Acute Myeloid Leukemia: The Path to Personalized Medicine - Episode 10

Significance of FLT3 Mutations in AML


Harry Erba, MD, PhD: We’re going to move on to talk about the FLT3 mutations as a target for therapy in acute myeloid leukemia [AML]. And what I’d like to start with is, Raajit, some comments about the prognostic significance of FLT3 and how and when we should be testing for those mutations.

Raajit K. Rampal, MD, PhD: So certainly FLT3 mutations are approximately a third of the AML cases that we see, and we know them to be among the most important prognostically significant mutations in terms of how patients do long term—their rates of relapse, their rates of survival. And so without a doubt, a lot of effort has been put into targeting this. Now the question is, when do you test for it? I think at this point, given the data from the RATIFY trial, it’s our expectation that this is among the genes that should be tested as a standard of care when a patient presents with newly diagnosed leukemia. We would also add to that things like IDH1 and IDH2 mutations because the FLT3 mutations are clinically actionable.

The turnaround time becomes an issue, pragmatically speaking, when we think about this. And we know from the RATIFY trial that we don’t need to start the midostaurin until day 8, and so that gives us more time for turnaround. I think that, in general, a lot of mutation testing has seen a marked reduction in the turnaround time. But I certainly know this is an issue in a lot of community practices, possibly less so in academic practices, where we can turn around these assays in several days and have an answer as to whether or not we have to modify therapy.

I think another question that arises from this, then, is do you test it in patients who have relapsed disease who perhaps didn’t initially present with FLT3-mutant disease? Is that an important thing to do? And I would certainly argue it is, because there’s more than enough literature to support that patients can relapse with a FLT3 clone, or that those who presented initially with a FLT3 clone relapse with a non-FLT3-driven clone. So I think the issues of testing are very important, but obviously there is variance across institutions, particularly when we talk about academic versus in the community.

Harry Erba, MD, PhD: You know the problem with next-generation sequencing, though, for the FLT3 ITD [internal tandem duplication] is that based on the technology and the length of these ITDs, the next-generation panel may miss it or underestimate it. So not only do they take longer, but they may actually miss it. So doing a PCR [polymerase chain reaction]—based assay is really important when thinking about prognostication and also picking a therapy. But Mark, there was an important abstract presented about the position of the internal tandem duplication, which you could get only by next-generation sequencing and how that may affect outcomes. Did you want to say a word about that?

Mark J. Levis, MD, PhD: Yes. I have been aware of this data for years and pointed out it’s very simple. So the issue is, where does this insertion mutation occur within the FLT3 gene? Is it in the juxta-membrane domain, or is it in the kinase domain? Well the kinase domain, which is further on down, it’s an artifact. If you have a longer insertion, it starts in the kinase domain. That was pointed out as a bad thing a long time ago, and I think their data are real, but you don’t need to know exactly where the insertion is. Just look at the length of the insertion itself. Usually the report will tell you, but not always.

Harry Erba, MD, PhD: OK, I’m going to ask something controversial now. Sorry to do this, but in that report, they showed, based on this position, that there was not a benefit in the subset analysis of the midostaurin. So would you advocate using position or length of the ITD to choose patients for midostaurin as opposed to some other therapy?

Mark J. Levis, MD, PhD: The problem is the length is a continuous variable. And while in the single-agent studies we do see responses with so-called long ITDs, or kinase domain insertions, we don’t have enough numbers to know whether the response is lower. I would speculate the response is lower. But that doesn’t mean you still shouldn’t use it.

Harry Erba, MD, PhD: And the second-generation drugs are more potent and specific and still have the same problem, I understand.

Mark J. Levis, MD, PhD: I believe they do, yes. Those long ITDs are bad actors.

Harry Erba, MD, PhD: So we do what we can do.

Transcript Edited for Clarity