In brief
What breaks: a cell stuck in adolescence
The familiar picture of leukaemia is of cells dividing furiously. It is more accurate to say the cells cannot grow up.
A blood-forming cell travels from immature precursor to mature working blood cell. Along that road programmes of genes switch on and off. In leukaemia with a KMT2A rearrangement a chromosomal break occurs: the KMT2A gene is severed and fused to another gene, producing a fusion protein.
That fusion protein settles on particular stretches of DNA and keeps the genes of immaturity permanently switched on — the very ones that should be switching off as maturation proceeds. The cell is stuck at the blast stage: it does not mature but keeps dividing. The marrow fills with immature cells and there is no room left for normal blood formation.
The rearrangement occurs in infants in their first year and in adults alike. Its prognosis has traditionally been among the worst, and until recently, in relapse after standard therapy, no targeted treatment existed at all.
The target chosen was the guide, not the fault
The fusion protein is a poor target in itself: it offers no convenient pocket into which a drug might fit. But it does not work alone.
To reach the right stretches of DNA the fusion uses the services of menin — an ordinary cellular protein encoded by the MEN1 gene. Menin is not an oncogene and does nothing wrong by itself; its role here is purely technical. It is the guide: it takes the fusion protein and delivers it to its landing site.
Hence the idea: remove the guide and the fusion cannot find the address. The immaturity programme switches off and the cell goes on to mature.
That is exactly what revumenib does — it wedges itself between menin and KMT2A and breaks their bond.
The same pathway turns out to be engaged by a different fault — mutation of the NPM1 gene, the commonest in adult acute myeloid leukaemia. The cause differs, the dependence on menin is the same, and so the drug works there too.
The comparison that fits: chemotherapy is burning the field to destroy the weeds along with everything else. A menin inhibitor is releasing a jammed brake: the cell drives itself to where it was supposed to arrive.
Differentiation syndrome: the other side of the same mechanism
Here is the point usually lost in a list of side effects. Differentiation syndrome is not a complication off to one side. It is the direct consequence of the drug working.
When millions of stalled cells are simultaneously given permission to mature, the body answers with violent inflammation. The picture is recognisable: fever, breathlessness, low oxygen, pulmonary infiltrates, pleural or pericardial effusion, rapid weight gain and oedema, falling blood pressure, acute kidney failure, rash.
The state can be fatal, which is why the FDA placed it in a boxed warning — the strictest format of caution available. In the trials the syndrome occurred in 60 of 241 patients (25%), and among those with myeloid leukaemia carrying a KMT2A translocation in 33%, one patient in three.
The management is prescribed firmly: on suspicion, corticosteroids are begun immediately and haemodynamics monitored until symptoms resolve.
The second half of the boxed warning is QTc prolongation and torsades de pointes, a dangerous rhythm disturbance. Hence the rules: do not start treatment when QTcF exceeds 450 msec, correct potassium and magnesium deficits before and during therapy, and interrupt, reduce or permanently stop the drug if the interval lengthens.
What the AUGMENT-101 trial showed
The trial ran at 22 sites in five countries. Patients from 30 days of age with relapsed or refractory leukaemia were enrolled. The drug is taken every 12 hours in 28-day cycles.
The age range in the KMT2A-translocated group is telling in itself: 94 patients treated, median age 37, from 1.3 years to 75. One fault, and both infants and the elderly.
| KMT2A translocation (57 evaluable) | Result |
|---|---|
| Complete remission + remission with partial recovery | *22.8%* (95% CI 12.7–35.8) against a 10% hypothesis, p=0.0036 |
| Overall response | *63.2%* (95% CI 49.3–75.6) |
| No detectable residual disease | 15 of 22 responders (*68.2%*) |
| Grade ≥3 adverse events | Febrile neutropenia 37.2%, differentiation syndrome 16.0%, QTc prolongation 13.8% |
The second cohort — myeloid leukaemia with an NPM1 mutation — drew older and more heavily pretreated patients: median age 63, a third with three or more prior lines, and three quarters already given venetoclax.
| NPM1 mutation (64 evaluable) | Result |
|---|---|
| Complete remission + remission with partial recovery | *23.4%* (p=0.0014) |
| Overall response | *46.9%* |
| Median duration of complete remission | *4.7 months* |
| Proceeded to transplantation | 5 of 30 responders (16.7%), three resumed the drug afterwards |
| Discontinuation for related adverse events | 4 patients (4.8%) |
How to read these percentages properly
Twenty-two per cent sounds modest, and this is the place not to err in either direction.
One more detail says much about the mechanism: the label directs treatment for at least six months in the absence of progression or intolerable toxicity. Maturation is a slow process, and stopping early 'for lack of efficacy' here means never giving the drug its chance.
How the tumour escapes
A separate study established why the response is lost in some patients. The cause is precise and almost elegant: mutations appear in the MEN1 gene — in menin itself — at exactly the site where the drug binds.
The tumour does not invent a detour and does not abandon the dependence that feeds it. It changes the lock the key was cut for. In targeted therapy this is the classic scenario, familiar from other targets.
The practical point: knowing the specific mechanism of resistance sets the direction for next-generation compounds designed for the altered lock.
What it costs
▸A two-part boxed warning — differentiation syndrome and rhythm disturbance; both call for action, not observation. ▸Close monitoring: electrolytes and ECG before and during treatment; therapy is not begun until the white cell count falls below 25 Gi/L. ▸Dosing every 12 hours, in 28-day cycles, continuously until progression or intolerance. ▸Interactions: strong CYP3A4 inhibitors require nearly halving the dose — 160 mg instead of 270 mg twice daily in patients from 40 kg. ▸Haematological toxicity: grade 3 or higher febrile neutropenia in 37.2%. ▸Effects on the fetus: the drug can cause harm, so effective contraception is required both for women and for men with partners of childbearing potential.
Who it is not for
▸Without a confirmed fault — the KMT2A translocation is determined by an FDA-authorised test, and the NPM1 mutation must be among the susceptible ones. Without that the drug cannot work at all. ▸With QTcF above 450 msec — treatment is not started until this is corrected. ▸With white cells above 25 Gi/L — these are lowered by other means first. ▸As a replacement for transplantation — the drug is most often a bridge to it, not an alternative. ▸There are no formal contraindications in the label — that section reads 'None', which does not cancel anything listed above.
What cannot be claimed yet
▸There was no randomised comparison. AUGMENT-101 is a single-arm trial: everyone received the drug and no control group existed. The assessment rests on a pre-specified threshold rather than on direct comparison. ▸Overall survival was not a primary endpoint. That the drug prolongs life has not been strictly demonstrated; what has been demonstrated is that it produces remissions where there were none. ▸Remissions are short — a median of 4.7 months in the NPM1 cohort; long-term outcomes depend above all on whether transplantation could be carried out. ▸Its place in the treatment sequence is undefined: combinations with venetoclax and others, and use in the first line, are under study but not yet reported. ▸The youngest are the least studied — enrolment began at 30 days of age, but few children were included.
Bottom line
▸The first menin inhibitor — a new class acting not by poison but by releasing a block on maturation. ▸The target is the guide, not the fault: removing menin is easier than attacking the fusion protein. ▸22.8% complete remissions in KMT2A-translocated disease and 23.4% with an NPM1 mutation — where no targeted treatment existed at all. ▸Differentiation syndrome in one patient in four — not a side effect off to one side but the other face of the mechanism; it sits in the boxed warning alongside rhythm disturbance. ▸Remission as a bridge to transplantation — in most cases that is the drug's practical purpose. ▸Resistance arises through mutations in menin itself — the tumour changes the lock to fit the key.
Your own case — which fault has been confirmed and whether a route to transplantation exists — can be reviewed at a consultation; the drug can be ordered here.
References
1. Issa GC, et al. Menin Inhibition With Revumenib for KMT2A-Rearranged Relapsed or Refractory Acute Leukemia (AUGMENT-101). J Clin Oncol. 2025;43(1):75–84. PMID 39121437
2. Arellano ML, et al. Menin inhibition with revumenib for NPM1-mutated relapsed or refractory acute myeloid leukemia: the AUGMENT-101 study. Blood. 2025;146(9):1065–1077. PMID 40332046
3. Issa GC, et al. The menin inhibitor revumenib in KMT2A-rearranged or NPM1-mutant leukaemia. Nature. 2023;615(7954):920–924. PMID 36922593
4. Perner F, et al. MEN1 mutations mediate clinical resistance to menin inhibition. Nature. 2023;615(7954):913–919. PMID 36922589
5. REVUFORJ (revumenib) — US Prescribing Information, Syndax Pharmaceuticals, Inc.
Key facts
- Revumenib (brand name Revuforj, Syndax Pharmaceuticals) is the first menin inhibitor, approved by the FDA in November 2024; in 2025 the indication was extended to a second molecular subtype.
- Indications: relapsed or refractory acute leukaemia with a KMT2A translocation in patients from one year of age, and relapsed or refractory acute myeloid leukaemia with a susceptible NPM1 mutation where no satisfactory alternative exists.
- Menin is not an oncogene but a chaperone protein: it delivers the fusion KMT2A protein to the stretches of DNA where the genes of cellular immaturity are switched on.
- The drug breaks the menin–KMT2A bond, the immaturity programme switches off, and leukaemic cells begin maturing into normal blood cells — the mechanism is differentiating, not cytotoxic.
- In AUGMENT-101 in KMT2A-translocated disease, among 57 evaluable patients complete remission or complete remission with partial haematologic recovery was achieved in 22.8% against a null hypothesis of 10% (p=0.0036), with an overall response of 63.2%.
- In 15 of 22 responders (68.2%) no residual disease could be detected with sensitive assays.
- In the NPM1-mutated cohort — 64 evaluable patients, three quarters of whom had received venetoclax — the complete remission plus partial-recovery rate was 23.4% (p=0.0014), overall response 46.9%, median duration of remission 4.7 months.
- The FDA boxed warning has two parts: differentiation syndrome, which can be fatal, and QTc prolongation with torsades de pointes.
- Differentiation syndrome occurred in 60 of 241 patients (25%), and in 33% of those with myeloid leukaemia carrying a KMT2A translocation.
- Treatment is started only once the white cell count falls below 25 Gi/L and continued for at least six months to give maturation time; the drug is not started if QTcF exceeds 450 msec.
