Hypothesis: menin-inhibitor sensitivity falls with cell-cycle exit, so the compartment that causes relapse is the one the drug reaches least — and nobody has sorted by cycle state to check
also called quiescence blunts menin inhibition · cell-cycle-state hypothesis for menin inhibitors
Menin inhibitors work by letting a leukaemia cell mature instead of staying stuck as a blast. Maturing is something a cell has to actively do, and the cells that survive treatment are the ones that have stopped doing almost anything. The prediction is that these drugs are weakest against exactly the cells that come back, and the experiment to check it has never been run.
The same thing, for a scientist
Menin inhibition displaces menin from protein complexes, reduces chromatin occupancy of the fusion at target genes, and produces changes in gene expression, **differentiation** and apoptosis [E1]. Differentiation is an active transcriptional programme. The compartment that survives therapy and seeds relapse is quiescent: in MLL-rearranged AML, FOXM1 is required for LSC quiescence, survival and self-renewal through beta-catenin stabilisation [E2]; in ALL patient-derived xenografts, a rare dormant, treatment-resistant, stem-like subpopulation matches primary cells recovered at MRD, and its adverse properties are reversible on release from the niche [E3]. The hypothesis is that **ex vivo and in vivo menin-inhibitor sensitivity is a decreasing function of cell-cycle exit**, so that measured sensitivity in bulk blasts systematically overstates activity against the residual compartment. This graph could find no study that sorts KMT2A-rearranged blasts by cycle state before measuring menin-inhibitor sensitivity, which makes this a cheap and unclaimed experiment rather than a speculation about an area already covered.
The argument in three steps
- Menin inhibitors act through differentiation. VTP50469 displaced menin from protein complexes, inhibited chromatin occupancy of MLL at select genes, and loss of MLL binding led to changes in gene expression, differentiation and apoptosis; PDX models of both MLL-rearranged AML and ALL showed dramatic reductions in leukaemia burden, with multiple ALL-engrafted mice disease-free beyond a year 1.
- Differentiation requires transcriptional engagement. A cell in G0 is transcriptionally quiescent by definition. There is no published mechanism by which a non-cycling, transcriptionally disengaged cell would execute a differentiation programme at the rate a cycling one does.
- The relapse compartment is quiescent. FOXM1-dependent quiescence sustains MLL-AF9 LSCs 2, and the dormant ALL subpopulation isolated by dye dilution matches MRD cells from patients 3.
If all three hold, the drug class most specific to this disease is weakest against the population that determines whether the disease returns.
What makes this testable rather than rhetorical
The experiment is small. Take KMT2A-rearranged primary blasts or PDX cells, stain with Hoechst plus Pyronin Y to separate G0 (2N DNA, low RNA) from G1 (2N DNA, higher RNA) and from S/G2M, sort, and measure menin-inhibitor sensitivity in each fraction with matched exposure. Read out apoptosis, a differentiation panel (CD11b, CD14, CD15) and target-gene suppression (MEIS1, HOXA9) so that failure to differentiate can be distinguished from failure to engage the target. See exp-cell-cycle-sorted-menin-sensitivity.
The result discriminates cleanly:
| observation | reading |
|---|---|
| G0 fraction shows equal MEIS1/HOXA9 suppression but less apoptosis and less differentiation | the drug engages its target in dormant cells but the downstream programme cannot run; combination must supply the missing step |
| G0 fraction shows less target suppression | drug access or chromatin state differs in G0; a different problem with different solutions |
| No difference across fractions | the hypothesis is wrong and menin inhibitors reach the residual compartment, which would be genuinely reassuring and is worth knowing |
The third row is a real possibility and would be a positive result for the field, which is part of why the experiment is worth running rather than assuming the pessimistic answer.
The uncomfortable corollary for combinations
Two combinations currently under discussion interact with this hypothesis in opposite directions.
CDK4/6 inhibition plus menin inhibition. A CDK4/6 inhibitor holds cells in G1 and may deposit them in G0. On this hypothesis that is the wrong direction. Against that, CDK6 depletion in MLL-rearranged AML inhibits growth through enhanced myeloid differentiation, the same direction as menin inhibition, so the two mechanisms might reinforce. The graph cannot currently say which dominates, and cell-state-cell-cycle-g1-cdk6 records the same uncertainty from the other side.
Anything that releases cells from the niche. The dormant ALL compartment regained drug sensitivity and resumed proliferation once dissociated from the in vivo environment 3. If dormancy is niche-imposed and reversible, then a niche-disrupting agent given before or with a menin inhibitor would be predicted to convert the resistant fraction into a sensitive one — a mechanism-based rationale for a combination that is not currently being pursued in this disease.
| statement | Menin-inhibitor sensitivity in KMT2A-rearranged leukaemia is a decreasing function of cell-cycle exit: blasts in G0 show less apoptosis and less differentiation in response to menin inhibition than cycling blasts from the same sample, so bulk sensitivity measurements systematically overstate activity against the compartment that causes relapse. |
|---|---|
| origin | literature-gap |
| prior probability | 0.45 |
| decision value | Determines whether menin-inhibitor therapy should be combined with an agent that mobilises or cycles the residual compartment, and whether a CDK4/6 inhibitor is the wrong partner. Paediatric menin-inhibitor trials are enrolling now and combination design decisions are being made without this information. |
| gap claim scope | This Queen could find no study sorting KMT2A-rearranged blasts by cell-cycle state before measuring menin-inhibitor sensitivity. That is a statement about searches run on 2026-09-19 with Title/Abstract terms, not a proof of absence; a study reporting this in a figure without saying so in its abstract would not have been found. |
predictions
[ "KMT2A-rearranged blasts sorted into G0 by Hoechst/Pyronin Y show lower apoptosis and less differentiation-marker induction after matched menin-inhibitor exposure than G1 and S/G2M fractions from the same sample.", "MEIS1 and HOXA9 suppression is preserved in the G0 fraction, so the target is engaged but the downstream programme does not run; this distinguishes a downstream block from a drug-access problem.", "Cells recovered at MRD after menin-inhibitor therapy are enriched for a G0 signature relative to diagnosis in the same patient.", "A niche-disrupting agent that releases dormant cells into cycle increases menin-inhibitor sensitivity of the residual compartment.", "Adding a CDK4/6 inhibitor to a menin inhibitor reduces, rather than increases, the differentiation response in the fraction that enters G1 arrest." ]
experiments
[ "exp-cell-cycle-sorted-menin-sensitivity" ]
analyses
[]
falsifiers
[ "Sorted G0, G1 and S/G2M fractions show equal apoptosis and equal differentiation-marker induction after matched menin-inhibitor exposure — this retires the hypothesis and is a genuinely reassuring result for the drug class.", "The G0 fraction shows GREATER sensitivity than cycling fractions, which would invert the clinical implication and argue for extending rather than intensifying therapy.", "MRD cells after menin-inhibitor therapy are not enriched for a G0 signature relative to matched diagnostic samples.", "Dormancy proves not to survive the sort, making the ex vivo design unable to address the question at all — in which case the hypothesis stands but must be tested in vivo." ]
- If menin inhibitors spare the quiescent compartment, then deep responses measured in bulk marrow will be followed by relapse from a population that was never exposed to meaningful drug pressure, and response depth will be a misleading trial endpoint.
- The obvious fix — combining with an agent that drives cells into cycle — would also drive normal haematopoietic stem cells into cycle and risks exhausting the pool an infant needs to recover.
- If instead the hypothesis is false and menin inhibitors do reach dormant cells, then the resistance seen clinically is genuinely chromatin-mediated (ENL persistence, PRC1.1 loss) and no scheduling or combination change will address it.
- What fraction of KMT2A-rearranged blasts is in G0 at diagnosis, and at MRD? Neither number appears to exist, and the hypothesis is about a compartment whose size is unmeasured.
- If dormancy does not survive ex vivo sorting, is there an in vivo design — for instance dye-dilution-marked PDX cells treated in the mouse and analysed without culture — that could answer the question?
- Does the reported year-long disease-free survival of VTP50469-treated MLL-rearranged ALL PDX mice mean the dormant compartment is reached, or that the murine niche does not sustain it?
- Is differentiation-state stratification (GMP-like) already a proxy for proliferative state, making this experiment confirmatory rather than new?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The menin-MLL inhibitor VTP50469 displaced menin from protein complexes and inhibited chromatin occupancy of MLL at select genes, with loss of MLL binding leading to changes in gene expression, differentiation and apoptosis; patient-derived xenografts of MLL-rearranged AML and ALL showed dramatic reductions in leukaemia burden, and multiple mice engrafted with MLL-rearranged ALL remained disease free for more than one year after treatment.
A Menin-MLL Inhibitor Induces Specific Chromatin Changes and Eradicates Disease in Models of MLL-Rearranged Leukemia · Krivtsov AV, Evans K, Gadrey JY, Eschle BK, Hatton C, Uckelmann HJ, et al., Armstrong SA · Cancer Cell 2019What this evidence cannot show
Cell lines and PDX models; cell-cycle state was not a variable in the study and no fraction of the treated population was characterised for quiescence. The year-long disease-free survival in ALL-engrafted mice is in fact evidence AGAINST this hypothesis's strongest form, since a residual quiescent compartment surviving drug should have produced relapse in that window. Read: abstract only.
Evidence that argues against this (1)
- The most direct counter-evidence comes from the same study: multiple mice engrafted with MLL-rearranged ALL remained disease free for more than a year after VTP50469 treatment. If a quiescent compartment were systematically spared, durable disease-free survival in a PDX model would be surprising. Either the murine setting does not sustain the human dormant compartment, or menin inhibition reaches it better than this hypothesis predicts. Read: abstract only (same paper). Krivtsov AV et al. 2019
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Foxm1 is required for survival, quiescence and self-renewal of MLL-AF9-transformed leukaemia stem cells in vivo, acting by binding beta-catenin and stabilising it against degradation, and FOXM1 inhibition suppresses leukaemogenic potential and induces apoptosis of primary LSCs from MLL-rearranged AML patients.
FOXM1 regulates leukemia stem cell quiescence and survival in MLL-rearranged AML · Sheng Y, Yu C, Liu Y, Hu C, Ma R, Lu X, et al., Qian Z · Nature Communications 2020What this evidence cannot show
Principally a mouse MLL-AF9 model; quiescence is scored by cell-cycle distribution rather than by label retention, and the result is unusual in assigning FOXM1, normally described as pro-proliferative, a quiescence-maintaining role. Read: abstract only.
Evidence that argues against this (1)
- The founding characterisation of the MLL-AF9 leukaemia stem cell describes a cell that retains the immunophenotype and global expression profile of a normal granulocyte-macrophage progenitor — a proliferative compartment — with only a limited self-renewal programme re-activated. If the MLL-AF9 LSC is fundamentally a cycling progenitor, the quiescent fraction is a minority within the LSC pool and a drug that spares it would still clear most of the disease-initiating population. Read: abstract only. Krivtsov AV et al. 2006
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In ALL patient-derived xenografts, a rare subpopulation combining long-term dormancy, treatment resistance and stemness resembled primary ALL cells isolated from paediatric and adult patients at minimal residual disease, and its adverse characteristics were reversible: resistant dormant cells became treatment-sensitive and began proliferating once dissociated from the in vivo environment.
Characterization of Rare, Dormant, and Therapy-Resistant Cells in Acute Lymphoblastic Leukemia · Ebinger S, Ozdemir EZ, Ziegenhain C, Tiedt S, Castro Alves C, Grunert M, et al., Jeremias I · Cancer Cell 2016What this evidence cannot show
PDX models not selected for KMT2A rearrangement, in a murine niche; dye dilution marks cells that have not divided, which is necessary but not sufficient for G0. Read: abstract only.
Evidence that argues against this (1)
- Reversibility is a serious methodological problem for the experiment this hypothesis proposes. If dormancy is niche-imposed and lost on removal from the mouse, then sorting cells and exposing them to drug ex vivo measures sensitivity in cells that have already begun leaving the state of interest, and a null result would be uninterpretable. The proposed experiment therefore needs a short-exposure design or an in vivo sorting approach, and this limitation is inherited from the same paper that motivates the hypothesis. Read: abstract only (same paper). Ebinger S et al. 2016
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Ex vivo menin-inhibitor sensitivity in primary AML has been stratified by proteomic and genetic class — GMP-like AML most sensitive, NPM1c/DNMT3A-mutant also sensitive — establishing that the field measures this sensitivity in primary patient cells and stratifies it by cell state, but by differentiation state rather than by cell-cycle state.
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 for what the study did and did not stratify by, not for a claim about cell cycle. The absence of cell-cycle stratification in this study does not establish that no study anywhere has done it. Read: abstract only.
Evidence that argues against this (1)
- Differentiation-state stratification may already capture most of what cycle-state stratification would add, because differentiation state and proliferative state are correlated in the myeloid hierarchy: the GMP-like class that is most bleximenib-sensitive is a proliferative progenitor class. If sensitivity tracks proliferation, the existing GMP-like finding is the cell-cycle result in different clothing and the proposed experiment would be confirmatory rather than novel. Read: abstract only (same paper). Hogeling SM et al. 2025
the hypothesis is testable but not by block-and-release; this entry names the method that did work (centrifugal elutriation)
supplies the separation method that makes the phase-stratified sensitivity question answerable in cells refractory to chemical blocks
the sorting experiment that discriminates the outcomes
the compartment the hypothesis is about
predicts reduced activity against the G0 compartment
- related to →The G1 state in KMT2A-rearranged leukaemia — where CDK6 does a job that is not about the cell cycle at all
predicts CDK4/6 inhibition is the wrong combination partner