Oncofusion

start / graph / drug-chidamide

Drugmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q39-acetylation-cycle

Chidamide (tucidinostat) - the HDAC inhibitor in both of this graph's menin-inhibitor combination papers, and the only one paired with a menin inhibitor in the lymphoid index disease

also called tucidinostat · CS055 · HBI-8000 · Epidaza · chidamide

Chidamide is an HDAC-blocking drug approved in China and Japan. It is the one drug in this area that shows up twice alongside menin inhibitors - the class this whole project is built around - once in myeloid leukaemia and once in the infant t(4;11) leukaemia that is the project's central disease. That makes it the most obvious partner drug in the field, and almost nobody outside China has noticed.

The same thing, for a scientist

Chidamide is an orally available **benzamide class I HDAC inhibitor** covering HDAC1, HDAC2, HDAC3 and HDAC10, approved in China for relapsed/refractory peripheral T-cell lymphoma and in Japan (as tucidinostat) for adult T-cell leukaemia/lymphoma. It earns an entry because of a pattern no other agent in this beat has: **it is the HDAC inhibitor in BOTH menin-inhibitor combination papers this graph holds.** In MLL-rearranged **AML**, subtoxic chidamide plus the menin-MLL inhibitor **MI-3** is *"highly synergistic"* in vitro and in vivo **but not in cells without this genetic aberration** - a genotype-restricted synergy - acting through apoptosis, loss of mitochondrial membrane potential, ROS, disruption of the CHK1/CHK2 DNA-damage checkpoint and homologous-recombination repair [E1]. In **t(4;11) infant B-ALL**, chidamide **enhances the menin inhibitor MI-538's restoration of HDAC7**, and the combination with standard chemotherapy reduces PDX engraftment and delays relapse [E2]. **Two independent groups, two lineages, two different menin inhibitors, the same HDAC inhibitor.** Separately, chidamide plus cladribine acts in AML through an HDAC2/c-Myc/RCC1 axis, with an active trial (NCT05330364) [E3]. The caution this entry must carry: **the genotype-restricted synergy in [E1] is the strongest single argument in the beat for HDAC inhibition in this disease, and it is one paper, in AML, with a tool menin inhibitor, unreplicated.**

Go deeper
Why the genotype restriction in the AML paper matters more than the synergy

Synergy claims are cheap; this one has the feature that makes a synergy claim informative. Ye et al. report that chidamide plus MI-3 displayed "a highly synergistic anti-tumor activity against human MLL-r AML cells in vitro and in vivo, but not those without this genetic aberration" 1.

That comparator is exactly what mechanism-kat8-mof-requirement-is-myeloid-not-fusion-specific demands and what almost nothing else in this beat has. Every other HDAC result here - the romidepsin screen, the connectivity mapping, the class IIA knockdowns, the panobinostat xenografts - lacks a non-KMT2A comparator arm, and this graph's standing position is that such a result cannot distinguish a fusion dependency from a lineage one. This paper ran the arm.

What restrains the reading:

  • The comparator is described in one abstract phrase and the cell lines composing it were not read in this beat; 'those without this genetic aberration' could be a small or poorly matched panel.
  • MI-3 is a tool compound, not revumenib, ziftomenib, bleximenib or enzomenib. Menin inhibitors differ in potency and off-target profile and a combination effect need not transfer.
  • AML, adult framing. The paper's opening sentence is about adult AML prognosis.
  • The mechanism reported - CHK1/CHK2 checkpoint disruption and homologous-recombination impairment - is a DNA-damage mechanism, not a chromatin one. That is consistent with the pattern in mechanism-acetylation-is-a-cycle-not-a-mark: HDAC-inhibitor mechanisms in this disease keep turning out not to be about histone acetylation. It also resembles the MOF mechanism this graph already judged to be a housekeeping requirement rather than a fusion-directed one.
  • One paper, one group, unreplicated.
The convergence, and what to do with it

Set the two combination papers side by side:

Ye et al. 2019 [E1]de Barrios et al. 2025 [E2]
diseaseMLL-rearranged AMLt(4;11) infant B-ALL
menin inhibitorMI-3MI-538
HDAC inhibitorchidamidechidamide
mechanism proposedapoptosis, ROS, CHK1/CHK2 checkpoint, HR repairHDAC7 de-repression, differentiation, RUNX2, glucocorticoid sensitisation
non-fusion comparatoryesnot stated
in vivoyesyes, PDX with chemotherapy

The proposed mechanisms have nothing in common. That is either two real and distinct routes by which class I HDAC inhibition assists menin inhibition in two lineages, or it is two groups each finding a plausible mechanism for a general cytotoxic interaction. This graph cannot currently tell which, and the entry says so rather than choosing.

What makes it actionable anyway. Menin inhibitors are the clinical reality of this disease - revumenib is approved, four more are in this graph. The question "what should be combined with a menin inhibitor in infant ALL?" is live and urgent. Chidamide is the only HDAC inhibitor with a published menin-inhibitor combination in the index disease, it is approved somewhere (so it has human safety data and a supply chain), and it has never been given with a clinical menin inhibitor to anyone. That is an unusually short distance between the graph's evidence and a trial concept - and drug-vorinostat shows the infant regulatory path for an HDAC inhibitor already exists.

Facts and figures
targetgene-hdac2
target noteBenzamide class I HDAC inhibitor covering HDAC1, HDAC2, HDAC3 and HDAC10. gene-hdac2 is named in the schema-required target field because it is the isoform to which a chidamide effect has been molecularly attributed (by co-IP and ChIP, not by selectivity). Chidamide is NOT HDAC2-selective.
modalitysmall molecule, benzamide class, orally available, class I-selective HDAC inhibitor
statusApproved in China (relapsed/refractory peripheral T-cell lymphoma) and in Japan as tucidinostat (adult T-cell leukaemia/lymphoma); NOT approved in the US or EU; never given with a clinical menin inhibitor to any patient
trials in this diseasenone - no chidamide trial in KMT2A-rearranged leukaemia was found in this beat, and no menin-inhibitor combination trial with any HDAC inhibitor
why this agentthe only HDAC inhibitor appearing in BOTH menin-inhibitor combination papers this graph holds, across both lineages
genotype restricted synergytrue
genotype restriction strengthone paper, AML, tool compound, unreplicated - but it is the ONLY non-KMT2A comparator arm in any HDAC result in this beat
mechanisms proposed across the two papers have nothing in commontrue
attribution to an hdac enzymepartial - HDAC2 by co-IP and ChIP in a different (cladribine) combination, not by compound selectivity
shortest path to a trialchidamide plus a clinical menin inhibitor (revumenib, ziftomenib, bleximenib or enzomenib); the agent is approved somewhere so human safety data and supply exist, and the infant HDAC-inhibitor regulatory path is established by NCT02553460
isoform coverage
[
  "HDAC1",
  "HDAC2",
  "HDAC3",
  "HDAC10"
]
trials
[
  "NCT05330364"
]
menin combination evidence
[
  {
    "partner": "MI-3 (tool menin inhibitor)",
    "disease": "MLL-rearranged AML",
    "result": "highly synergistic in vitro and in vivo, NOT in cells without the rearrangement",
    "mechanism": "apoptosis, ROS, CHK1/CHK2 checkpoint disruption, HR repair disruption",
    "pmid": "31590682",
    "title": "Co-inhibition of HDAC and MLL-menin interaction targets MLL-rearranged acute myeloid leukemia cells via disruption of DNA damage checkpoint and DNA repair.",
    "authors": "Ye J, Zha J, Shi Y, et al.",
    "venue": "Clin Epigenetics",
    "year": 2019,
    "title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
  },
  {
    "partner": "MI-538 (tool menin inhibitor)",
    "disease": "t(4;11) infant B-ALL",
    "result": "enhances MI-538's restoration of HDAC7; with chemotherapy reduces PDX engraftment and delays relapse",
    "mechanism": "HDAC7 de-repression, differentiation, RUNX2 repression, glucocorticoid sensitisation",
    "pmid": "40722046",
    "title": "HDAC7 induction combined with standard-of-care chemotherapy provides a therapeutic advantage in t(4;11) infant B-cell acute lymphoblastic leukemia.",
    "authors": "de Barrios O, Ocón-Gabarró I, Gusi-Vives M, et al.",
    "venue": "Biomark Res",
    "year": 2025,
    "title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
  }
]
key ae
[
  "not enumerated in this entry; chidamide's reported class effects include thrombocytopenia, neutropenia, fatigue and QT prolongation, recorded as a pointer and not verified against a label in this beat"
]
resistance
[
  "not characterised in KMT2A-rearranged leukaemia",
  "if the synergy runs through DNA-damage checkpoint disruption, any restoration of CHK1/CHK2 signalling or of HR competence would defeat it without altering HDAC engagement",
  "menin-inhibitor resistance mutations (MEN1 M327I and related) would remove the partner arm entirely"
]
How this could bite us back
  • Combining an HDAC inhibitor with a menin inhibitor in an infant stacks two epigenetically active agents on a chemotherapy backbone. The TINI experience shows that infection, not leukaemia, killed four of fifty infants on a regimen with one epigenetic agent, and three of those deaths came after that agent had stopped.
  • The synergy mechanism in the AML paper is DNA-damage checkpoint disruption and impaired homologous recombination. Deliberately impairing HR in an infant who will receive years of genotoxic chemotherapy and who has a long remaining lifetime is a second-malignancy argument that has to be made explicitly, not assumed away.
  • Menin inhibitor resistance through MEN1 binding-site mutations removes the partner arm of this combination entirely, and leaves an HDAC inhibitor whose monotherapy record in acute leukaemia is a trial terminated for futility.
  • The genotype-restricted synergy - the strongest single argument in this beat for HDAC inhibition in this disease - is one paper, in AML, with a tool compound, from one group, unreplicated, and with a proposed mechanism that this graph previously judged to be the signature of a non-fusion-directed requirement. It should be replicated before it is built on.
  • Chidamide is not approved in the US or EU. A trial concept that depends on it inherits a regulatory and supply problem that vorinostat, panobinostat and romidepsin do not have, and that is a real reason the field may have ignored the most interesting combination data it has.
Still open
  1. Does chidamide synergise with a CLINICAL menin inhibitor - revumenib, ziftomenib, bleximenib or enzomenib - as it does with the tool compounds MI-3 and MI-538? Nobody has tested it, and it is the single most direct question this beat generates.
  2. The two menin-combination papers propose mechanisms with nothing in common. Are these two genuine lineage-specific routes, or two post-hoc explanations of a general cytotoxic interaction?
  3. How robust is the genotype restriction in the AML paper? It is the only non-KMT2A comparator arm in any HDAC result in this beat, and the composition of that comparator panel has not been read.
  4. Is the chidamide effect in the infant B-ALL paper attributable to class I HDAC inhibition at all, or would any cytotoxic agent enhance the menin inhibitor's effect in that assay?
  5. Chidamide is approved in China and Japan and essentially absent from the Western KMT2A literature. How much of the relevant evidence is in Chinese-language publications that this beat's English-language PubMed queries did not reach?
Every source behind this entry (3)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Combination treatment with subtoxic doses of the HDAC inhibitor chidamide and the menin-MLL interaction inhibitor MI-3 displayed highly synergistic anti-tumour activity against human MLL-rearranged AML cells in vitro and in vivo, but not against those without this genetic aberration; co-exposure led to robust apoptosis in MLL-r AML cells with loss of mitochondrial membrane potential and a sharp increase in ROS generation, disrupted the DNA damage checkpoint at the level of CHK1 and CHK2 rather than their upstream kinases ATR and ATM, and disrupted DNA repair likely via homologous recombination but not non-homologous end joining.

    PMID 31590682DOI
    What this evidence cannot show

    AML, and the paper frames its motivation around adult AML prognosis - it says nothing about infants or about lymphoid disease. MI-3 is a tool menin inhibitor, not any of the five clinical agents in this graph, and combination effects need not transfer between menin inhibitors. Chidamide covers HDAC1, HDAC2, HDAC3 and HDAC10 so nothing attributes to an enzyme. The 'not those without this genetic aberration' comparator is a single abstract phrase and the composition of that comparator panel was not read in this beat. One group, unreplicated. The reported mechanism is a DNA-damage mechanism, which in this graph's KAT8/MOF precedent was the signature of a housekeeping rather than a fusion-directed requirement. Abstract read; PMCID PMC6781368, full text not fetched.

    Evidence that argues against this (1)
    • A DNA-damage-and-ROS mechanism is what this graph concluded was NOT fusion-directed when it appeared for MOF: loss of DNA-damage-repair gene expression, increased gamma-H2AX foci and global histone acetylation loss were read as a housekeeping requirement of a rapidly dividing cell rather than a lesion-directed one. The same reasoning should be applied here before the genotype restriction is taken as settling the matter, because a genotype-restricted synergy and a non-specific mechanism are in tension. Valerio DG et al. 2017
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In t(4;11) infant B-ALL, the menin inhibitor MI-538 restores HDAC7 expression and the effect is enhanced by the class I HDAC inhibitor chidamide; the treatment drives leukaemic pro-B cells toward a more differentiated state and impairs aberrant proliferation in an HDAC7-dependent manner, increases glucocorticoid sensitivity of PDX cells ex vivo by repressing RUNX2, and combined with standard chemotherapy reduces PDX cell engraftment in vivo and delays relapse.

    HDAC7 induction combined with standard-of-care chemotherapy provides a therapeutic advantage in t(4;11) infant B-cell acute lymphoblastic leukemia. · de Barrios O, Ocon-Gabarro I, Gusi-Vives M, Collazo O, Meler A, Romecin PA, et al. · Biomarker Research 13(1):99 2025
    PMID 40722046DOI
    What this evidence cannot show

    Chidamide's role here is as an ENHANCER of the menin inhibitor's effect on HDAC7, not as the primary agent, and the abstract does not quantify its contribution. MI-538 is a tool compound. The therapeutic benefit is stated to be only partially HDAC7-mediated. No non-KMT2A comparator arm is described. A separate group reports that knocking HDAC7 down kills these same cells, and this beat does not resolve that conflict - see gene-hdac7. Abstract read; PMCID PMC12305908, full text not fetched.

    Evidence that argues against this (1)
    • Three class IIA-selective inhibitors failed in vivo at maximum achievable dose in a KMT2A-rearranged ALL xenograft, and shRNA knock-down of HDAC4/5/7 kills KMT2A-rearranged ALL cells - the opposite sign to the induction strategy chidamide is here assisting. The combination's benefit may not run through HDAC7 at all. Verbeek TCAI et al. 2024
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Chidamide combined with cladribine is synergistic on proliferation arrest, G0/G1 arrest and apoptosis in AML cell lines (U937, THP-1, MV4-11) and primary AML cells, acting through an HDAC2/c-Myc/RCC1 axis with HDAC2 shown to interact with c-Myc and c-Myc shown to bind the RCC1 promoter; high RCC1 and HDAC2 expression associate with poor survival in AML patients; the combination was active in primary cells from patients including one with an MLL-AF6 fusion; and the work provides experimental evidence for the clinical trial NCT05330364.

    PMID 36849955DOI
    What this evidence cannot show

    AML. MV4-11 is KMT2A::AFF1-rearranged and is in the panel, and one primary sample carried MLL-AF6, but no KMT2A-versus-non-KMT2A comparison is reported, so this paper contains KMT2A-rearranged material without containing a KMT2A-rearranged finding. Chidamide's four-isoform coverage means the HDAC2 attribution rests on the co-IP and ChIP evidence rather than on compound selectivity. NCT05330364 was not independently retrieved from the registry in this beat and is recorded here as the trial the authors name. Abstract read in full; PMCID PMC9972767, full text not fetched.

    Evidence that argues against this (1)
    • Synergy in AML cell lines including THP-1 and MV4-11 does not transfer to infant KMT2A-rearranged B-lineage disease, which is this graph's index population: U937 carries no KMT2A rearrangement at all, and the two that do are myeloid and from patients aged one and above. Chidamide is a class I-selective benzamide, so a synergy attributed to it still inherits the enzyme-attribution problem this beat documents. Verbeek TCAI et al. 2024
Connected to