Oncofusion

start / graph / fusion-kmt2a-mllt1

Fusionstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q11-comparative

KMT2A::MLLT1 (MLL-ENL), t(11;19)(q23;p13.3)

also called MLL-ENL · MLL::ENL · KMT2A-ENL · MLL/MLLT1 · t(11;19)(q23;p13.3)

Here the broken KMT2A gene joins ENL, a close cousin of the AF9 partner, and the result is usually a lymphoid leukemia - sometimes of B cells, distinctively often of T cells. Its most surprising feature is that outcome depends almost entirely on the patient's age: infants do badly, while children between one and nine and those with T-cell disease have done remarkably well.

The same thing, for a scientist

KMT2A::MLLT1 is the third commonest KMT2A rearrangement (460/3401, 13.5%), accounting for ~18.5% of KMT2A-rearranged ALL, ~24% of infant ALL, and it is the leading KMT2A partner in T-ALL (n=43 in the DCAL series) [E1]. MLLT1/ENL is the paralogue of MLLT3/AF9: both carry an N-terminal YEATS acetyl-lysine reader and a C-terminal ANC1-homology domain (AHD) that docks AFF1/AFF4 and DOT1L. ENL breakpoints are heterogeneous, falling at ENL exons 2, 4, 6 or 7, so the amount of the YEATS reader retained in the chimera varies between patients - an under-appreciated source of biological heterogeneity within a single named fusion [E2]. On the KMT2A side, MLLT1 is one of only two partners (with SEPTIN6) that prefer **intron 11 constitutively** rather than as a function of age or therapy [E1]. Wild-type ENL is itself an acetylation-dependent leukemia dependency whose YEATS domain sustains oncogenic transcription, and YEATS-domain displacement is anti-leukemic in AML models [E3][E4]. Clinically the fusion is a striking exception to the KMT2A-r rule: in 35 children with t(11;19)(q23;p13.3) ALL, 11 of 13 infants failed treatment whereas no relapses occurred among B-precursor patients aged 1-9 years or among T-cell patients [E5]. Menin inhibition applies partner-agnostically [E6].

What it looks like
loading 5J9S…
PDB 5J9S ↗
shape

This is real experimental data — atoms placed by crystallography or cryo-EM, deposited by the people who did the experiment.

I have never read one of these pictures before

A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.

Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.

Anything in yellow sticks is a bound small molecule, which is usually the drug.

Go deeper
What is kept and what is lost on each side1 deeper

Kept from KMT2A: AT-hooks 1-3, SNL/repression region, MBM1/MBM2, the LEDGF-binding region, CXXC. Lost from KMT2A: PHD1-3, bromodomain, ePHD4, FYRN/FYRC, TAD, WIN motif, SET 1. The same partner-agnostic half as every other major-BCR fusion.

Kept from MLLT1/ENL: the C-terminal region including the AHD, the disordered docking module that binds the AFF1/AFF4 CHD and DOT1L - structurally and functionally homologous to the AF9 AHD.

Variably kept or lost - the YEATS domain. ENL's N-terminal YEATS domain is the acetyl-lysine reader 3. Genomic and transcript breakpoints in ENL have been mapped at exons 2, 4, 6 and 7, all in frame 2. An exon-2 junction retains nearly the whole ENL body and therefore most of the YEATS fold; an exon-6 or exon-7 junction removes it. The field routinely reports 'KMT2A::MLLT1' as if it were one protein; it is at least two, and which one a patient has determines whether the chimera itself can read acetylation or merely recruits SEC/DOT1L. We flag this as a genuine open question rather than asserting a single architecture.

ENL versus AF9: near-paralogues with different clinical outputs

ENL (559 aa, Q03111) and AF9 (568 aa, P42568) are paralogues with the same two-module architecture and overlapping interactomes. Yet MLLT1 is a lymphoid partner (18.5% of KMT2A-r ALL, top T-ALL partner) and MLLT3 a myeloid one (30.4% of KMT2A-r AML) 1. Whatever distinguishes them is not gross domain composition. Candidate explanations: differential expression of the two paralogues across haematopoietic stages (so the translocation is only productive where the partner locus is transcribed and 3D-proximal), different YEATS specificity for acetyl versus crotonyl marks, and different AHD affinity for AFF1/AFF4 versus repressive partners. This is one of the most informative unexplained contrasts in the whole KMT2A family.

Mechanism: acetylation reading plus the same elongation hijack

Class: chromatin-elongation-hijack, with an acetylation-reading layer that is uniquely prominent for this partner.

The shared steps are as for KMT2A::AFF1: menin/LEDGF docking, CXXC-directed delivery to CpG-island promoters, and SEC/DOT1L-driven release of paused RNA polymerase II.

The partner-specific layer is ENL's YEATS domain. Wild-type ENL binds acetylated histone H3 through YEATS and is required to sustain oncogenic transcription in AML; in two simultaneous 2017 studies, disrupting ENL - by degradation or by point mutations that abolish acetyl-lysine binding - suppressed leukemic gene expression and proliferation, establishing YEATS as a bona fide druggable reader pocket 34. Erb and colleagues showed with a chemical-genetic displacement system that acute loss of ENL YEATS function collapses transcription at key leukemia genes, which is as close to a pharmacological proof of concept as a reader domain gets 4.

The strategic point for a comparative graph: this is the one KMT2A partner that brings its own small-molecule-shaped pocket. A YEATS-binding ligand would be a partner-specific drug with a real binding site, in contrast to AFF1, whose contribution is entirely disordered scaffold. Whether that pocket lives on the fusion protein or only on endogenous wild-type ENL depends on the breakpoint (see above), which makes the breakpoint question therapeutically, not just academically, important.

Clinical phenotype, lineage and the age paradox

Lineage skew: lymphoid, and distinctively T-lineage. MLLT1 is ~18.5% of KMT2A-r ALL and the commonest KMT2A partner in T-ALL (n=43) 1. A small St Jude/POG series of 15 MLL/ENL cases split 7 B-precursor, 4 T-ALL, 4 AML - roughly the same distribution 2. No other KMT2A partner produces this much T-cell disease except AFDN via the minor BCR.

Age skew: bimodal. ~24% of infant ALL carries MLLT1, so it is a major infant partner; but it is also well represented in children aged 1-9 and in T-ALL adolescents 15.

The age paradox - the most clinically important fact about this fusion. In 35 children with t(11;19)(q23;p13.3) ALL identified from 3,578 karyotyped ALL cases, all had MLL rearrangement and MLL-ENL transcripts, yet outcome tracked age and immunophenotype, not genotype: only 2 of 13 infants remained in remission, versus 6 of 11 older children, and strikingly no relapses at all occurred among B-precursor patients aged 1-9 years or among T-cell patients 5. The authors concluded that these two subsets should not be transplanted in first remission 5.

Contrast with AFF1. KMT2A::AFF1 is uniformly bad at every age. KMT2A::MLLT1 is bad in infants and good afterwards. Same 5' half, same retained module set, same menin dependence - so the prognostic difference cannot be a property of the fusion protein alone. It is a property of the cell the fusion lands in, and it is the strongest single argument in this graph that cell-of-origin, not oncoprotein chemistry, sets curability in the KMT2A family.

Druggability and therapeutic status

Menin inhibition: clinically validated for the class, partner-agnostic. The retained MBM1/MBM2 module is identical, and AUGMENT-101 enrolled KMT2A-translocated disease without partner restriction (CR+CRh 22.8%, ORR 63.2%) 6.

YEATS-domain ligands: emerging, and the most interesting partner-specific opportunity in the family. The ENL YEATS pocket is a genuine, shallow-but-tractable acetyl-lysine binding site, validated genetically and by chemical-genetic displacement 34. Probe compounds against ENL/AF9 YEATS exist; none is a clinical agent.

DOT1L: applies as for MLLT3, with the same modest clinical result.

The honest framing: for the 1-9-year-old and T-ALL groups, current chemotherapy already cures most patients 5, so the therapeutic need is concentrated in infants - exactly the group where the fusion behaves like KMT2A::AFF1 and where a partner-specific YEATS drug would have to beat a cell-of-origin problem rather than a protein problem.

Facts and figures
gene5KMT2A
gene3MLLT1
cytobandt(11;19)(q23;p13.3)
cytoband noteMust be distinguished from t(11;19)(q23;p13.1), which is KMT2A::ELL - a different partner on the same chromosome arm with a myeloid rather than lymphoid phenotype.
prognosisAge- and lineage-dependent rather than fusion-dependent: 2/13 infants remained in remission versus 6/11 older children in a 35-patient t(11;19)(q23;p13.3) ALL series, with no relapses among B-precursor patients aged 1-9 years or among T-cell patients; large modern series still show poor outcome for KMT2A-r infants regardless of partner.
uniprot5Q03164
uniprot3Q03111
uniprot5 verifiedQ03164 = KMT2A_HUMAN, 3969 aa (UniProt REST, checked 2026-09-18)
uniprot3 verifiedQ03111 = ENL_HUMAN, gene MLLT1, protein ENL, 559 aa, Homo sapiens (UniProt REST, checked 2026-09-18)
exon numbering noteKMT2A exon numbers follow the Meyer/Marschalek DCAL consensus. ENL exon numbers follow Fu et al. 2007. The mapping of ENL exon boundaries onto the YEATS fold is our inference from domain coordinates, not a published protein-level result.
PDB noteNo structure of the KMT2A::MLLT1 fusion protein exists. 5J9S = ENL YEATS domain in complex with an H3K27-acetylated peptide - the acetyl-lysine reader pocket, which is the only genuine small-molecule-shaped site any KMT2A partner contributes, and which is retained in the fusion only for upstream ENL junctions; 2LM0 = AF4(AFF1)-AF9 complex, a paralogue proxy for the ENL AHD:AFF1-CHD interface (no ENL-AHD:AFF1 structure exists); 7UJ4 and 4GQ6 = menin with revumenib and with an MLL peptide, the partner-agnostic drugged node. Verified against the RCSB entry API on 2026-09-18.
druggability classemerging
druggability noteClinically validated only through the partner-agnostic menin node. Uniquely among the six KMT2A partners here, MLLT1 contributes a real ligandable pocket (the YEATS aromatic cage), genetically and chemical-genetically validated but with no clinical-stage inhibitor.
stage of hematopoiesisbimodal: a fetal/neonatal lympho-myeloid-primed progenitor in the infant cases, and a committed lymphoid progenitor (including a T-lineage-committed thymic precursor) in the older children and T-ALL cases that do well
age peakbimodal: <1 y (poor-outcome infant subset) and 1-9 y plus adolescent T-ALL (good-outcome subsets)
developmental order rank3
developmental order noteRanked 3 of 10 as a weighted compromise: the infant subset arises in a cell nearly as primitive as KMT2A::AFF1's, while the T-ALL and 1-9-year subsets arise in committed lymphoid progenitors that would rank 7-8. The bimodality is the biologically important fact and a single integer conceals it.
domain name notedomains_retained_5 / domains_lost_5 use the exact domain names from data/graph/nodes/king/gene-kmt2a.json so the dashboard's 3D domain ruler can match them. The LEDGF/PSIP1-binding region is retained biologically but has no lane in the parent node, so it is recorded in each breakpoint's phenotype_notes instead.
frequency
all_kmt2a_r_acute_leukemi…13.5all_kmt2a_r_acute_leukemia_pct: 13.5kmt2a_r_all_pct18.5kmt2a_r_all_pct: 18.5infant_all_pct24.0infant_all_pct: 24kmt2a_r_t_all_n43.0kmt2a_r_t_all_n: 43n_cases_dcal_cohort460.0n_cases_dcal_cohort: 460
diseases
[
  "infant B-ALL",
  "childhood B-ALL",
  "T-ALL",
  "AML",
  "MPAL"
]
frequency — the numbers behind the chart
{
  "all_kmt2a_r_acute_leukemia_pct": 13.5,
  "kmt2a_r_all_pct": 18.5,
  "infant_all_pct": 24,
  "kmt2a_r_t_all_n": 43,
  "n_cases_dcal_cohort": 460
}
breakpoints
[
  {
    "id": "bp-kmt2a-intron-11-mllt1-e2",
    "gene5_exon_last": 11,
    "gene3_exon_first": 2,
    "frequency_pct": 39.5,
    "population": "infant and childhood ALL; MLLT1 prefers KMT2A intron 11 constitutively. frequency_pct is the share of all KMT2A breaks in BCR subregion C (exon 11 to exon 13), not the share of MLLT1 fusions",
    "domains_retained_5": [
      "MBM1",
      "MBM2",
      "AT-hook 1",
      "AT-hook 2",
      "AT-hook 3",
      "SNL1/2",
      "CXXC"
    ],
    "domains_lost_5": [
      "PHD1",
      "PHD2",
      "PHD3",
      "Bromodomain",
      "PHD4",
      "FYRN",
      "TAD",
      "FYRC",
      "WIN",
      "SET"
    ],
    "domains_retained_3": [
      "most of the ENL YEATS domain (exon-2 junction)",
      "ENL central disordered region",
      "AHD (ANC1 homology domain; binds AFF1/AFF4 and DOT1L)"
    ],
    "domains_lost_3": [
      "ENL exon-1-encoded N-terminus (partial truncation of the YEATS fold)"
    ],
    "phenotype_notes": "Exon 2 is the commonest reported ENL junction. HOXA-low, infant-associated KMT2A break. This architecture is the one in which the chimera itself may retain acetyl-lysine reading capacity. The retained KMT2A half also contains the LEDGF/PSIP1-binding region (no counterpart lane in the parent gene-kmt2a domain map, so it is recorded here rather than in the domain arrays)."
  },
  {
    "id": "bp-kmt2a-intron-11-mllt1-e6-e7",
    "gene5_exon_last": 11,
    "gene3_exon_first": 7,
    "frequency_pct": null,
    "frequency_pct_note": "Not separately quantified in the source series; 4 of 15 MLL/ENL cases used the novel exon-4 or exon-6 junctions, and exon 7 was one of two previously described junctions, but per-junction frequencies in an unselected population are unknown.",
    "population": "subset of childhood ALL and T-ALL",
    "domains_retained_5": [
      "MBM1",
      "MBM2",
      "AT-hook 1",
      "AT-hook 2",
      "AT-hook 3",
      "SNL1/2",
      "CXXC"
    ],
    "domains_lost_5": [
      "PHD1",
      "PHD2",
      "PHD3",
      "Bromodomain",
      "PHD4",
      "FYRN",
      "TAD",
      "FYRC",
      "WIN",
      "SET"
    ],
    "domains_retained_3": [
      "ENL central disordered region",
      "AHD"
    ],
    "domains_lost_3": [
      "YEATS domain (acetyl-lysine reader)"
    ],
    "phenotype_notes": "Downstream ENL junctions (exons 6-7) exclude the YEATS reader entirely, leaving a fusion whose partner contribution is purely the AHD docking hub - architecturally equivalent to KMT2A::MLLT3. Whether this changes phenotype or drug response has never been tested. The retained KMT2A half also contains the LEDGF/PSIP1-binding region (no counterpart lane in the parent gene-kmt2a domain map, so it is recorded here rather than in the domain arrays)."
  }
]
PDB IDS
[
  "5J9S",
  "2LM0",
  "7UJ4",
  "4GQ6"
]
mechanism class
[
  "chromatin-elongation-hijack"
]
drugs
[
  "drug-revumenib",
  "drug-ziftomenib",
  "drug-pinometostat"
]
key interactors
[
  "gene-aff1",
  "gene-aff4",
  "gene-dot1l",
  "gene-men1",
  "gene-psip1",
  "gene-paf1",
  "gene-hoxa9",
  "gene-meis1"
]
trials
[
  "NCT04065399"
]
How this could bite us back
  • Resistance: menin-inhibitor resistance via MEN1 binding-site mutation applies identically, because the drugged interface sits on the shared KMT2A half.
  • Compensation: ENL and AF9 are redundant paralogues with overlapping YEATS and AHD functions, so a selective ENL ligand is likely to be compensated by AF9 (and vice versa) - the family may need dual inhibition, with correspondingly worse selectivity.
  • Essential-function toxicity: the YEATS aromatic cage is conserved across ENL, AF9, GAS41 and YEATS2, all of which have essential roles in normal transcription and H2A.Z deposition; a pan-YEATS ligand would be broadly cytotoxic, and a selective one is chemically hard.
  • Clinical-trial hazard: because 1-9-year-old and T-ALL patients with this fusion already do very well on chemotherapy, testing a new targeted agent in them risks harm without headroom for benefit. The drug is needed in infants, where the cohort is small, the pharmacology hardest and the biology closest to KMT2A::AFF1.
  • Diagnostic delivery problem: conventional karyotyping missed t(11;19) in nearly half of confirmed cases in one series, so any partner-specific therapy depends on molecular detection that is not universally performed.
Still open
  1. How much of the ENL YEATS fold is actually present in the fusion protein in each patient? ENL junctions at exons 2, 4, 6 and 7 should produce chimeras that either retain or lack the acetyl-lysine reader, which would make 'KMT2A::MLLT1' at least two different oncoproteins. Nobody has stratified phenotype, HOXA signature or menin-inhibitor response by ENL junction.
  2. Why is MLLT1 lymphoid and T-lineage-skewed while its near-identical paralogue MLLT3 is myeloid? Is this partner-locus expression and 3D proximity in the cell of origin, or a real biochemical difference in YEATS specificity or AHD partner preference?
  3. The same fusion is lethal in infants and near-curable in 1-9-year-olds. If that difference is entirely cell-of-origin, then a faithful infant model of KMT2A::MLLT1 must begin in a fetal progenitor, and the good-prognosis cases tell us what a permissive-but-containable cellular context looks like. What, concretely, does the older child's progenitor have that the infant's lacks?
  4. Would a YEATS-domain ligand add anything on top of a menin inhibitor, or do both drugs simply block different steps of one linear pathway with identical escape routes?
Every source behind this entry (6)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    KMT2A::MLLT1 accounts for 460/3401 (13.5%) of genomically characterised KMT2A rearrangements, about 18.5% of KMT2A-rearranged ALL and 24% of infant ALL, is the leading KMT2A partner in T-ALL (n=43), and prefers KMT2A intron 11 constitutively rather than as a function of age or therapy.

    The KMT2A recombinome of acute leukemias in 2023. · Meyer C, Larghero P, Almeida Lopes B, et al. · Leukemia 2023
    PMID 37019990DOI
    What this evidence cannot show

    Referral-based single-centre cohort; the T-ALL subcohort is small in absolute terms, so 'leading T-ALL partner' rests on tens rather than hundreds of cases. Percentages are of cases with an identified partner.

    Evidence that argues against this (1)
    • MLLT1's share is inflated by how it is detected, and this graph holds the number that shows it. About half of KMT2A::MLLT1 cases - 241 of 460 - are SPLICED fusions in which the break lands upstream of an intact partner, and those are recoverable only when RNA is examined. Centres doing RNA-level characterisation therefore find proportionally more MLLT1, so the 13.5% figure is partly a statement about assay mix across contributing centres rather than about disease frequency. Leukemia 2023
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    MLL/ENL fusion transcripts use ENL breakpoints at exons 2, 4, 6 or 7, all in frame, and in 15 patients with MLL/ENL the phenotype split 7 precursor-B ALL, 4 T-ALL and 4 AML.

    Analysis of acute leukemias with MLL/ENL fusion transcripts: identification of two novel breakpoints in ENL · Fu JF, Liang DC, Shih LY · American Journal of Clinical Pathology 2007
    PMID 17145626DOI
    What this evidence cannot show

    Only 15 MLL/ENL patients from a single Taiwanese series, detected by RT-PCR and cDNA panhandle PCR; conventional karyotyping missed t(11;19) in 6 of 13 cases, so ascertainment is method-dependent. The paper reports transcript breakpoints, not protein-level confirmation of which part of the YEATS fold survives, so the inference that YEATS retention varies between patients is ours, not the authors'.

    Evidence that argues against this (1)
    • Fifteen patients is too few to carry a phenotype distribution, and a series two orders of magnitude larger exists. Genomic breakpoint characterisation of 3401 acute leukaemia patients analysed between 2003 and 2022 catalogued 107 in-frame KMT2A fusions with resolved partner genes and breakpoints; against that denominator a 7 / 4 / 4 split across precursor-B ALL, T-ALL and AML has confidence intervals wide enough to be compatible with most plausible true distributions. The breakpoint observation (in-frame joins at MLLT1 exons 2, 4, 6 or 7) is the durable part of this item; the lineage proportions are not. Meyer C et al. 2023
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The ENL YEATS domain binds acetylated histone H3 and this acetyl-reading function sustains oncogenic gene expression in acute myeloid leukemia, making ENL a chromatin-reader dependency.

    PMID 28241141DOI
    What this evidence cannot show

    The dependency was established for wild-type ENL in AML cell lines and mouse models, including NPM1-mutant and KMT2A-rearranged contexts; it is not a study of the KMT2A::MLLT1 fusion protein, and it does not show that the fusion's own YEATS domain (where retained) is the relevant target.

    Evidence that argues against this (1)
    • Acetyl-reading is one of ENL's contributions and this entry treats it as the whole of them. ENL and its intrinsically disordered region are by themselves sufficient to initiate liquid droplet formation of the super elongation complex, AFF4 fluidises those droplets, the droplets assemble within minutes of a stimulus and are required for rapid transcriptional induction - and fusing ENL to MLL boosts SEC phase separation. So ENL brings a multivalent self-association property in addition to its YEATS pocket, and a therapeutic strategy aimed only at the acetyl-binding pocket leaves the condensate-nucleating function of the same protein in place. Which of the two the fusion actually needs has not been separated. Guo C et al. 2020
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Chemical-genetic displacement of the ENL YEATS domain collapses transcription at leukemia-dependency genes, establishing the YEATS pocket as a pharmacologically tractable node.

    Transcription control by the ENL YEATS domain in acute leukaemia · Erb MA, Scott TG, Li BE, et al. · Nature 2017
    PMID 28241139DOI
    What this evidence cannot show

    Uses an engineered degradation/displacement system rather than a drug-like ligand, so tractability is demonstrated in principle only; YEATS is a shallow aromatic-cage pocket and no clinical-stage YEATS inhibitor exists. Conducted in cell lines.

    Evidence that argues against this (1)
    • 'Pharmacologically tractable' is doing a lot of work here, because the displacement was chemical-genetic. The nearest clinical test in the same pathway is the DOT1L inhibitor pinometostat: 51 adults treated, pharmacodynamic reduction of H3K79 methylation achieved, maximum tolerated dose never reached, and two complete remissions - both, as it happens, in t(11;19), which is this fusion's cytogenetic group. That is simultaneously the strongest partner-specific clinical signal in this literature and a demonstration that on-target epigenetic engagement in this chain converts into single-agent remission in about 4% of treated patients. A tool-compound displacement experiment establishes that the pocket can be occupied; it does not establish tractability in the sense the entry needs. Stein EM et al. 2018
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Among 35 children with t(11;19)(q23;p13.3) ALL and confirmed MLL-ENL transcripts, only 2 of 13 infants remained in complete remission versus 6 of 11 older children, and no relapses occurred among B-precursor patients aged 1-9 years or among T-cell patients.

    Childhood acute lymphoblastic leukemia with the MLL-ENL fusion and t(11;19)(q23;p13.3) translocation · Rubnitz JE, Camitta BM, Mahmoud H, et al. · Journal of Clinical Oncology 1999
    PMID 10458233DOI
    What this evidence cannot show

    Retrospective, n=35 total with 11-13 patients per subgroup, and 'no relapses' in tiny subgroups has very wide confidence intervals. Therapy dates from the 1980s-90s and pre-dates MRD-directed stratification, so the absolute cure fractions are not current. Cytogenetic case-finding will have missed cryptic t(11;19).

    Evidence that argues against this (1)
    • Later large series show KMT2A-r infant ALL outcome is poor regardless of partner, so the favourable MLLT1 signal is specific to non-infant age groups and should not be generalised to the infant population where the therapeutic need lies. Pieters R et al. 2019
  6. E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Revumenib was studied and approved in KMT2A-translocated relapsed/refractory acute leukemia without restriction by partner gene, achieving CR+CRh 22.8% and ORR 63.2%.

    Menin Inhibition With Revumenib for KMT2A-Rearranged Relapsed or Refractory Acute Leukemia (AUGMENT-101) · Issa GC, Aldoss I, Thirman MJ, et al. · Journal of Clinical Oncology 2025
    PMID 39121437DOINCT04065399
    What this evidence cannot show

    No partner-stratified efficacy analysis, so activity specifically in KMT2A::MLLT1 is inferred from the shared retained MBM module and the partner-agnostic eligibility criteria.

    Evidence that argues against this (1)
    • The label is partner-agnostic; the evidence for partner-agnostic BENEFIT is thin, and the field says so. Current syntheses of menin-inhibitor development list unresolved challenges in patient selection and sequencing alongside emerging resistance mechanisms, and note that the dependency extends beyond KMT2A rearrangement to NPM1 mutation, NUP98 rearrangement and UBTF tandem duplication - i.e. that the drug tracks a HOX-driven transcriptional state rather than any particular fusion architecture. An entry for one partner should not read the class approval as evidence about that partner. Cited as a map of the current development landscape rather than as primary evidence. Bruzzese A et al. 2026
Connected to