Oncofusion

start / graph / ana-menin-dependence-untested-in-septin-and-cytoskeletal-kmt2a-fusions

Analysisstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q-nonelongation-partners

Menin dependence of a non-elongation KMT2A fusion has never been demonstrated — the counted absence, and what the pivotal trial actually enrolled

also called menin partner-independence gap · septin fusion menin dependence

Menin drugs are given to patients because their leukaemia is 'KMT2A-rearranged', on the reasoning that menin sticks to the KMT2A half, which every fusion keeps. Nobody has ever tested that in a leukaemia whose partner is a cytoskeletal protein, and the registrational trial did not knowingly enrol a single such patient.

The same thing, for a scientist

The menin thesis is partner-independent by construction: menin binds high-affinity motifs in the conserved amino-terminus of MLL that every oncogenic fusion retains, and that interaction is required for initiation and maintenance of MLL-associated leukemogenesis [E1]; the required MLL sequences lie between residues 5 and 44 [E2]; and the structure shows a single menin pocket binding the MLL peptide [E3]. From that, partner-independence is an inference about a binding site, not a measurement across partner classes. This analysis records how far the measurement actually goes. **PubMed [Title/Abstract] intersection of 'menin' (1264 records) with 'septin OR septins' (1936 records) returns exactly 0 records. The intersection of menin-inhibitor drug names with septin terms returns 0 after noise removal. The intersection of (MLL OR KMT2A) AND septin AND (menin OR MEN1 OR DOT1L OR H3K79 OR super elongation) returns 0** [E4]. No septin-partner fusion has been tested for menin dependence, for DOT1L dependence, for SEC recruitment or for H3K79 methylation, by any method, ever. Clinically, the pivotal AUGMENT-101 KMT2A-r cohort reports its efficacy population (n=57) by translocation: t(11;19) 13 (22.8%), t(9;11) 11 (19.3%), t(10;11) 7 (12.3%), t(6;11) 7 (12.3%), t(4;11) 2 (3.5%), t(1;11) 2 (3.5%), fusion partner unknown 13 (22.8%), other translocations 2 (3.5%) [E5]. **No septin translocation appears** — t(X;11), t(11;17)(q23;q25), t(11;22) and t(2;11) are all absent from the table, so at most they hide inside 'other' (n=2) or 'unknown' (n=13).

Go deeper
What partner-independence actually rests on

Three results, none of which involves a non-elongation partner explicitly in its abstract.

  1. Oncogenic MLL fusion proteins retain the ability to associate stably with menin through a high-affinity amino-terminal conserved binding motif, and that interaction is required for initiation of MLL-mediated leukemogenesis; menin is essential for maintenance of MLL-associated but not other oncogene-induced myeloid transformation, and acute genetic ablation reverses aberrant Hox expression and abrogates the differentiation arrest 1.
  2. MLL sequences between amino acids 5 and 44 are required for interaction with menin and for transformation of haematopoietic progenitors, and a dominant-negative N-terminal MLL fragment down-regulates Meis1 and inhibits proliferation 2.
  3. The crystal structure shows menin's deep pocket binding a short MLL1 peptide, with LEDGF binding at a distinct composite surface 3.

Each of these is about the retained KMT2A N-terminus. The inference — therefore every fusion is menin-dependent — is strong and reasonable. It is still an inference. The partners actually used in the foundational work are elongation partners, and neither the abstracts nor the structural work extend to a septin, a cytoskeletal adaptor or any other non-elongation partner.

The counted absence

Every count below was run on 2026-09-19 with both halves of the intersection controlled, so a zero can be read as a real absence rather than a dead query 4:

queryrecords
menin[Title/Abstract]1264
septin OR septins [Title/Abstract]1936
menin AND (septin OR septins)0
menin AND (SEPT6 OR SEPT9 OR SEPT5 OR SEPT11 OR SEPT2)0
(revumenib OR ziftomenib OR VTP-50469 OR MI-503 OR MI-2 OR 'menin inhibitor') AND septin terms0 after removing 19 wildcard artefacts (SEPT* matches 'September')
(MLL OR KMT2A) AND septin AND (menin OR MEN1 OR DOT1L OR H3K79 OR 'super elongation' OR SEC)0

Both halves have large literatures. The intersection is empty. This is a genuine, countable absence and not a search artefact, and the wildcard check was performed by reading the titles of all 19 apparent hits, which were about differentiation syndrome, dermatomyositis, lupus and dairy farming.

The closest thing to a demonstration: AFDN in AUGMENT-101

One non-elongation partner HAS been exposed to a menin inhibitor in a registrational setting. t(6;11)(q27;q23) — KMT2A::AFDN, whose transforming element is RA1-mediated self-association rather than an elongation module — comprised 7 of 57 patients (12.3%) in the AUGMENT-101 efficacy population and 8 of 94 (8.5%) in the safety population 5. The authors report that efficacy was observed across subgroups including KMT2A translocation partner, in an exploratory analysis with limited sample sizes whose per-partner data sit in a Data Supplement table.

This matters and should not be overstated. It is the best available evidence that a pure self-association partner responds to menin inhibition, and it is clinical rather than mechanistic, exploratory, not powered, and not resolved per partner in the main text. A case report of a KMT2A::AFDN patient achieving MRD-negative remission on revumenib plus venetoclax and azacitidine with donor lymphocyte infusion adds an anecdote in the same direction and cannot be attributed to the menin inhibitor 6.

So the honest position: menin dependence is demonstrated mechanistically for elongation partners, supported clinically and exploratorily for one self-association partner (AFDN), and entirely untested for the septin class and for ABI1.

Why the gap persists and what closes it

Three practical reasons. There is no KMT2A-septin-fusion cell line in DepMap or in the standard leukaemia panels. The one usable reagent that exists is HF6, an IL-3-dependent mouse line immortalised by MLL/SEPT6 and used as a background for cooperating-lesion screens 7. And the trials report by cytogenetic translocation rather than by molecular fusion, which means a septin patient enrolled as 'KMT2A-rearranged' would be counted in 'other' or 'unknown' and never surface.

The experiment that closes the gap is small: express KMT2A::SEPTIN6 (and a Δcoil control) in mouse progenitors, treat with a menin inhibitor, read colony formation and Hoxa9/Meis1. See exp-menin-inhibitor-sensitivity-of-kmt2a-septin-fusions.

Facts and figures
scriptnot a script; E-utilities queries and PMC full-text reads performed interactively on 2026-09-19
statusrun
run date2026-09-19
absence claimNo KMT2A-septin fusion, and no KMT2A-ABI1 fusion, has ever been tested for menin dependence, DOT1L dependence, super-elongation-complex recruitment or H3K79 methylation, by any method.
absence is countedtrue
verdictMenin dependence is demonstrated mechanistically for elongation partners, supported clinically and exploratorily for one self-association partner (AFDN, 12.3% of the pivotal efficacy population), and entirely untested for the septin class and for ABI1. The partner-independence claim is an inference from a binding site, and it is a good inference, but it is not a measurement.
counts
menin_AND_septin0.000menin_AND_septin: 0menin_AND_individual_SEPT…0.000menin_AND_individual_SEPT_symbols: 0menin_inhibitor_names_AND…0.000menin_inhibitor_names_AND_septin_after_noise_removal: 0mll_AND_septin_AND_mechan…0.000mll_AND_septin_AND_mechanism_terms: 0menin_TIAB1,264menin_TIAB: 1264septin_TIAB1,936septin_TIAB: 1936
augment101 efficacy population by translocation
septin_translocations_nam…0.000septin_translocations_named: 0t(4;11)2.000t(4;11): 2t(1;11)2.000t(1;11): 2other_translocations2.000other_translocations: 2t(10;11)7.000t(10;11): 7t(6;11)7.000t(6;11): 7t(9;11)11.0t(9;11): 11t(11;19)13.0t(11;19): 13fusion_partner_unknown13.0fusion_partner_unknown: 13n57.0n: 57
counts — the numbers behind the chart
{
  "menin_TIAB": 1264,
  "septin_TIAB": 1936,
  "menin_AND_septin": 0,
  "menin_AND_individual_SEPT_symbols": 0,
  "menin_inhibitor_names_AND_septin_after_noise_removal": 0,
  "mll_AND_septin_AND_mechanism_terms": 0
}
augment101 efficacy population by translocation — the numbers behind the chart
{
  "n": 57,
  "t(11;19)": 13,
  "t(9;11)": 11,
  "t(10;11)": 7,
  "t(6;11)": 7,
  "t(4;11)": 2,
  "t(1;11)": 2,
  "fusion_partner_unknown": 13,
  "other_translocations": 2,
  "septin_translocations_named": 0
}
experiments
[
  "exp-menin-inhibitor-sensitivity-of-kmt2a-septin-fusions"
]
How this could bite us back
  • The clinical exposure is running ahead of the evidence in a specific, checkable way: revumenib is approved for KMT2A-rearranged acute leukaemia without partner restriction, so septin-partner patients are eligible on a mechanism that has never been demonstrated in their fusion class.
  • If a septin-partner patient fails menin inhibition, the failure will be recorded as primary resistance in a KMT2A-rearranged patient, and the partner-class hypothesis will remain untested for another cycle. Recording the partner molecularly at enrolment is the cheap fix.
  • Conversely, if the graph's target strategy is reframed around the retained KMT2A N-terminus rather than the elongation machinery, menin becomes MORE central rather than less — but only if this gap is closed, because the reframing's whole point is that partner class does not matter, and that is the part nobody has checked.
Still open
  1. Is KMT2A::SEPTIN6 menin-dependent? One colony assay with a menin inhibitor and a genetic Men1 control would answer it.
  2. What do the AUGMENT-101 Data Supplement per-partner response rates show for t(6;11)? If AFDN patients responded at the cohort rate, that is the strongest existing evidence for partner-independence and it is currently invisible outside a supplement.
  3. Should trials report KMT2A partner molecularly rather than cytogenetically? A quarter of the pivotal efficacy population had an unknown partner, which makes every partner-stratified question unanswerable after the fact.
  4. Does oligomerisation order modulate menin dependence? If menin occupancy scales with local fusion concentration, a filament-forming partner and a dimerising partner could differ in menin-inhibitor sensitivity without differing in menin binding.
Every source behind this entry (7)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Oncogenic MLL fusion proteins retain an ability to associate stably with menin through a high-affinity, amino-terminal, conserved binding motif, this interaction is required for the initiation of MLL-mediated leukemogenesis, menin is essential for maintenance of MLL-associated but not other oncogene-induced myeloid transformation, and acute genetic ablation of menin reverses aberrant Hox gene expression and abrogates the differentiation arrest and oncogenic properties of MLL-transformed blasts.

    PMID 16239140DOI
    What this evidence cannot show

    The abstract does not state which fusion proteins were used, and the full text was not accessible for this entry (the article is not in PMC). This entry therefore does NOT assert that any particular partner class was tested, and the reader should not infer it.

    Evidence that argues against this (1)
    • Menin dependence is not universal even among KMT2A-related leukaemias in the way a partner-independent reading implies: in vitro responses to revumenib differ qualitatively between KMT2A-rearranged ALL and AML, with ALL showing rapid dose-dependent cell death and AML a much slower response with myeloid differentiation, so lineage — and plausibly partner — modifies the dependency. Adriaanse FRS et al. 2024
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    MLL sequences between amino acids 5 and 44 are required for interaction with menin and for transformation of haematopoietic progenitors, and expression of a dominant-negative inhibitor composed of amino-terminal MLL sequences down-regulates Meis1 and inhibits cell proliferation.

    Interaction of MLL amino terminal sequences with menin is required for transformation. · Caslini C, Yang Z, El-Osta M, et al. · Cancer Res 2007
    PMID 17671196DOI
    What this evidence cannot show

    Maps the requirement on the KMT2A side. The fusion partners used are not specified in the abstract and the full text was not consulted for this entry.

    Evidence that argues against this (1)
    • A requirement mapped on the KMT2A side cannot establish that the requirement holds regardless of what is attached on the other side: the partner determines oligomerisation order, and if menin occupancy at target loci depends on local fusion concentration then partner class could modulate the dependency without changing the binding motif. Ono R et al. 2005
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Crystal structures of human menin free and in complex with MLL1, with JUND, and with an MLL1-LEDGF heterodimer show that menin contains a deep pocket binding short MLL1 or JUND peptides in the same manner, and that menin promotes transcription by binding MLL1 through the peptide pocket while interacting with LEDGF at a distinct surface formed by both menin and MLL1.

    PMID 22327296DOI
    What this evidence cannot show

    Structures of menin with short MLL1 peptides, not with any fusion protein. No structure of menin bound to a full KMT2A fusion exists, so the structural argument for partner-independence is an argument about a peptide.

    Evidence that argues against this (1)
    • The same pocket binds JUND with opposite transcriptional consequences, which shows the pocket's occupancy does not determine the output; by the same logic, demonstrating that a septin fusion binds menin would not by itself demonstrate that it depends on menin. Huang J et al. 2012
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    PubMed [Title/Abstract] searches run on 2026-09-19 return 1264 records for menin and 1936 for septin OR septins, but exactly 0 for their intersection; 0 for menin intersected with the individual symbols SEPT2, SEPT5, SEPT6, SEPT9 and SEPT11; 0 for menin-inhibitor drug names intersected with septin terms once 19 wildcard artefacts are removed by title inspection; and 0 for (MLL OR KMT2A) AND septin AND (menin OR MEN1 OR DOT1L OR H3K79 OR super elongation OR SEC).

    What this evidence cannot show

    Title/abstract scoping. A menin-inhibitor experiment on a septin fusion buried in a supplementary figure of a paper whose title and abstract mention neither term would not be found. The search also cannot see conference abstracts or unpublished negative results, and a negative result here is exactly the kind that goes unpublished.

    Evidence that argues against this (1)
    • The zero is about the intersection of two named topics, not about whether septin-fusion patients have received menin inhibitors. Registrational trials enrol on 'KMT2A-rearranged' status without molecular partner typing for a quarter of patients, so septin-partner patients may already have been treated and simply not identified as such. Issa GC et al. 2025
  5. E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    In AUGMENT-101, the revumenib KMT2A-rearranged efficacy population (n=57, centrally confirmed KMT2A-r with at least 5% marrow blasts) comprised t(11;19) 13 (22.8%), t(9;11) 11 (19.3%), t(10;11) 7 (12.3%), t(6;11) 7 (12.3%), t(4;11) 2 (3.5%), t(1;11) 2 (3.5%, additional information not reported), fusion partner unknown 13 (22.8%) and other translocations 2 (3.5%); the primary endpoint CR+CRh was 22.8% and the authors report that efficacy was observed across subgroups including KMT2A translocation partner in an exploratory analysis with limited sample sizes.

    PMID 39121437DOI
    What this evidence cannot show

    Read from the PMC full text (PMC11687943) for this entry. The per-partner response data are in a Data Supplement table that was not accessible here, so this entry states the enrolment composition and the authors' qualitative statement, not per-partner response rates. No septin translocation is named anywhere in the paper.

    Evidence that argues against this (1)
    • A quarter of the efficacy population had an unknown fusion partner and a further 3.5% were 'other translocations', so the absence of a named septin translocation does not prove no septin-partner patient was treated — it proves the trial could not have known if one was. Meyer C et al. 2023
  6. E6weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper

    A 32-year-old with KMT2A::AFDN acute monocytic leukaemia relapsing with CNS involvement after allogeneic transplant achieved MRD-negative status in marrow and cerebrospinal fluid with sustained complete remission after one course of revumenib combined with venetoclax and azacitidine followed by donor lymphocyte infusion.

    PMID 42643519DOI
    What this evidence cannot show

    A single case with four simultaneous interventions (revumenib, venetoclax, azacitidine, DLI). No component can be credited. Cited only as the sole partner-identified clinical observation for a non-elongation KMT2A fusion.

    Evidence that argues against this (1)
    • Four concurrent active therapies including donor lymphocyte infusion make this uninformative about menin dependence; the registrational monotherapy data, which do include t(6;11) patients, are the only evidence with any weight and are exploratory. Issa GC et al. 2025
  7. E7moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    HF6, an IL-3-dependent mouse bone-marrow-derived cell line immortalised by the class II mutation MLL/SEPT6 and fully transformable by class I mutations such as FLT3 mutants, exists and has been used as a screening background for cooperating lesions.

    Possible involvement of RasGRP4 in leukemogenesis. · Watanabe-Okochi N, Oki T, Komeno Y, et al. · Int J Hematol 2009
    PMID 19350351DOI
    What this evidence cannot show

    The line is described in the context of an expression-cloning screen for cooperating mutations, not as a characterised reagent; its current availability is not established here.

    Evidence that argues against this (1)
    • A mouse IL-3-dependent line immortalised by a fusion is a weak surrogate for human leukaemia and a poor setting in which to measure a menin-inhibitor response, because growth-factor dependence can mask or mimic a transcriptional dependency; a primary-progenitor colony assay with a genetic menin control would be more interpretable. Yokoyama A et al. 2005
Connected to