Precision Medicine in Acute Myeloid Leukemia Treatment - Episode 2

Best Practices for NGS Testing in AML


Harry Erba, MD, PhD: There are a lot of different mutations that have been discovered in people with acute myeloid leukemia [AML], and cytogenetic changes we’ve known about. I think what can be a challenge is taking this ever-shifting amount of data and then applying it to the next patient you see. I think most of us have set up some kind of algorithm for how we evaluate patients with acute myeloid leukemia. For example, at Duke University, we look at things that we need immediately for treatment planning and then other things where we might be able to wait. For example, at the time of diagnosis, we follow the 2017 ELN [European LeukemiaNet] guidelines and get an aspirate and a biopsy. Now, keep in mind, if the biopsy is a dry tap and you have 100,000 white cells and they’re all blasts, you don’t really need to get a good bone marrow aspirate. You can do all the tests I’m about to talk about on the blood.

You need a hematopathologist looking at the slides, you need flow cytometry for immunophenotyping. One reason is just for making the diagnosis. The second reason might be, and we’ll talk about it later, identifying a leukemia-associated immunophenotype that you might use later on for MRD [minimal residual disease] assessments. Maybe I’ll turn to Dan later for that discussion. You also want to get cytogenetics and molecular tests. But the question is what do you need right at the time of diagnosis?

In my opinion, what you would really like to know is does the patient have core binding factor translocations, 15;17 of APL [acute promyleocytic leukemia], or complex changes, [chromosome] 5 and 7 abnormalities, that make you think of myelodysplastic-related cytogenetic changes? Cytogenetics may take some time, so we have developed a rapid turnaround for FISH [fluorescence in situ hybridization] analysis that can give us much of that data. So we’ll do FISH, so that we get the data quickly, within a couple of days, to allow us to do treatment planning, and we could talk more about what sets of patients we might use different therapies.

The molecular markers that you need at the time of diagnosis are really just FLT3 and IDH1/2. And those are best done by PCR [polymerase chain reaction] because at our institution, I’ll ask you about yours, the turnaround time is quicker. But I agree with you, Eunice, we also have to get a next-generation sequencing [NGS] panel. The reason for that is these markers have become prognostically important for planning future therapies, such as does the patient need a transplant. But you might not need them right away. There’s definitely argument, considerations of how many different markers you need in those panels. But there is consensus about some of the ones that are important. In the ELN guidelines, RUNX1, TP53, and ASXL1 have been associated with a poor prognosis, so those are very important to get. If you were to do those by PCR, you’ll quickly get to a point where it’s actually a cost savings to do them by next-generation sequencing. So you can argue about how big your panel is, but it should have at least the markers that are recommended by NCCN [National Comprehensive Cancer Network] and ELN. And of course, the other reason for getting PCR at baseline, a FLT3 and IDH1/2, is that at least for FLT3, you get a quicker turnaround time and NGS panels may sometimes miss large internal tandem duplications. So that’s what we do at the time of diagnosis.

I guess my final point on this would be in the vast majority of patients with AML, you can wait to get this information. Of course, you have to move your pathologist along and tell them what’s important in real time, but you can get this done quickly. I’m going to now turn back to Eunice and say, well, do you agree with this and how quickly do you get your results?

Eunice Wang, MD: I completely agree with that, Harry. At our institute we do use the single analyte PCR and we are really emphasizing the turnaround time for FLT3 and IDH1/2. We both know that there are targeted therapies commercially available now for those. In fact, there are some therapies that we’re now testing both in the upfront setting. So having those results back before day 8 or early on in the treatment process can be essential for clinical trial enrollment, as well as for initiating those FLT3 inhibitors in frontline therapy, which I’ll talk about a little bit later.

At our institute, we have good availability of a molecular diagnostics laboratory, which is just down the street, so we get the FLT3, IDH1/2 turned around within 2 or 3 days. They run the test Monday, Wednesday, and Friday for us. Once you move to a commercial lab, regardless of which commercial lab you use, you’re going to have additional delays because you have to send the sample out. So that is by definition, an outside commercial lab is going to be adding anywhere from 24 to 72 hours to your turnaround time, depending on where that weekend falls and you collect the sample.

For that reason moving forward, it would be best to not have that additional 2- to 3-day turnaround because then that makes the results really not available to the clinician, we’re talking realistically, until 7 to 10 days. In my perspective, that can be a little bit too long when you’re trying to decide on therapy, whether you’re going to add a FLT3 inhibitor but also what baseline chemotherapy you’re going to give because some chemotherapy regimens you can add and some you can’t. And if you’ve already started a cytotoxic regimen, it might not be possible 7 or 10 days later to add that beneficial drug on top of it. So that’s my impression of the importance of these tests. I think there’s some question about which tests, because of cost issues, insurance coverage, and availability, need to be done at different times in the treatment course.

Alexander E. Perl, MD, MS: I think that’s really important because nowadays it isn’t one-size-fits-all in terms of what regimen you give frontline. There are patients who would really benefit from 7+3 [cytarabine/daunorubicin] alone or other patients who might benefit by 7+3 [cytarabine/daunorubicin] with gemtuzumab, and other patients who might benefit by 7+3 [cytarabine/daunorubicin] plus midostaurin or even CPX-351 instead of 7+3 [cytarabine/daunorubicin]. Now we need to really figure out what’s the best regimen for the patient, and often these genetic tests are driving that choice. So you need enough information to make a regimen choice. You don’t want to be part way through that and say, “Oh, wait a second, I want to add this but I’ve already committed.” You really need to have enough information to commit to a regimen. At the same time, you can’t wait weeks and weeks and weeks before you commit to anything. Patients need treatment.

So I generally agree with exactly what you said, that in many patients we have a few days to maybe a week to get information if they’re clinically stable. But in some patients who really have aggressive disease, we certainly want to get therapy started that we can add another agent to if it makes sense. That’s been our approach, certainly for FLT3 mutations, where we have about a week to get that information. But we don’t want to be going longer than a week with any test that’s turning around. So I agree with you, I think it’s really important to get PCR testing back in that first week of therapy because we can start 7+3 [cytarabine/daunorubicin] and add midostaurin to it and feel comfortable that we’re giving the right therapy starting at day 8.

Jorge E. Cortes, MD: I would add to that, I think it’s important to emphasize that this waiting, it’s not just that we think that it’s OK to wait. It’s important nowadays certainly because of the treatment options. But it’s been documented. There have been studies that have shown that waiting those few days because of the additional information that you gain, and even more so when it impacts your treatment approach, is it doesn’t adversely impact the outcome. It’s not just that we think that it doesn’t impact. So it’s important to educate our patients to say it’s fine, we will monitor you. Transfuse them and supportive care and all of that, of course. But the wait is I think justified and actually more and more critical because of the information you gain, except for the few exceptions of the very proliferative patients.

A couple of things I think would be valuable to bring up. One of the issues is that the reports nowadays still vary greatly from lab to lab, so that sometimes it’s a little hard to understand what the lab is telling you. We all become familiar with our own labs and we know how to read them, but sometimes you get these reports from outside of patients referred to you, and they’re difficult to read. And some of the elements that we’re valuing more is we need to recognize that there are sometimes variants that are not the mutations that we recognize as having prognostic or therapeutic implications, and it’s important to recognize them.

Also the importance of the variant allele frequency for some of these mutations that may have particularly prognostic implications and maybe even some therapeutic implications in some instances. I think more and more it’s going to become critical to make these reports uniform across different labs. There has to be some entity that regulates that, so that we all get the same information in the same format and then understand it well, so that we know what that means and when it’s a real mutation, when it’s a variant, and then what to do with this variant allele frequency. I think when we talk about some of the therapeutic implications, we can discuss that further. But I think it is important to recognize that in addition to just the presence of a mutation, what this means.

Harry Erba, MD, PhD: You all bring up very important points here. I think understanding these reports can be challenging. The place where the variant allelic frequency may be most important is in FLT3-ITD [internal tandem duplication], where the European LeukemiaNet now actually uses that to help stratify risk in AML patients. It hasn’t quite caught on. There are some challenges in standardizing those assays the way they’re done, but clearly it does seem to carry some importance. So again, another good reason for doing PCR analysis of FLT3-ITDs at the time of diagnosis.

Transcript Edited for Clarity