Evaluating emerging mutations in patients with acute myeloid leukaemia who relapsed while receiving gilteritinib therapy in the ADMIRAL trial

Dr Catherine Smith - University of California San Francisco, San Francisco, USA

I presented work looking at resistance mechanisms on the phase III trial of gilteritinib randomised to salvage chemotherapy in FLT3 mutant AML. What we did was we looked at 40 patients who relapsed on gilteritinib that had paired baseline and relapse samples. We analysed them using next generation sequencing for a targeted panel of genes associated with myeloid leukaemias. We noted that the most common mutations associated with gilteritinib resistance were mutations within the RAS/MAP kinase signalling pathway as well as a small number of patients that instead developed mutations within FLT3 itself at the gatekeeper residue F691L largely.

The patients were treated in the phase III clinical trial. We looked at the patients treated in the gilteritinib arm. We took all the patients that responded and subsequently relapsed and looked for the patients that had paired samples. Again these samples were analysed using a next generation sequencing panel. We sequenced them using a commercial panel called the ArcherDx Core Myeloid sequencing panel that sequences 37 genes associated with AML. It sequenced the entire FLT3 gene and then targeted hotspots in other genes.

The main conclusions are that the most common mechanisms of relapse do appear to be RAS/MAP kinase gene mutations and there’s only a small proportion of patients that develop resistance due to F691L gatekeeper mutations. One of the most intriguing findings in the study, however, is that some patients do have RAS mutations at the time of drug initiation and about 40% of these patients, actually 39% to be clear, still achieve some mechanism of complete remission. So just because you have a RAS mutation doesn’t mean you might not respond. We’re not yet sure why this might be except that we noted that there were fewer mutations per patient in the baseline samples compared to the relapse samples.

Could this be used to predict response?

There should be concern if there is a pre-treatment RAS mutation that that patient should be monitored very closely for relapse. But the data suggests at this point that you should still try gilteritinib because some of these patients still respond.