Survival Improvement Seen With Gilteritinib Across FLT3-ITD+ AML Subgroup Analyses

In an interview with Targeted Oncology, Mark J. Levis, MD, PhD, discussed the findings from the follow-up analyses of the ADMIRAL trial that were presented at the 2019 ASCO Annual Meeting. He explained what these findings mean for the patient and what questions future research will aim to answer.

Mark J. Levis, MD, PhD

In the phase III ADMIRAL trial, the FLT3 inhibitor gilteritinib (Xospata) significantly improved overall survival (OS) in patients with relapsed/refractoryFLT3-mutant acute myeloid leukemia (AML).1Finding from the trial led to the FDA's approval of gilteritinib in this setting, and theseOS data were later added to gilteritinib’s label.

Overall 371 patients were randomized 2:1 to receive either gilteritinib monotherapy or salvage chemotherapyin the ADMIRALtrial.The median OS was 9.3 months with gilteritinib versus 5.6 months with salvage chemotherapy. This led to a 36% reduction in the risk of death.

The overall response rate (ORR) was 68% in the gilteritinib arm versus 26% with salvage chemotherapy. However, the median duration of drug exposure was significantly longer with gilteritinib at 4.1 versus 0.9 months with salvage therapy. The median time to achieve a complete response was 1.8 versus 1.1 months with gilteritinib versus salvage chemotherapy, respectively.

Further analyses investigated these results to find whether specific co-mutations or theFLT3-ITD allelic ratio (AR) may have an impact on responses to gilteritinib in patients with relapsed/refractoryFLT3­-mutated AML.2Data from 361 of the 371 patients enrolled in the ADMIRAL trial were analyzed; 4 major co-mutation cohorts were identified, includingNPM1,DNMT3A,DNMT3A/NPM1, andWT1mutations in addition toFLT3mutations.

Almost half of the patients had anNPM1mutation (47.9%), while 31.9% had theDNMT3Amutation, and 23.8% had both of these mutations. An additional 18% of patients harbored aWT1mutation. The analysis demonstrated that gilteritinib treatment was superior in terms of both ORR and OS in all 4 cohorts of co-mutations. Patients with both theNPM1andDNMT3Amutations had the greatest survival benefit compared with the other cohorts.

Long OS times were seen in patients with both high and low FLT3-ITD ARs with gilteritinib treatment compared with chemotherapy. In patients treated with gilteritinib, the OS was longer in patients with lowFLT3-ITD AR than those with highFLT3-ITD AR (7.1 vs 10.6 months), but this difference was not statistically different (HR, 1.341;P= .0712). However, Levis noted that the AR findings were even less significant for patients receiving chemotherapy.

“[FLT3-ITD AR] can’t be that accurately measured and in chemotherapy it doesn’t matter. If you have a high AR, chemotherapy [won’t provide any benefit], and if you have a low AR, [chemotherapy won’t provide much benefit],” said Mark J. Levis, MD, PhD. “You should never be using AR, in my view, to change your management.”

In an interview withTargeted Oncology, Levis, program leader of Hematologic Malignancies and Bone Marrow Transplant Program at the Sidney Kimmel Comprehensive Cancer Center, and professor of oncology at John Hopkins University, discussed the findings from the follow-up analyses of the ADMIRAL trial that were presented at the 2019 ASCO Annual Meeting. He explained what these findings mean for the patient and what questions future research will aim to answer.

TARGETED ONCOLOGY: What was your rationale for running further analyses on this group of patients?

Levis:This abstract is a follow-up study that is based on the ADMIRAL trial. The ADMIRAL trial tested gilteritinib, which is a novel FLT3 inhibitor that is active against the TKD and ITD versions of the receptor. Patients with relapsed/refractoryFLT3-mutant AML were randomized to receive either gilteritinib monotherapy or salvage chemotherapy. The trial was positive, and the drug was approved on the basis of that trial.

For patients who haveFLT3-ITD—mutated AML when they present at diagnosis, there is more to the story than just the fact that they have that mutation. How much of the mutation do they have? That’s actually reflective in this term calledFLT3-ITD AR, and that’s like a variant allelic fraction (VAF) for our field. How much of the mutation do they have? The more they have, the worse it is.

Co-mutations have an impact also, so what co-mutations do they present with?NPM1is a very common mutation found in AML, and it’s particularly common inFLT3-mutant AML. If you are a patient withFLT3-ITD—mutant AML and have theNPM1mutation, it’s a little less bad. TheNPM1has a mitigating effect. It doesn’t make it good, but it’s less bad.

From this, the obvious question came up that now we have a way of treating these patients with relapsed/refractoryFLT3-mutant AML, is there anything that will predict how they will do, referring to theFLT3-ITD AR or co-mutations? What happens there? We actually had on-study samples from almost all of the patients, 361 of the 371 trial patients, that we could do these analyses on. We used the Archer Core Myeloid Panel, and we used a VAF of co-mutations of 2.7%, and asked what impact did this have on the patients who got chemotherapy, gilteritinib, or either?

TARGETED ONCOLOGY: What were the findings from these analyses?

Levis:The bottom-line was in terms of response rate, it didn’t matter; gilteritinib had a way better response rate than chemotherapy no matter what group you looked at, didn’t matter which mutation.

However, something interesting came up when you actually broke the data down. If you look atNPM1mutations overall in the whole population, they did worse, but that wasn’t supposed to happen. They were supposed to do better, but they were clearly worse. They were clearly worse in the patients we were trying to treat with chemotherapy. Gilteritinib actually could have overcome that worseness, butNPM1mutations normally mitigate that fact when you treat with chemotherapy at diagnosis. Treating at relapse with chemotherapy actually has the opposite effect.

TheFLT3-ITD AR pretty much behaved as expected; the more you have, the worse you did, although gilteritinib was actually able to overcome that. Again, theFLT3-ITD AR is a reflection of how much tumor you have, and if you presented with advanced disease, it’s harder to fix. Gilteritinib clearly did better than chemotherapy. Chemotherapy was pretty hopeless with a highFLT3-ITD AR, but even gilteritinib had some trouble with a highFLT3-ITD AR compared to lowFLT3-ITD AR.

There was an interesting category of patients who had anNPM1mutation, aDNMT3Amutation, and of course, they had aFLT3mutation. Those patients [did very poorly] when treated with chemotherapy. However, with gilteritinib, we actually saw a remarkably impressive survival rate, a plateau. It looked like patients with newly diagnosed AML being treated. This subgroup, which was 25% of the patients enrolled, is a uniquely defined subgroup that you can say will really benefit from gilteritinib, but now we need to know why.

What if we treat these patients upfront? Are they going to do better if treated with gilteritinib upfront? This study raises as many questions as it answers, and we have a lot to do going forward. It has a lot of “gee-wiz” components to it.

TARGETED ONCOLOGY: Could you speak more to the frequency of these mutations that you saw in this group of patients?

Levis:Virtually half of the patients, 47% in the study, had anNPM1mutation, as expected for this patient population. About 30% had aDNMT3Amutation, and the group that hadDNMT3AandNPM1was about 23.9%, so roughly a quarter [of patients] had both theDNMT3AandNPM1.WT1mutations [were present in] 18% of patients, and they actually looked like that double mutation combination in terms of the way everybody [did poorly] if you gave them chemotherapy. Gilteritinib seemed to handle that well.

We don’t fully understand how to explain these results. Interestingly, the patients withNPM1mutations had a typically highFLT3-ITD AR. We think ofNPM1as a myeloproliferative mutation with very high white counts, so those go along. Having lots of disease when you relapse, it’s probably bad, but theDNMT3Acomponent has such an distinct effect on theNPM1-mutated,FTL3-mutated, we don’t understand. Again, it was a pretty dramatic effect.

TARGETED ONCOLOGY: What are the key questions we still need to try to answer in the next analysis or other research?

Levis:In the relapsed setting, you’re not going to get much with monotherapy. Most patients are still going to die. However, you can actually identify this subgroup as 25% of patients that might have long-term survival with gilteritinib alone, so you might have a patient population that you don’t necessarily need to be that aggressive in treating, which is pretty cool, but moving forward, I think we are going to be stratifying trials based on these new data. We want to make sure to really look at this double mutation group to see if, maybe in fact, this is a group that doesn’t need bone marrow transplant; this is a group that is susceptible to gilteritinib. Trying to explain why they do this, I think, is going to lead us back to the lab, I suspect. We will do some mouse studies with mouse models to try and understand this better.

TARGETED ONCOLOGY: Do you think this further necessitates the need for next-generation sequencing in patients with AML and making sure to test all patients for these mutations?

Levis:Of course. We learn so much more when we have this information. This is a component of virtually any trial moving forward, you have to have [next-generation sequencing] information to understand who’s going to respond afterwards, what are the mechanisms of resistance, how to predict resistance, etc. It’s an obvious component.

TARGETED ONCOLOGY: What were your thoughts on the FDA’s addition of the OS data to the gilteritinib label for AML?

Levis:We already knew those data, which was why none of us were surprised when the FDA granted regulatory approval “seemingly” without that data; I think the FDA had those data too, they were just waiting for it to mature. You could see where the curves were separating nicely, and it was very reassuring. I think it’s reassuring just when you look on the label and see a dramatic difference between the gilteritinib and salvage chemotherapy arms, you won’t be particularly cynical in that this drug will only get a little benefit—it will actually get a lot of benefit and it makes the approval more reassuring and a patient’s use of the drug more reassuring.

This study was done in part because the response rate to this monotherapy wasn’t all that high in this high-risk population, look for some combination where maybe chemotherapy does have a chance. No, chemotherapy does not have any chance against this population within any group you look at. That was one of the bottom lines of this study.


  1. Perl AE, Martinelli G, Cortes JE, et al. Gilteritinib significantly prolongs overall survival in patients with FLT3-mutated (FLT3mut+) relapsed/refractory (R/R) acute myeloid leukemia (AML): Results from the Phase III ADMIRAL trial. Presented at: 2019 AACR Annual Meeting; March 29 to April 3, 2019; Atlanta, GA. Abstract CTPL04.
  2. Levis MJ, Perl AE, Martinelli G, et al. Effect of gilteritinib on survival in patients with FLT3-mutated (FLT3mut+) relapsed/refractory (R/R) AML who have common AML co-mutations or a high FLT3-ITD allelic ratio.J Clin Oncol.2019;37(suppl; abstr 7000).