Maintenance decitabine after intensive therapy for AML in older FLT3-ITD-negative patients

Dr James Foran - Mayo Clinic Florida, Jacksonville, USA

We performed a large phase III randomised study of intensive therapy in patients with acute myeloid leukaemia over 60 and we studied a standard intensive induction and consolidation strategy and we compared it to a slightly lower intensity drug with clofarabine. In our first reports, which we reported a couple of years ago at ASH, we found that the intensive therapy was superior. We made a number of important observations from that study about the importance of complete remission, the impact of MRD positive disease in remission and even the role of allogeneic transplant for older adults in a prospective study.

This abstract deals with the second randomisation that was embedded in the E2906 trial. Patients who were in complete remission and had completed and recovered from two cycles of intensive consolidation chemotherapy were randomised one-to-one onto a maintenance treatment for a year versus observation. We picked a year because that was what our preliminary data was at the time and the sponsors, the National Cancer Institute, wanted us to adhere to that. So originally we had intended to have 172 patients but because our original study, the parent protocol, had stopped early because of the survival changes we didn’t complete the accrual to the second randomisation. So we had 70% accrual, we had 120 patients. They were randomised one-to-one to observation versus an intervention with low dose decitabine. Decitabine was given three consecutive days, 20mg/m2, every four weeks for a year. Patients had to have adequate hematologic recovery so there were some dose delays for some people.

Then we looked at disease free survival and overall survival afterwards. What we found, first of all, was that decitabine was generally well tolerated. We had some cytopenias but no deaths. We had a 9% incidence of fever neutropenia but no major problems with it. It had a significant impact on disease free and overall survival. We had an unusual statistical design just because of the fact it was embedded in a large phase III study and the number of patients we had. But that design allowed for a one-sided p-value of 0.1 to be significant and we met that endpoint on the multivariate analysis for disease free survival and, importantly, for overall survival.

So we believe it supports a definitive randomised study; we think it’s potentially an important innovation in this setting. The patients who went on it tended to be the intermediate risk patients. Almost half of them were over the age of 70 so it reflects a real life population who would get to that phase. But, importantly, when we looked back at the FLT3-ITD status of patients most did not have a FLT3 mutation or did not have FLT3-ITD and that group had a strong survival advantage of decitabine maintenance versus observation. So it suggests to us that there’s a group that we can target to really make a survival impact in that space or in that population.

We’re planning to take it forward and to do this in a definitive study, although we think our results are very compelling, and possibly look for alternate strategies such as some of the new oral agents that are out there to try to capitalise on this.

How are comorbidities managed in these patients?

It’s a great question. Acute myeloid leukaemia we give intensive therapies and get younger adults through it and we think we cure a substantial number of them. But the median age of AML is 70 or higher and so it really is a disease of older adults. The study was already selected for people who were fit enough for an intensive approach so we know that that’s not for everybody, there are some people who aren’t strong enough for that. We have data on comorbidities, we have not analysed that in detail yet, but we also acknowledge that we have selected a more fit population. On that basis we didn’t really struggle… let me say it differently. We didn’t see a difficulty of people getting on the decitabine. That being said, the median number of cycles was six; they were eligible for up to thirteen cycles over a year, every four weeks over a year. The median number was six, partly because five patients never got it, even though they were randomised to that, and partly because we think patients were having some dose delays. So it’s possible that a slightly lower dose would keep people on schedule. It’s also possible that going beyond a year will help. But we’re going to have to ask that in a prospective study.

The major messages really are we think that there’s an advantage of maintenance decitabine. We think it has to be definitively shown in a powered randomised phase III trial but our results are compelling, particularly in the large, intermediate risk population who are FLT3-ITD negative. We think that’s where it’s most likely to have its impact.