Raising MHC class I is not a free good: the same molecule is the inhibitory ligand that holds NK cells off, and in this fusion family it is an in vivo leukaemia dependency
also called class I upregulation trade-off · T cell versus NK cell trade-off under menin inhibition · missing-self cost of antigen presentation
Turning up a leukaemia cell's identity cards makes it easier for one kind of immune killer to find, and harder for another kind. The second kind is the one that matters most after a mismatched transplant, which is exactly the setting where this drug is being proposed.
The same thing, for a scientist
Every proposal in this beat - menin inhibition, EZH2 inhibition, interferon - raises MHC class I in order to make blasts visible to T cells. **The same molecules are the ligands for inhibitory KIR (human) and Ly49 (murine) receptors**, and in this fusion family the evidence that the inhibitory arm is load-bearing is stronger than the evidence that the T-cell arm is. **H2-K1, a classical class I molecule, scored among the top in vivo dependencies of MLL-AF9 leukaemia-initiating cells**; knocking it down depleted leukaemia and extended survival, and **NK-cell depletion rescued the leukaemia completely** [E1]. Three KMT2A-rearranged lines are NK-resistant with inhibitory KIR ligands present and activating ligands low, and **blocking class I increases NK degranulation** [E2]. Against that, the T-cell benefit of raising class I in this disease is entirely extrapolated: the menin-inhibitor class I experiments are in CML, SCLC, neuroblastoma and NPM1c AML [E3][E4]. **The trade is being proposed on the strength of the weaker half.** The quantity that decides it - whether the inhibitory brake is already saturated at baseline class I levels - has never been measured.
The setting where the conflict is sharpest
Both leading proposals for menin-inhibitor immunology put the drug after allogeneic transplant. cell-state-interferon-herv-mhc2 records Fetsch's proposal of maintenance after allo-HCT 4; the bleximenib paper's functional arm is allogeneic T-cell killing 3.
Haploidentical transplantation is, in large part, a deliberate missing-self strategy: KIR-ligand mismatch between donor and recipient reduces relapse risk in AML, which is the reason to choose a mismatched donor at all 1. A drug given as post-transplant maintenance that raises the recipient blast's class I is raising the inhibitory ligand that the deliberate mismatch was engineered to avoid.
So the two most-discussed immunological uses of menin inhibition in this disease - GVL enhancement through antigen presentation, and NK-mediated missing-self control after a mismatched graft - are in mechanistic conflict, in the same patient, at the same time. This graph could find no source that states the conflict, and no experiment that measures which effect is larger.
What would resolve it, and why nobody has done it
The resolving measurement is a dose-response, and every published experiment is a loss-of-function. Chan blocks class I; Ghosh knocks it out; Nakata removes NK cells. None raises class I and asks what NK cells do.
Two possibilities, with opposite conclusions:
- Saturated. If baseline class I on a KMT2A-rearranged blast already engages inhibitory KIR maximally, then raising it adds nothing on the NK side and the menin-inhibitor immune rationale is clean. This is plausible - inhibitory receptor signalling is generally considered to have a low threshold.
- Unsaturated. If degranulation still falls as class I rises, the drug trades a demonstrated NK brake for an extrapolated T-cell benefit.
The experiment is small: titrate class I on a KMT2A-rearranged line with interferon-gamma or a menin inhibitor, measure NK degranulation and killing across the range with KIR-genotyped donor NK cells. It is written as exp-mhc-class-i-and-nk-ligand-timecourse-under-menin-inhibition-kmt2a-r.
Three more edges of the same trade
Graft-versus-host disease. Raising class I and class II raises presentation of every self-peptide, not only leukaemia peptides. cell-state-interferon-herv-mhc2 already records that the GVL mechanism has no published separation from a GVHD mechanism, and the class I arm adds the CD8 compartment to that concern 4.
Post-transplant immunosuppression. Calcineurin inhibitors and post-transplant cyclophosphamide suppress exactly the T-cell effector functions the benefit requires, while doing comparatively less to the NK inhibition the cost operates through. The direction of the net effect under standard GVHD prophylaxis is therefore worse than the in vitro data suggest, not better.
Downstream of recognition. Protease inhibitor-9 was reported on all three KMT2A-rearranged lines 2. It neutralises granzyme B inside the target cell and so defeats NK and T-cell killing alike, after recognition has already succeeded. Raising the display does nothing about it, and nobody has measured it in this disease since 2012.
| kind | essential-function |
|---|---|
| severity | unquantified - the direction is established and the magnitude has never been measured |
| mechanism | MHC class I is simultaneously the T-cell recognition element and the ligand for inhibitory KIR/Ly49 receptors; raising it strengthens both |
| deciding measurement | class I dose-response of NK degranulation on a KMT2A-rearranged line - never performed |
| mitigation | pair class I induction with an MHC-independent effector (CAR, BiTE, ADCC) rather than relying on endogenous recognition; or measure saturation before assuming a cost |
| setting of maximal conflict | menin inhibitor as maintenance after KIR-ligand-mismatched haploidentical or cord blood transplant |
affects
[ "drug-revumenib", "drug-ziftomenib", "drug-bleximenib", "drug-tazemetostat" ]
evidence asymmetry
{
"NK_cost": "in vivo dependency in MLL-AF9 (PMID 39844759); NK resistance with inhibitory KIR ligands in three KMT2A::AFF1 lines (PMID 23014531)",
"T_cell_benefit": "extrapolated to KMT2A-rearranged disease from K-562/SCLC/neuroblastoma (PMID 36635503) and NPM1c AML (PMID 39704147)"
}- The strategy raises the ligand for an inhibitory receptor family in a disease where a class I molecule is a demonstrated in vivo leukaemia dependency.
- It is proposed for the post-transplant setting, which is simultaneously the setting with the most NK-dependent therapeutic mechanism and the most T-cell suppression from GVHD prophylaxis.
- The benefit is measured as mean MHC upregulation, which the source paper states does not predict killing, so the endpoint most likely to be adopted is the one shown not to correlate with the outcome.
- Both arms are defeated downstream by a granzyme-B inhibitor reported on all three index cell lines, which no current strategy addresses and nobody has re-measured in fourteen years.
- Is the NK inhibitory brake saturated at baseline class I on a KMT2A-rearranged blast? This single dose-response decides whether the trade-off is real or notional.
- In a KIR-ligand-mismatched transplant, does post-transplant menin inhibition reduce the missing-self effect the mismatch was chosen for? Measurable in a mouse model and never asked.
- Does protease inhibitor-9 / serpin B9 still mark KMT2A-rearranged blasts with modern reagents, and does it change under a menin inhibitor? A granzyme-B brake downstream of recognition would cap every strategy in this beat.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
H2-k1, a classical MHC class I molecule, scored among the top in vivo dependencies in a CRISPR dropout screen in MLL-AF9 leukaemia-initiating cells; its disruption had no ex vivo growth effect but strongly depleted leukaemia in bone marrow and spleen and extended survival in vivo; NK-cell depletion accelerated leukaemia and completely rescued the antileukaemic effect of H2-k1 knockdown while macrophage depletion did not; and the authors note that in AML, haematopoietic stem cell transplantation with KIR-mismatching of the donor reduces the risk of relapse.
H2-K1 protects murine MLL-AF9 leukemia stem cells from natural killer cell-mediated immune surveillance. · Ghosh S, Rodriguez-Zabala M, Dushime GT, Reinbach K, Ramakrishnan R, Sitnicka E, Jaras M · Haematologica 2025What this evidence cannot show
Murine MLL-AF9, myeloid. The KIR-mismatch statement is the authors' framing of the transplant literature, not a result of this paper. The experiment removes class I and therefore establishes that the brake exists; it says nothing about whether the brake can be tightened further. Read: full text via the PMC HTML page (PMC12130766), 2026-09-20, the PubMed abstract field being empty.
Evidence that argues against this (1)
- This is an immunocompetent-mouse result. Children receiving menin inhibitors are usually lymphopenic after intensive chemotherapy or immunosuppressed after transplant, and NK function in AML is independently documented as impaired, so the size of the NK brake in the real clinical setting is unknown and could be small enough that the trade does not arise. Ghosh S et al. 2025
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Three MLL-rearranged leukaemia cell lines (RS4;11, SEM, MV4-11) and primary blasts were resistant to NK-cell lysis, carried inhibitory KIR ligands and protease inhibitor-9 and expressed low NKG2D activating ligands and adhesion molecules; adding a pan-MHC class I antibody to block the inhibitory KIR-HLA interaction further augmented NK degranulation across all KIR2DL1, KIR2DL2/3 and KIR3DL1 subsets.
Antibody-dependent cell-mediated cytotoxicity overcomes NK cell resistance in MLL-rearranged leukemia expressing inhibitory KIR ligands but not activating ligands. · Chan WK, Kung Sutherland M, Li Y, Zalevsky J, Schell S, Leung W · Clin Cancer Res 2012What this evidence cannot show
In vitro degranulation and killing with a xenograft survival arm; all three lines are t(4;11) KMT2A::AFF1, so the genotype is narrow. 2012, with no modern replication found in this beat's searches. Blocking class I is the opposite manipulation from raising it, so this establishes the direction of the axis and not its dose-response. Read: abstract only; identifiers fetched 2026-09-20.
Evidence that argues against this (1)
- The paper's own preferred solution bypasses class I entirely - an antibody against CD19 or CD33 restored NK killing of all three lines through ADCC regardless of the class I state, with class I blockade as an add-on. If MHC-independent engagement works, the trade-off this risk entry describes can be sidestepped rather than resolved. Chan WK et al. 2012
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The functional immune benefit of a clinical-stage menin inhibitor was demonstrated as enhanced allogeneic and autologous T-cell cytotoxicity against primary AML blasts, in 4 of 8 samples in the allogeneic assay, all four of which carried NPM1c and DNMT3A mutations rather than a KMT2A rearrangement; and the strength of the T-cell response was not directly correlated with the level of MHC upregulation.
Bleximenib, the novel menin-KMT2A inhibitor JNJ-75276617, impairs long-term proliferation and immune evasion in acute myeloid leukemia. · Hogeling SM, Le DM, La Rose N, Kwon MC, Wierenga ATJ, Van den Heuvel FAJ, et al., Schuringa JJ · Haematologica 2025What this evidence cannot show
Cited here to establish the weakness of the benefit side of the trade in the index genotype: the T-cell benefit of raising MHC under a menin inhibitor has been demonstrated in NPM1c AML, not in KMT2A-rearranged disease, and not in any lymphoid or infant material. Read: full text (PMC12130779), 2026-09-20.
Evidence that argues against this (1)
- The absence of a KMT2A-rearranged arm is a gap in the evidence, not evidence of absence; MV4-11 and MOLM13, both KMT2A-rearranged, did show raised HLA-DR and enhanced killing by peripheral blood mononuclear cells in the same paper, which is a weaker but genuine KMT2A-rearranged signal. Hogeling SM et al. 2025
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin inhibition enhanced graft-versus-leukaemia effects through raised MHC class II and directly promoted donor T-cell effector function with increased TNF-alpha, interferon-gamma, perforin and granzyme production and reduced exhaustion, with the authors proposing menin inhibition as maintenance after allogeneic haematopoietic cell transplantation.
Menin inhibition enhances graft-versus-leukemia effects by T-cell activation and endogenous retrovirus induction in AML. · Fetsch V, Schwobel L, Ozyerli-Goknar E, Stell AV, Punta M, Plenge T, et al., Zeiser R · Blood 2026What this evidence cannot show
Class II, AML, murine and xenograft transplant models; read abstract only. Cited here for the proposed clinical setting, which is the setting in which the NK trade-off is sharpest. This graph's entry for the paper is `cell-state-interferon-herv-mhc2`.
Evidence that argues against this (1)
- The proposal is for the post-transplant setting, which is where KIR-ligand mismatch is deliberately engineered and where patients receive calcineurin inhibitors or post-transplant cyclophosphamide that suppress the T-cell functions the benefit requires. Both the cost and the practical obstacle are largest exactly where the drug is being proposed. Ghosh S et al. 2025
- confers resistance to →MHC class I is a bivalent gene set, and a menin inhibitor derepresses it - shown in K-562, SCLC and neuroblastoma, and measured on HLA-A in primary AML, but never once in a KMT2A-rearranged cell
the cost side of the intervention the mechanism entry describes
the NK brake is the mechanism the risk is built on
- confers resistance to →Revumenib (SNDX-5613, Revuforj) — first approved menin–KMT2A interface inhibitor
class effect; no revumenib-specific class I or NK experiment exists
- confers resistance to →Tazemetostat - an approved EZH2 inhibitor, and the drug this disease's own biology predicts will not work
the same trade applies to any EZH2 inhibitor used to restore MHC presentation
arm C measures NK degranulation as a function of achieved class I level
the same antigen-presentation increase is the substrate for graft-versus-host disease