EHA 2019 | ADMIRAL analysis: survival by mutation status in gilteritinib-treated FLT3+ R/R AML

This is relatively new data, and I think it’s going to emerge, potentially, is quite important. We looked at the data from the ADMIRAL study, which was a randomized study of gilteritinib versus salvage chemotherapy for relapsed refractory FLT3 mutant patients. And that was a positive trial, and had an overall survival benefit for

This is relatively new data, and I think it’s going to emerge, potentially, is quite important. We looked at the data from the ADMIRAL study, which was a randomized study of gilteritinib versus salvage chemotherapy for relapsed refractory FLT3 mutant patients. And that was a positive trial, and had an overall survival benefit for gilteritinib monotherapy. But then we broke the patients down by mutation status to ask, are there any that responded better or worse? Were there any differences?

And I think that the big points were, we could find no subgroup in which gilteritinib did not thoroughly outperform chemotherapy. But even then, we found some interesting findings that we haven’t fully explained. One, NPM1 mutations, which are normally considered less unfavorable, I don’t like to say favorable, they’re less unfavorable. At diagnosis, they’re less unfavorable. It turns out, in this setting, they’re quite unfavorable, at least with chemotherapy.

So a patient with an NPM1 mutation who’s relapsed, and has a FLT3 mutation, does appallingly poorly with chemotherapy. Gilteritinib doesn’t seem to mind. And as I mentioned, that there really no subgroup where it doesn’t do just fine.

We noted, in addition, that there was a distinct subgroup, one quarter of the patients had a combination of FLT3 ITD, DNMT3A, and NPM1 mutation. And they seemed to respond remarkably well to gilteritinib, such that it looked very different than a relapsed AML population. It looked like a newly diagnosed population. They had a remarkably good survival, and, in contrast, the chemotherapy did appallingly poorly.

Finally, we wanted to look at the variant allele frequency or allelic ratio, the burden of the ITD mutation in this population. And that came out pretty much as expected, but it’s often not what people think. They would think that a highly allelic burden is going really work… gilteritinib will be very effective in a patient with a high allelic burden. No, it is effective. It’s much more effective than chemotherapy, it’s true, but that’s a very difficult population to deal with. They have a very high tumor burden. They’re typically in sort of advanced disease when they relapse.

And so, while gilteritinib is far more effective than chemotherapy for those patients, gilteritinib would rather have a low allelic ratio patient to treat. So the lower allelic ratio patients did better with gilteritinib than the highly allelic ratio. Low allelic ratio, you still have a chance with chemotherapy. Chemotherapy can still salvage some of those, although it gilteritinib still looks better.

So the big pictures are, NPM1 at relapse is bad for chemo. DNMT3A NPM1 FLT3 ITD is a uniquely responsive subgroup for gilteritinib. And we want to look at that with other FLT3 inhibitors. And allelic ratio. The more disease you have, the worse things are. But gilteritinib, again, clearly better than chemotherapy.