Oncofusion

start / graph / drug-ziftomenib

Drugstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q6-therapeutics

Ziftomenib (KO-539, Komzifti) — once-daily menin inhibitor approved in NPM1-mutant AML

also called KO-539 · Komzifti · KO539

Ziftomenib is a second menin-blocking pill, taken once a day, approved by the FDA in November 2025 for adults whose NPM1-mutated acute myeloid leukaemia has returned. Its KMT2A-rearranged arm had to be paused early because too many patients developed a dangerous reaction called differentiation syndrome.

The same thing, for a scientist

Ziftomenib is an orally bioavailable, highly potent and selective inhibitor of the menin–KMT2A interaction (C33H42F3N9O2S2, MW 717.89; a 4-aminopyrimidine/thieno-fused bicyclic bearing an indole-carbonitrile and a methanesulfonyl-piperazine; 3 Lipinski violations, a notably high-MW/lipophilic chemotype). It was developed from the University of Michigan (Grembecka/Cierpicki) menin-inhibitor series and licensed to Kura Oncology; Kyowa Kirin is the co-development partner. FDA approved 2025-11-13 at 600 mg once daily for adults with R/R NPM1-mutant AML with no satisfactory alternative (KO-MEN-001/KOMET-001, NCT04067336; n=112, CR+CRh 21.4%). It is NOT approved for KMT2A-rearranged disease and is not approved in children. Its distinguishing scientific feature is a partially non-overlapping resistance profile with revumenib: it retains activity against MEN1 T349M and G331R but not M327I or G331D.

Go deeper
Approval status — and what it is NOT approved for

2025-11-13 — FDA approved ziftomenib (Komzifti, Kura Oncology) for adults with relapsed or refractory AML with a susceptible NPM1 mutation who have no satisfactory alternative treatment options, at 600 mg orally once daily until progression or unacceptable toxicity 3.

Three limits matter for this lab:

  1. NPM1-mutant only. KMT2A-rearranged disease is not in the label.
  2. Adults only. No paediatric indication, in contrast to revumenib's 1-year floor.
  3. Efficacy: CR+CRh 21.4% (95% CI 14.2–30.2) in 112 patients, median duration of CR+CRh 5 months (95% CI 1.9–8.1), and 14/66 (21.2%) transfusion-dependent patients converted to independence 3.
Why the KMT2A-rearranged arm was halted — the differentiation-syndrome signal

This is the most instructive safety story in the whole menin class. In KOMET-001 phase 1, 83 patients received 50–1000 mg once daily. Grade 3 or worse differentiation syndrome occurred in 12/83 (15%), and there were two treatment-related deaths — one from differentiation syndrome and one cardiac arrest. The rate and severity of differentiation syndrome specifically drove the decision to halt enrolment of patients with KMT2A rearrangements 1.

Read this together with the revumenib label, where DS was 33% in KMT2A-r AML versus 18% in NPM1m AML [see drug-revumenib E3]. The consistent pattern across two chemically distinct drugs is that KMT2A-rearranged myeloid disease differentiates harder and more dangerously than NPM1-mutant disease. That is not a chemistry problem; it is disease biology, and it means any KMT2A-r menin programme must build in cytoreduction, steroid prophylaxis and dose ramp-up from the start.

At the recommended phase 2 dose of 600 mg, 9/36 (25%) patients with KMT2A-r or NPM1m achieved CR or CRh; among NPM1m patients at 600 mg, 7/20 (35%) achieved CR. No responses were seen at 200 mg — a clean dose–response that validated 600 mg 1.

The resistance asymmetry — the most actionable finding in this node

The 2026 discovery/preclinical paper tested ziftomenib against the four MEN1 mutants selected clinically by revumenib 2:

MEN1 mutantZiftomenib activity
T349Mretained antileukaemic activity
G331Rlow-nanomolar potency (GI50 25 nM or less), though several-fold weaker than WT
M327Iresistant
G331Dresistant

Co-crystal structures of ziftomenib with wild-type menin, menin-T349M and menin-G331R showed a similar binding mode across all three, rationalising the retained potency 2. In other words the two drugs, despite sharing a pocket, have partially orthogonal escape routes.

This directly generates a clinical strategy — genotype the MEN1 pocket at menin-inhibitor relapse and switch agent rather than abandon the class — and it also means the lab's menin-interface hypothesis is not refuted by the appearance of resistance mutations. The interface remains druggable; it is the individual chemotype that fails. See hyp-mutation-guided-menin-inhibitor-alternation.

Combinations (KOMET-007) and frontline ambitions

Ziftomenib is in combination development with intensive chemotherapy (7+3) and with venetoclax/azacitidine in newly diagnosed and R/R NPM1m and KMT2A-r AML; the programme is reviewed with the combination data in 4. The strategic logic is that a once-daily agent with no meaningful QTc signal is easier to layer onto a cytotoxic backbone than a twice-daily CYP3A4-boosted agent with a QT boxed warning — a real, if unglamorous, differentiator.

Facts and figures
chembl IDCHEMBL5095038
targetgene-men1
target interactionmenin–KMT2A (MBM1) protein–protein interface
modalitysmall molecule (oral, reversible PPI inhibitor)
chemical class2,4-diaminopyrimidine fused thiophene core with indole-carbonitrile and methanesulfonyl-piperazine arms; trifluoroethyl substituent
molecular formulaC33H42F3N9O2S2
mw freebase da717.89
alogp5.01
ro5 violations3
statusFDA approved 2025-11-13 (adults, R/R NPM1-mutant AML, no satisfactory alternative). NOT approved for KMT2A-rearranged disease. No paediatric indication.
sponsorKura Oncology (co-development Kyowa Kirin); chemotype originated University of Michigan (Grembecka/Cierpicki)
notableKMT2A-rearranged enrolment in KOMET-001 was halted because of differentiation-syndrome rate and severity
efficacy
phase 1b NPM1m at 600 mg20.0phase 1b NPM1m at 600 mg: 20phase 1b, KMT2A-r or NPM1…36.0phase 1b, KMT2A-r or NPM1m at 600 mg: 36R/R NPM1m AML (registrati…112.0R/R NPM1m AML (registration, KO-MEN-001): 112
dosing
{
  "route": "oral",
  "dose_mg": 600,
  "schedule": "once daily, continuous"
}
trials
[
  "NCT04067336"
]
efficacy — the numbers behind the chart
[
  {
    "population": "R/R NPM1m AML (registration, KO-MEN-001)",
    "n": 112,
    "cr_crh_pct": 21.4,
    "median_dor_months": 5
  },
  {
    "population": "phase 1b, KMT2A-r or NPM1m at 600 mg",
    "n": 36,
    "cr_crh_pct": 25
  },
  {
    "population": "phase 1b NPM1m at 600 mg",
    "n": 20,
    "cr_pct": 35
  }
]
key ae
[
  "differentiation syndrome 26% (fatal in 2)",
  "anaemia G3+ 24%",
  "febrile neutropenia G3+ 22%",
  "pneumonia G3+ 19%",
  "thrombocytopenia",
  "sepsis"
]
resistance
[
  "MEN1 M327I (resistant)",
  "MEN1 G331D (resistant)",
  "MEN1 T349M (retains activity)",
  "MEN1 G331R (retains low-nM activity)"
]
How this could bite us back
  • The KMT2A-r arm was halted for toxicity, not futility — the population this lab cares about is the one that tolerates the class worst.
  • M327I and G331D confer resistance to ziftomenib as well as revumenib, so a shared-pocket class has shared escape routes; a genuinely orthogonal mechanism (degrader, covalent, allosteric, or a different partner interface) is still needed.
  • High molecular weight (718 Da), aLogP 5.0 and three Lipinski violations make CNS penetration and paediatric formulation non-trivial; CNS disease is common in infant KMT2A-r ALL.
  • Approval in NPM1m but not KMT2A-r risks the class's commercial gravity drifting to adult AML, leaving infant KMT2A-r ALL under-studied.
Still open
  1. Ziftomenib retains potency against T349M and G331R menin. Would sequential or alternating menin inhibitors at MEN1-mutant relapse recover remissions in patients, or does the polyclonality of relapse defeat it?
  2. Why does KMT2A-rearranged disease produce more severe differentiation syndrome than NPM1-mutant disease with two structurally unrelated menin inhibitors — is it blast burden, lineage, or a qualitatively different differentiation trajectory?
  3. Is there a paediatric pharmacokinetic and safety path for a once-daily, non-QT-prolonging menin inhibitor in infants, where revumenib's twice-daily CYP3A4-boosted regimen is hardest to deliver?
Every source behind this entry (4)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In the phase 1 KOMET-001 trial (83 patients, 50–1000 mg once daily), grade 3+ differentiation syndrome occurred in 15% with two treatment-related deaths, and the rate and severity of differentiation syndrome drove the decision to halt enrolment of KMT2A-rearranged patients; at the 600 mg recommended phase 2 dose 25% of KMT2A-r or NPM1m patients achieved CR/CRh and no responses occurred at 200 mg.

    Ziftomenib in relapsed or refractory acute myeloid leukaemia (KOMET-001): a multicentre, open-label, multi-cohort, phase 1 trial · Wang ES, Issa GC, Erba HP, Altman JK, Montesinos P, DeBotton S, Walter RB, Pettit K, Savona MR, Shah MV, Kremyanskaya M, Baer MR, Foran JM, Schiller G, Ades L, Heiblig M, Berthon C, Peterlin P, Rodriguez-Arboli E, Salamero O, et al. · The Lancet Oncology 2024
    PMID 39362248DOI
    What this evidence cannot show

    Phase 1 dose-finding; 22 sites; adults only (18 years and older); small per-dose subgroups (KMT2A-r n=9-16 at relevant doses) so response estimates are imprecise; sponsor (Kura) co-authored.

    Evidence that argues against this (1)
    • The phase-1 per-dose figures did not survive into the registration dataset, and the differentiation-syndrome story reads differently at scale. In the registration-enabling phase 2 of the same trial (KOMET-001, R/R NPM1-mutant AML, n=92, median age 69), CR/CRh was 22% (95% CI 14-32) - below the 25% phase-1 subgroup estimate at 600 mg - ORR 33%, median response duration 4.6 months and median overall survival 6.6 months (95% CI 3.6-8.6). Differentiation syndrome occurred in 25% but was 15% grade 3 with NO grade 4-5 events and was 'manageable with protocol-defined mitigation', and only 3% discontinued for a ziftomenib-related adverse event. The two phase-1 treatment-related deaths therefore reflect dose-finding without mitigation rules, not a fixed property of the drug. Note also that the KMT2A-rearranged cohort halted in phase 1 was never carried to registration, so no later dataset has resolved the KMT2A-r differentiation-syndrome signal at all. Wang ES et al. 2025
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Ziftomenib retains antileukaemic activity against the revumenib-resistance mutant MEN1 T349M and low-nanomolar potency against G331R, but M327I and G331D are resistant; crystal structures of ziftomenib with wild-type, T349M and G331R menin show a similar binding mode.

    Discovery and preclinical activity of the menin-KMT2A inhibitor ziftomenib in acute leukemia models · Miao H, Wu T, Purohit T, Chen D, Klossowski S, Borkin D, Clegg B, Ray J, Park S, Stevens R, Kim E, Kempinska K, Wang Y, He M, Wen B, Goldman JW, Agrusa J, Ding C, Sulis ML, Sun D, Mody R, Kim AS, Ren P, Li LS, Liu Y, Burrows F, Kessler L, Cierpicki T, Grembecka J · Blood 2026
    PMID 41849234DOI
    What this evidence cannot show

    Engineered isogenic cell models and structural biology, not clinical data; the mutants were introduced experimentally rather than sampled from patients on ziftomenib; academic inventors of the series are senior authors.

    Evidence that argues against this (1)
    • No clinical series yet demonstrates that switching from revumenib to ziftomenib at MEN1-mutant relapse produces remissions; the prediction is untested in humans. Issa GC et al. 2025
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.regulatory

    FDA approved ziftomenib (Komzifti) on 2025-11-13 for adults with relapsed or refractory NPM1-mutated AML with no satisfactory alternative treatment, at 600 mg once daily; in KO-MEN-001 (NCT04067336, n=112) CR+CRh was 21.4% (95% CI 14.2–30.2) with median duration 5 months, and differentiation syndrome occurred in 26% (29/112) and was fatal in 2 patients.

    FDA approves ziftomenib for relapsed or refractory acute myeloid leukemia with a NPM1 mutation · US Food and Drug Administration · FDA Oncology (Cancer) Approval Notifications 2025
    What this evidence cannot show

    Single-arm registration; adults only; no KMT2A-rearranged indication, so the label cannot be extrapolated to the population of interest here. CITATION TYPE CORRECTED 2026-09-20: this was stored as `type: news`. It is an FDA approval announcement on fda.gov, which is a PRIMARY regulatory record, not a news report of one. Retyped to `regulatory`. URL fetched and the page title and approval date confirmed against the claim in the same session.

    Evidence that argues against this (1)
    • The label's n=112 and the peer-reviewed registrational publication's n=92 are DIFFERENT data cuts of the same cohort, and they give different numbers: CR/CRh 21.4% (label) vs 22% (publication); differentiation syndrome 26% vs 25%; median duration of CR+CRh 5 months vs a median duration of response of 4.6 months. Neither is the other. More importantly, the approval summary does not carry the number that bounds the benefit: median overall survival in the published cohort was 6.6 months (95% CI 3.6-8.6). A CR+CRh rate in the low twenties with a median OS under seven months is a bridging result, and the label alone does not let a reader see that. Wang ES et al. 2025
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.review

    Ziftomenib and revumenib have moved from proof-of-concept to approved therapy, and combination strategies (including KOMET-007 with intensive chemotherapy and with venetoclax/azacitidine) are now the main development path because on-target resistance emerged under monotherapy.

    PMID 42738274DOI
    What this evidence cannot show

    Review; the authors themselves note that the combination evidence base is early-phase, single-arm, with short follow-up and no comparative data.

    Evidence that argues against this (2)
    • The stated CAUSE of the pivot to combinations - 'because on-target resistance emerged under monotherapy' - is not what the resistance data show. In patient-derived xenografts from multiply relapsed KMT2A::AFF1 ALL, menin-inhibitor resistance was already present in patients who had NEVER received a menin inhibitor, with sustained on-target transcriptional suppression of canonical menin-inhibitor target genes in the resistant cells; RAS-pathway and TP53 comutations emerged but neither was sufficient to confer resistance in vitro. Much of the failure the combination programmes are designed to pre-empt is therefore prior-therapy-acquired and pocket-independent, which predicts that combinations will help by adding an orthogonal mechanism rather than by preventing on-target mutation. Mahdavi L et al. 2026
    • A second strand of the same argument is also unsupported by the registration data: prespecified subgroup analyses in KOMET-001 phase 2 found CR/CRh comparable regardless of previous therapy (including venetoclax) and regardless of comutation type. If prior-therapy class and comutation do not move the monotherapy response rate, the case for combination rests on response DURABILITY (median 4.6 months) rather than on response rate, and should be argued on that footing. Wang ES et al. 2025
Connected to