but it still goes in a sequence.

Harry Erba, MD, PhD: What I heard you say is that you order everything, but what you need right away is an AML FISH panel looking for core binding factor translocations and chromosome 5 or 7 changes. Those FISH panels have other probes; they might find other deletions or additions that might help you in looking for that conflict stereotype, which will have a TP53 mutation 75% of the time. We have an approved drug for de novo AML with an MDS [myelodysplastic syndrome] like karyotype, like daunorubicin cytarabine, GO [gemtuzumab ozogamicin] for the core binding factors, and a PCR assay for FLT3 and IDH because you get them back quickly. The next-generation panel provides important prognostic information. That can wait, and the karyotype can wait. But you order it.

Mark Levis, MD, PhD: It’s important that if there is any change in disease status if the treatment didn’t work, we reassess. That usually involves a next-generation-sequencing approach to reassessing, and they’re not going to come up suddenly with core binding factors out of the blue. For a targetable mutation, a FLT3 might come, and a FLT3 might go. Whenever there’s a change in disease status—the patient was refractory, the patient is progressing, the patient has now relapsed—I want to reassess. The emphasis tends to shift more toward next-generation sequencing and a FLT3.

Rami, should next-generation-sequencing panels be part of standard of care in AML?

Rami Komrokji, MD: I agree. At the beginning they are very informative, but we do the same. Whenever there’s a change, we check because that gives you a wonderful opportunity, whether it’s in patients in current clinical trials or using some of the agents we have now. Like in FLT3, sometimes you see escape or new mutations coming up. There is value in checking them. We’ll also touch base at some point about the value of achieving MRD [minimal residual disease] status with a treatment, and that could be partly evaluated by next-generation sequencing as well.

Mark Levis, MD, PhD: There’s a problem with next-generation sequencing making that statement. What a lot of oncologists don’t realize is that diagnostic companies are bundling everything together. You get your final report 2 or 3 weeks later, but you didn’t know about that core binding factor, and you didn’t know about the FLT3, so you haven’t started treatment in time.

You didn’t consider the fact that there was an IDH mutation; maybe that patient is on the border. Are they going to be a good responder to AZA-VEN [azacitidine-venetoclax]? Choosing what you want [and] when and this turnaround time is still a problem.

Transcript Edited for Clarity