Harry Erba, MD, PhD: Let me see if I can synthesize this because I’m still confused, to tell you the truth, about where we’ve ended up here. I agree it’s a moving target. I think there are some important issues that we have to deal with. One of them was the mention of, “Well, if I know the patient had a normal karyotype, then I might do additional testing.” It brings up the point that some pathology labs are doing reflexive testing. I’m going to ask you point blank right now, based on what we’re going to talk about later, should FLT3 testing be done as soon as the patient is diagnosed? Or do you wait for a normal karyotype to be shown? What’s the answer?

Alexander E. Perl, MD: I do it on everyone.

Harry Erba, MD, PhD: Right up front?

Alexander E. Perl, MD: Right up front.

Harry Erba, MD, PhD: Is there a reason for that?

Alexander E. Perl, MD: Yes, because if they’re going to get induction chemotherapy, I’m going to want to add a FLT3 inhibitor to their frontline therapy.

Eunice Wang, MD: Even if they’re not getting induction chemotherapy, right?

Alexander E. Perl, MD: If they’re not getting induction therapy, I want to know, because I might use that in their frontline or their relapsed therapy. I want to know that as quickly as I can. The benefit is if the mutation is present, it gives us a target to go after. Even on the RATIFY study, which is a randomized study, a significant number of the patients who were in the midostaurin arm did not have a normal karyotype, and we know there’s a benefit.

Harry Erba, MD, PhD: Again, I want to bring this back to very concrete recommendations for people practicing today. I agree with what you just said. We need to get that information about FLT3, tyrosine kinase domain mutations, in internal tandem duplications because we now have a drug approved for that population of patients and you have to start it quickly. If you don’t test, you won’t get that information. If you do next-generation sequencing, you may get it as part of a panel, but it might be too late because these panels take too long.

Alexander E. Perl, MD: Yes.

Eunice Wang, MD: You have to remember that when we are starting a patient on therapy, we do have 7 days of 7 and 3 chemotherapy. But we’re supposed to be starting the midostaurin on day 8. So, if you don’t send the sample on day 1 or day 2, you may not get it back by the time you start the midostaurin on day 8. There is an urgency. So yes, I agree, you should do FLT3 on everybody up front, and that I think is going to change the immediate treatment of that patient. With some of these other mutations, I would argue they’re going to change the longer term outcome with a prognostic outcome, or predict for resistance. What do you think, Jorge?

Jorge E. Cortes, MD: Absolutely, yes, we also do it on everybody. I think it has practical implications. We obviously have to acknowledge that if the patient has other karyotypes, complex karyotypes, it’s going to be much less frequently mutated.

Eunice Wang, MD: But don’t you think, until we get these agents, they could wait a week or 2 to get the molecular data?

Alexander E. Perl, MD: Oh, sure.

Harry Erba, MD, PhD: Yes.

Jorge E. Cortes, MD: Although for many of these, in terms of some other practical elements—at least for our lab—it’s less costly to do a panel than individual mutations. So, we send a panel for that reason.

Eunice Wang, MD: But for people who don’t have that luxury…

Jorge E. Cortes, MD: Absolutely. Then, you have to send the critical ones.

Harry Erba, MD, PhD: Again, the other issue is for the other mutations. Nucleoplasmin and CEBP alpha are clearly important in normal karyotype AMLs as being prognostically important. We agree that those should be tested. ASH and CAP guidelines also strongly recommend testing many of the others that you have mentioned: p53, IDH1 and IDH2. But I’m going to make a statement, and tell me if this is correct or not. Other than prognostic information that you get from those, is there any, at this point, clear evidence that one treatment algorithm is better than another?

Alexander E. Perl, MD: I think we have those data in older patients with secondary AML, because we have a randomized study saying that patients who got Vyxeos had a better survival than those who got plain 7 and 3 chemotherapy. Vyxeos here is the liposomal encapsulation of both daunorubicin and Ara-C [cytarabine] in one molecule. So in that setting, we have evidence that one induction strategy is better than another. And to define who benefits from that, we need to know a history of prior MDS [myelodysplastic syndromes], prior chemotherapy, or genetic features that would define that.

Eunice Wang, MD: But, I would argue that molecular features do not make a difference in terms of selecting liposomal 7 and 3.

Alexander E. Perl, MD: But, you could look at a P53 mutation and say that in and of itself, though it wasn’t a defining criterion in the study.

Eunice Wang, MD: That was not a defining criterion, right?

Harry Erba, MD, PhD: Although I agree with you, I do wait for some of this information, especially in older patients whom I’m getting more and more concerned about giving intensive therapy to if you find a p53 mutation. It leads me towards thinking about a 10-day regimen of decitabine as was published by Washington University.

Alexander E. Perl, MD: Yes.

Harry Erba, MD, PhD: Do you know if that’s better than intensive chemotherapy?

Eunice Wang, MD: I don’t, but I also argue that for a lot of my patients who are older—and you can disagree with me—I’m not in a big rush. If a younger person comes in and their white blood cell count is 400,000, they have hyperleukocytosis, I’m going to start 7 and 3 chemotherapy. For somebody who’s maybe not a candidate for 7 and 3 chemotherapy, I maybe have a little extra time to wait for some of those results before jumping in, right? What do you think, Jorge?

Jorge E. Cortes, MD: That is true, but you know the problem here, going back to the point that Sasha was trying to make, is this subset of patients with secondary AML that’s defined by cytogenetic abnormalities associated with MDS. The cytogenetic results can take 3 or 4 weeks. I’m in agreement with waiting, but for 3 or 4 weeks?

Alexander E. Perl, MD: Not 3 or 4 weeks.

Eunice Wang, MD: No, no.

Jorge E. Cortes, MD: That’s a problem if that’s going to define your treatment. If it takes that long, then it starts to become impractical.

Transcript Edited for Clarity