An Introduction to Biomarker Testing in Acute Myeloid Leukemia

Dr. Naval G. Daver: So in your practice, which mutations are you requesting? Do you get all of them together? Do you get a rapid panel first, then the others later? And then are you doing this all as NGS or do you sometimes for mutations specifically request PCR initially? So when you see a patient, how do these data come to you? And then importantly, what's the turnaround time for these? And which ones can you wait for before starting?

Dr. Eunice S. Wang: Well, that's an excellent question, because I think now in the era of precision targeted therapy for AML, you need to have the data in order to select the therapies. So many panels have been recommending a comprehensive mutational profiling of patients at the time of diagnosis. For example, the ELN 2022 has recently put out new prognostic classification systems that are largely based on cytogenetic and mutational features and incorporate many, many more mutations than we think are probably on most of our panels. For example, they incorporate a whole number of what they call secondary mutations, SRSF2, U2AF, etc., which previously many of us were not looking at because they seem to be just for academic purposes, but now seem to prognosticate and be associated with an adverse outcome. So there are now a number of these mutations which are important for therapeutic purposes. What we have been focusing on our center initially was a rapid heme panel that included NMP1, FLT3, ITD, FLT3-TKD mutations, IDH1, IDH2, also included P53 and KMT2A or MLL. We've now expanded that to a larger panel, including many of these secondary mutations and other mutation types, which probably now incorporates about 75 mutations. So our turnaround time in-house is an admirable three to five days. But many centers still use PCR for FLI3-ITD because patients who have very large internal gen duplications can be missed on a next-gen sequencing panel. So many of us will still do PCR for FLT3-ITD of note, the FLT3- ITD allelic ratio, which was a prognostic category in and of itself for the last European leukemia. Net classification is no longer considered a prognostic factor. So anybody now with FLT3-ITD is considered intermediate risk because of the advances we've made with incorporation of our FLT3 inhibitors. So I think even ideally, the many panels have recommended NCCN, ELN, a turnaround time somewhere in the range of five to seven days. We have three to five days, but we are really encouraging people to use that to better tailor and select therapy for patients.

Dr. Naval G. Daver: That's very impressive. So you get the whole 75 gene panel in three to five days?

Dr. Eunice S. Wang: Yes.

Dr. Naval G. Daver: That is really good. So just- at MD Anderson, we do a rapid panel that covers about 10 genes, the ones you mentioned, NPM1, FLT3, TP53, IDH1 and 2, MLL, JAK2, some of those for which targeted therapies are there, those usually come in about three to four days. And then the broader panel, 81 gene panel takes about seven days. So I think that is great. But as you said, I think for most centers and especially community and outside, this is not the reality. I still see a lot of patients coming to us and they've come to me two, three weeks later after their diagnosis and the molecular data is still not available to the referring community doctor or physician. So I think this is one of the big challenges that we are seeing a widening in the gap in the management of AML where the larger academic centers are moving more and more rapidly towards getting the data. But in the community, I think in the last four or five years, I don't know if that timeline has shortened at all. So I think this is one of the big problems. Any thoughts on that?

Dr. Margaret T. Kasner: So I think one of the things that I would talk about frequently is that there was a paper, the abstract was originally published, I think at ASH in 2019. And then subsequently the paper was published that really said, outcomes in AML patients remain equal no matter how long you wait and are better as long as you wait, as long as the patient is stable enough to do so. And we all know that to be the case. And so I think if FLT3-ITD, because your patients are going to present with very high white counts and very sick, those patients need to come to academic centers where it's going to be turnover very quickly. And if ideally I think the most of the panels or most of the guidance is saying seven to 10 days, and ideally it's not taking three weeks, but patients who can wait seven to 10 days are by far going to do better. And actually the BAML trial said the same thing, which is all those patients waited to be categorized. And every patient who either joined BAML or another clinical trial that matched their target did better than patients who got standard of care or supportive care. And so I think the summary version is faster is better, but waiting is best. So if the patient can wait and the data will be available in 10 days, that is the best possible option for the patient. And when they can't wait, which we all know, there are some patients that truly can't wait and they're like five seconds from a ventilator or even on a ventilator, for those patients, we do the best we can.

Dr. Naval G. Daver: And I think that's really important to highlight. Because I think a lot of community and smaller academic or even fellows when you talk to them, this is a different message that was told to us 20 years ago. The motto was the sun should not set on AML. It was like appendicitis. So now I think a lot of times we call people and we say, oh, it's OK, you can wait 10 to 14 days. And I think that is because now we actually have treatments that have a bigger impact on changing the course of the disease than just rushing in with Anthracycline Cytarabine and hoping it works. Now, I do think it's important, like you said, to make sure that the patient is appropriate. Because the study you mentioned, Kristof Roleg [ph] in the German AMLSG group, they published that. But those were all patients who actually made it onto the AMLSG trials. So they were actually patients who were reasonably stable. And I think for 80, 90%, that's probably true. But of course, if you have somebody with the new FLT3 proliferative, you start and then you have to tailor the therapy later to whatever you find is a mutation.