About one KMT2A-rearranged patient in twenty has no direct fusion at all - only the reciprocal one
also called reciprocal-only patients · 114 of 2454 · no MLL-X allele · 5 percent reciprocal only
In a small but real group of patients the famous fusion gene simply is not there. The only KMT2A fusion they have is the mirror-image one - and that one has no handle for the menin drug to grab.
The same thing, for a scientist
The Frankfurt DCAL reports that 114 of 2,454 KMT2A-rearranged patients - about 4.6% - were diagnosed only with the reciprocal fusion allele, with no KMT2A-X allele present [E1]. In those patients the entire menin-interface model has no substrate: the menin-binding motif lives on KMT2A exon 1, which is on the direct product, and there is no direct product. This is not a hypothetical population and it is not tiny; at 4.6% of KMT2A-rearranged acute leukaemia it is comparable in size to several partner classes that carry dedicated entries in this graph. It is also almost completely uncharacterised: the number appears in one sentence of one review and this beat could find no report of which partners are involved, which diseases they presented as, how they were treated or how they did. Two adjacent facts sharpen it. Ten of 102 KMT2A fusion genes are per se out-of-frame, so a direct product can be absent for structural reasons as well as by deletion [E1]. And NUP98::KMT2A, an inversion in which KMT2A is the 3' partner, deletes KMT2A exon 1 and is reported resistant to menin inhibition with reduced rather than elevated HOXA output [E2] - a worked example of what a menin-blind KMT2A oncoprotein looks like clinically.
Why this is the sharpest single fact in the beat
Most of the reciprocal-product argument is about a second protein alongside the first, where one can always argue the first is what matters. Here there is no first. If KMT2A-rearranged leukaemia can be driven by a KMT2A fusion with no menin-binding motif in 4.6% of patients, then menin binding is not necessary for KMT2A-fusion leukaemogenesis in general - it is necessary for the common architecture. That is a weaker and more accurate claim than the one the field makes, and it has a natural clinical corollary: those patients should not be expected to respond to a menin inhibitor, and nothing in any label or eligibility criterion identifies them.
What would have to be true for this to be an artefact
Three ways the number could be wrong, all worth checking before it is leaned on: the direct allele exists but is not detected by that centre's assay panel (the mirror of the RT-PCR blindness documented on the reciprocal side); the patients are a referral artefact, sent to a specialist centre precisely because routine typing failed; or the reciprocal allele is present and inert and the leukaemia is driven by something else entirely, with the KMT2A rearrangement a passenger. None of these is excluded by a single-sentence count in a review, and the entry's confidence is set accordingly.
| n reciprocal only | 114 |
|---|---|
| denominator | 2454 |
| PCT | 4.6 |
| characterisation available | none beyond the count |
| menin binding motif location | KMT2A exon 1 (residues 6-25), on the direct product only |
out of frame direct fusions
{
"n": 10,
"of": 102
}- A menin inhibitor licensed on KMT2A-rearrangement status would be given to these patients with no molecular basis for a response, and no assay in routine use would flag them.
- If reciprocal-only disease responds to menin inhibition anyway, the menin-interface model is wrong in a way nothing else in this graph could detect.
- Counting 4.6% from one referral centre is the kind of number that moves when a second centre counts it, and no second centre has.
- Which is it: 4.6% or 0.47%? The two counts come from one archive three years apart, the later one is worded as a detection failure, and no paper reconciles them. Every use of the reciprocal-only population as an argument depends on the answer.
- Which partner genes, which diseases and which outcomes are represented among the 114 reciprocal-only patients? The count exists; the description does not.
- Has any reciprocal-only patient ever been treated with a menin inhibitor? The architecture predicts no response and there is no case report either way.
- Is the absent direct allele truly absent, or absent from that centre's assay panel? Whole-genome sequencing on a handful of these cases would settle it.
- Do the 10 out-of-frame direct fusions overlap with the 114 reciprocal-only patients, or are they separate populations?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In the Frankfurt DCAL series, 114 of 2,454 patients (about 5%) were diagnosed only with the reciprocal fusion allele and displayed no MLL-X allele; of 102 fusion genes, 10 were per se out-of-frame; balanced translocations account for about 70% of rearrangements and a reciprocal fusion gene is generated in 94% of cases.
The reciprocal world of MLL fusions: A personal view. · Marschalek R · Biochim Biophys Acta Gene Regul Mech 2020What this evidence cannot show
A single diagnostic centre's referral series, framed by its author as a personal view; the 114 patients are counted but not otherwise described - no partner breakdown, no disease distribution, no outcome. Referral bias is the obvious alternative explanation and is not addressed.
Evidence that argues against this (1)
- The same laboratory's t(4;11) genomic work shows that patients scored as lacking one allele frequently carry that allele fused to an unexpected partner, so an assay-based absence in this series should be treated with the same suspicion in the other direction. Kowarz E et al. 2007
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
NUP98::KMT2A, arising from inv(11)(p15q23) and deleting the KMT2A exon that encodes the menin-binding region while retaining the WIN motif and SET domain, is transforming in an inducible transgenic mouse and is reported resistant to menin inhibition, with reduced rather than elevated HOXA output.
Transforming activities of the NUP98-KMT2A fusion gene associated with myelodysplasia and acute myeloid leukemia. · Fisher JN, Thanasopoulou A, Juge S, et al. · Haematologica 2020What this evidence cannot show
A murine inducible transgenic with a median latency of about 80 weeks; menin-inhibitor resistance is reported with the tool compound MI-2-2, not a clinical agent, and this is a single rearrangement rather than a class.
Evidence that argues against this (1)
- A menin-blind architecture that is transforming does not establish that menin-blind disease is common or that it behaves the same way in humans; the direct-fusion architecture remains the overwhelming majority and the clinical responses to menin inhibitors were obtained in it. Ozyerli-Goknar E et al. 2026
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Read in full text (PMC8530991): the reciprocal t(4;11) fusion gene is AF4 exons 1-3 fused to MLL exons 12-37, while the direct fusion is MLL exons 1-10 fused to AF4 exons 4-20, and KMT2A breakpoints cluster in introns 9-11.
The role of reciprocal fusions in MLL-r acute leukemia: studying the chromosomal translocation t(4;11). · Wilhelm A, Marschalek R · Oncogene 2021What this evidence cannot show
Read in full text via Europe PMC fullTextXML (40,194 characters), not abstract only. The exon composition describes that laboratory's constructs, chosen to model the infant intron-11 breakpoint. This entry uses it only for the exon arithmetic; the further step - that a product starting at KMT2A exon 12 cannot carry the menin-binding motif encoded at residues 6-25 - is an architectural inference from UniProt coordinates, not a statement in this paper, and no study has shown directly that menin fails to bind a reciprocal KMT2A product.
Evidence that argues against this (1)
- No menin co-immunoprecipitation on any reciprocal KMT2A fusion protein has been published, so the absence of menin binding is an inference from architecture, not a measurement. Marschalek R 2020
- E-later-cohortmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.journal-article
The same laboratory's later and larger cohort reports a reciprocal-only rate about ten times lower. In 3,401 patients analysed for the 2023 recombinome, the reciprocal-only category is 16 patients - 0.47% - against 114 of 2,454 (4.6%) in the 2020 review.
The KMT2A recombinome of acute leukemias in 2023. · Leukemia 2023What this evidence cannot show
Verified 2026-09-19 against both sources. The 2020 abstract states verbatim: *"So far, 114 patients (out of 2454 = ~5%) have been diagnosed only with the reciprocal fusion allele, displaying no MLL-X allele."* The 2023 full text (PMC10169636) states: *"16 cases in which no direct KMT2A fusion but only the reciprocal KMT2A fusion could be detected"* and *"16 patients where a 5'-KMT2A fusion could not be identified, but with the presence of reciprocal fusion allele"*. **These may not be the same denominator.** The 2023 figure is enumerated inside the complex-rearrangement analysis (n = 426 of 3,401) and the two papers draw on overlapping patients from one archive, so this is not two independent estimates. Neither paper cross-references the other's count.
Evidence that argues against this (2)
- **The most economical explanation is the one that damages this entry most: 'reciprocal-only' is partly an assay-failure category, and it shrinks as the assay improves.** The 2023 wording is explicitly about detection - the direct fusion *could not be identified* - not about absence. Three more years of long-distance inverse PCR and RNA-seq on the same archive would then convert most of the 114 into ordinary two-allele patients. If that is right, the refutation this entry carries weakens by an order of magnitude: the population of patients whose leukaemia has no menin-binding fusion at all may be well under 1%, not 5%. Leukemia 2023
- Against that reading, the 2020 review frames its 114 as a positive finding worth discussing rather than a residual, and the 2023 paper never retracts or mentions it. Until someone states the denominators side by side, both numbers should be quoted together and neither alone. Biochimica et Biophysica Acta - Gene Regulatory Mechanisms 2020
The population in which the reciprocal product is the only KMT2A fusion present.
- confers resistance to →MEN1 / menin — the adaptor the fusion cannot work without
No menin-binding motif is present in these patients at all, because the allele carrying it is absent.
- confers resistance to →Revumenib (SNDX-5613, Revuforj) — first approved menin–KMT2A interface inhibitor
Predicted non-response on architectural grounds; never tested, and no eligibility criterion identifies these patients.
Adds a third and larger class of menin-blind KMT2A oncoprotein: patients who have no direct fusion at all.
The 114/2454 figure is one row of that entry's frequency accounting.
Supplies a functional rationale for why a reciprocal-only patient could have leukaemia at all.