Adding a third AML drug missed overall, but hinted at a benefit in one mutation group

Acute myeloid leukaemia in older adults is one of oncology's hardest problems. Many patients are too frail for the intensive chemotherapy that can cure younger people, so treatment leans on gentler combinations — and even those often fail when the cancer carries a mutation called FLT3-ITD, which drives the disease and resists therapy. A new randomised trial tested whether adding a targeted drug to a standard gentle regimen would help. The headline answer was no; the more interesting answer, buried in a subgroup, was a qualified maybe [s1].

The trial, INTERVENE, was published in Blood and tested whether adding midostaurin — a drug that blocks the FLT3 signalling protein — to low-dose cytarabine and venetoclax improves outcomes in newly diagnosed AML in patients aged 60 or older with a non-adverse karyotype [s1]. After a run-in phase in 18 patients established that the three drugs could be combined safely, with low-dose cytarabine and venetoclax given on days 1 to 28 and midostaurin added on days 11 to 28, the trial randomly assigned 120 patients 2:1 to the three-drug regimen (abbreviated LVM) or the two-drug backbone (LV) [s1].

What the trial asked

The reasoning behind the trial reflects a running frustration in the field. Earlier randomised studies adding FLT3 inhibitors to treatment for older, unfit patients had failed to improve survival [s1]. Newer FLT3 drugs such as gilteritinib and quizartinib looked more promising when added to venetoclax-based therapy, but at a cost: the combinations hit the bone marrow hard, forcing doctors to cut doses substantially [s1]. Midostaurin, given sequentially rather than continuously, was the trial's attempt to get a FLT3 benefit without that toxicity penalty.

The primary endpoint was composite complete remission — the proportion of patients whose leukaemia was cleared from the bone marrow, counting both full and near-full responses [s1]. It is a standard early marker of whether a regimen works, though not the same as cure or long-term survival, which depend on whether the remission holds.

What it found

On its primary endpoint, the three-drug regimen fell flat. Composite complete remission occurred in 72% of patients on LVM versus 68% on LV — a four-point gap that did not meet the trial's threshold for a real improvement across the whole population [s1]. For the average enrolled patient, in other words, the extra drug added nothing measurable.

The signal was in the patients the drug was designed for. Among the 29 participants whose leukaemia carried the FLT3-ITD mutation, composite complete remission was 82% with the three-drug regimen versus 57% with two drugs — a 25-percentage-point difference, with an 0.85 posterior probability that the true difference exceeds five points and a 95% credible interval running from −10 to 59% [s1]. The regimen also cleared the FLT3-ITD mutation far more often on ultrasensitive sequencing (65% versus 0%), and in that subgroup overall survival was longer: a median of 24.2 months versus 8.8 months, with a hazard ratio of 0.34 (95% credible interval 0.071 to 0.960) [s1].

How to read it

This is a trial whose most striking numbers come with the heaviest caveats, and holding both at once is the whole point. The survival and remission differences in the FLT3-ITD group are large and biologically coherent — this is precisely the mutation midostaurin targets. But the subgroup contained just 29 patients, the analysis was not the trial's primary question, and its credible interval for the remission difference still touches below zero [s1]. Subgroup findings of this kind are where oncology has repeatedly been fooled, which is why they are treated as hypothesis-generating rather than practice-changing. The authors' own framing is cautious: midostaurin "may have a role" in newly diagnosed FLT3-ITD AML unfit for intensive chemotherapy, not that it is established [s1].

What the trial does deliver cleanly is a negative primary result and a safety profile — the three drugs were well tolerated — that make the FLT3-ITD question worth pursuing properly [s1]. It joins a wave of trials rebuilding AML and related blood-cancer treatment around venetoclax, including azacitidine plus venetoclax as AML induction and pirtobrutinib combinations in chronic lymphocytic leukaemia.

What to watch

The logical next step is a trial powered specifically for the FLT3-ITD population, where this regimen's apparent benefit could be confirmed or dissolve; the registry record (ACTRN12619001655134) marks INTERVENE as complete [s1][s2]. Until then, a four-point miss overall and a dramatic subgroup should be read as a lead, not a verdict.

This article describes research and is not medical advice. Leukaemia treatment decisions are for patients and their haematology teams.