Valemetostat - an approved dual EZH1/EZH2 inhibitor that has already worked in an MLL-AF4 xenograft, in a disease nobody has enrolled it in
also called valemetostat · valemetostat tosilate · DS-3201 · DS-3201b · EZHARMIA
Valemetostat blocks both EZH2 and its back-up EZH1, which is the combination this leukaemia's biology says is needed. It is already approved in Japan for a different blood cancer, and in the laboratory it shrank a mouse tumour carrying the exact fusion that causes infant leukaemia - but it has never been given to a patient with that disease.
The same thing, for a scientist
Valemetostat tosilate (DS-3201, EZHARMIA) is an oral **dual EZH1/EZH2** inhibitor developed by Daiichi Sankyo, approved in Japan in September 2022 for relapsed/refractory adult T-cell leukaemia/lymphoma and subsequently for peripheral T-cell lymphoma [E1][E4]. Its registrational phase 2 in R/R ATL gave a centrally reviewed ORR of **48.0%** (90% CI 30.5-65.9) in 25 heavily pretreated patients (median 3 prior lines, 24 of 25 prior mogamulizumab), with 5 complete and 7 partial remissions, median duration of response not reached, and **haematological toxicity as the dominant adverse-event class** - thrombocytopenia, anaemia, lymphopenia, leukopenia and neutropenia at grade 3 or above [E2]. **The reason this entry sits in the Polycomb beat of a KMT2A graph is a single preclinical result**: in B-ALL cell lines DS-3201 inhibited growth **more significantly and strongly than the EZH2-selective EPZ-6438 (tazemetostat)**, up-regulated *CDKN1C* and *TP53INP1* as direct EZH1/2 targets confirmed by CRISPR knockout, and **significantly inhibited growth of a B-ALL patient-derived xenograft harbouring MLL-AF4** [E3]. The parent chemical programme reported activity against leukaemia lines with MLL-AF9, MLL-AF4 and AML1-ETO fusions [E5]. **So the drug class the biology points at is approved, the compound has been tested against the index fusion in vivo, and it worked - and `valemetostat AND (KMT2A OR MLL)` returns 1 PubMed record** [E6]. This is the clearest translational opening the PRC2 arm produced, and it comes with a specific and serious bite-back in an infant B-lineage disease.
The case for and the case against, stated separately
For:
- It covers both catalytic paralogues, which is the specific requirement the KMT2A genetics imposes (
gene-ezh1). - It has already worked against the index fusion in vivo, in a human PDX 3.
- It is approved, so the pharmacokinetics, formulation and adult safety profile exist 12.
- Its mechanism in B-ALL - CDKN1C and TP53INP1 de-repression - is a differentiation/arrest mechanism compatible with the menin-inhibitor differentiation logic this project already works in.
Against:
- The toxicity is haematological and dose-limiting 2. PRC2 is required for HSC maintenance, so the organ being treated is the organ being suppressed. A 69-year-old with R/R ATL and an infant with newly diagnosed ALL do not have the same tolerance for cytopenias.
- The one index-fusion result is a single PDX with no n, no dose and no stated patient age in the abstract 3, and tazemetostat was not run in that PDX, so 'dual beats selective' rests on the cell-line arm.
- EZH2 inhibition raises H3K27ac at the same residue and, in B cells with high AID activity or DNA-repair deficiency, increases chromosomal translocation frequency - see
risk-ezh2-inhibition-b-lineage-translocations. In infants who will live for decades, in a translocation-defined B-lineage disease, this needs an argument rather than a citation of the monotherapy-safe headline. - Resistance routes are already mapped: PRC2 gatekeeper mutations and compensatory DNA methylation 4. The second matters specifically here, because t(4;11) infant ALL is one of the most promoter-hypermethylated leukaemia subtypes in the paediatric literature - so the disease may be pre-adapted to the escape route.
- No paediatric formulation, no paediatric dose and no paediatric exposure data are established by any of the cited work.
What the graph should ask for before anyone proposes a trial
Three things, none of them expensive, and in this order:
- Reproduce the PDX result in an infant KMT2A::AFF1 model with an H3K27me3 pharmacodynamic. The existing result is one PDX of unstated provenance;
exp-eed-allosteric-inhibition-in-kmt2a-aff1proposes the matched design, and valemetostat belongs in it as the comparator arm. - Measure the haematopoietic cost directly. The mouse shRNA therapeutic-window claim was made against immortalised non-transformed lines [see
gene-eed], not primary HSPCs, and the clinical toxicity says the window is narrower than that assay reported. - Test the translocation hazard in the intended combination. PMID 40825195 found monotherapy safe and combination with a PI3K-delta inhibitor hazardous; the intended paediatric partner would be glucocorticoids plus a menin inhibitor, and neither has been tested for this interaction.
Until those exist, the graph's position should be: the right drug class is identified and approved, and the evidence that it should be given to an infant is one xenograft.
| target | gene-ezh2 |
|---|---|
| secondary target | gene-ezh1 |
| modality | small molecule, oral, dual EZH1/EZH2 inhibitor |
| status | Approved in Japan September 2022 for relapsed/refractory adult T-cell leukaemia/lymphoma (EZHARMIA); subsequently for peripheral T-cell lymphoma. Not approved and not enrolled in any KMT2A-rearranged or paediatric leukaemia. |
| paediatric status | no paediatric formulation, dose or exposure data established by any cited work |
aliases development
[ "DS-3201", "DS-3201b" ]
trials
[ "NCT04102150" ]
key ae
[ "thrombocytopenia", "anaemia", "lymphopenia", "leukopenia", "neutropenia", "alopecia", "dysgeusia" ]
efficacy in approved indication
{
"orr_rr_atl_pct": 48,
"ci90": "30.5-65.9",
"complete_remissions": 5,
"partial_remissions": 7,
"n": 25,
"median_age": 69,
"pmid": "36150143",
"title": "An open-label, single-arm phase 2 trial of valemetostat for relapsed or refractory adult T-cell leukemia/lymphoma.",
"authors": "Izutsu K, Makita S, Nosaka K, et al.",
"venue": "Blood",
"year": 2023,
"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."
}index fusion evidence
{
"system": "B-ALL patient-derived xenograft harbouring MLL-AF4",
"result": "significant growth inhibition",
"n_pdx": 1,
"pmid": "36280659",
"limitations": "no n, dose, age or H3K27me3 pharmacodynamic stated in the abstract; tazemetostat not run in the PDX",
"title": "Dual inhibition of EZH1/2 induces cell cycle arrest of B cell acute lymphoblastic leukemia cells through upregulation of CDKN1C and TP53INP1.",
"authors": "Ito J, Yamagata K, Shinohara H, et al.",
"venue": "Int J Hematol",
"year": 2023,
"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."
}resistance
[ "PRC2 gatekeeper mutations", "compensatory DNA methylation - of particular concern in t(4;11) infant ALL, which is extensively promoter-hypermethylated" ]
counted absence
{
"query": "valemetostat AND (KMT2A OR MLL)",
"records": 1,
"aliases_not_searched": [
"DS-3201",
"DS-3201b"
],
"date": "2026-09-20"
}- The dose-limiting toxicity of this class is marrow suppression, in a disease where the marrow is both the tumour and the organ that must recover. Adult R/R ATL tolerance at median age 69 does not transfer to an infant receiving intensive chemotherapy.
- Dual inhibition depletes H3K27me3 more completely than EZH2-selective inhibition, and the same depletion raises H3K27ac and increases chromosomal translocation frequency in B cells with high AID activity or DNA-repair deficiency (PMID 40825195). A more complete inhibitor is a more complete exposure of that hazard, in a translocation-defined B-lineage leukaemia, in patients with decades of life ahead.
- Compensatory DNA methylation is a reported resistance route and t(4;11) infant ALL is already one of the most promoter-hypermethylated leukaemia subtypes - so the disease may be pre-adapted to escape, and the biomarker of response (CDKN1C, an imprinted DNA-methylation-controlled locus) is entangled with the resistance mechanism.
- EZH2 loss of function causes myeloid neoplasia, and dual inhibition removes the EZH1 buffer as well. The second-malignancy question is sharper for this drug than for an EZH2-selective one, and no cited study has a follow-up long enough to address it.
- Does valemetostat work in an INFANT KMT2A::AFF1 model, with a measured H3K27me3 pharmacodynamic? The existing index-fusion evidence is one PDX of unstated provenance, and no infant model has been tested.
- What does valemetostat cost normal infant haematopoiesis? Every grade 3+ adverse event in the registrational trial was a cytopenia, PRC2 is required for HSC maintenance, and the preclinical therapeutic-window claim was measured against immortalised lines rather than primary HSPCs.
- Does valemetostat combine with, or interfere with, a menin inhibitor? If menin drives EZH2 expression the two hit one axis twice; if menin inhibition works by shifting KMT2A against polycomb, removing polycomb could blunt it. The combination is the obvious next move and the graph can state that it is not obviously additive.
- Is the compensatory-DNA-methylation resistance route pre-loaded in t(4;11)? Infant t(4;11) and t(11;19) are extensively promoter-hypermethylated and the degree tracks relapse-free survival, so the escape mechanism reported for this drug class may already be present at diagnosis.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.review
Valemetostat tosilate (EZHARMIA), a selective dual inhibitor of EZH1 and EZH2 developed by Daiichi Sankyo, was approved in Japan in September 2022 for the treatment of patients with relapsed or refractory adult T-cell leukaemia/lymphoma.
Valemetostat Tosilate: First Approval. · Keam SJ · Drugs 2022What this evidence cannot show
A drug-approval summary, cited for regulatory status only. Approval is a Japanese PMDA decision for R/R ATL; it establishes nothing about efficacy or safety in leukaemia of any other type, in children, or in KMT2A-rearranged disease. Regulatory status is a dated fact that changes.
Evidence that argues against this (1)
- An approval in adult T-cell leukaemia/lymphoma - an HTLV-1-driven mature T-cell malignancy in adults - transfers almost nothing to infant pro-B ALL. The relevant shared feature is only that PRC2 is targetable in a haematological malignancy with acceptable adult toxicity; the lineage, the driver, the age and the marrow reserve are all different. Izutsu K et al. 2023
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
In a multicentre open-label single-arm phase 2 trial of valemetostat 200 mg/day orally in 25 patients with relapsed or refractory aggressive adult T-cell leukaemia/lymphoma (median age 69, median 3 prior lines, 24 with prior mogamulizumab), the centrally reviewed overall response rate was 48.0% (90% CI 30.5-65.9) with 5 complete and 7 partial remissions and median duration of response not reached; treatment-emergent adverse events in at least 20% included thrombocytopenia, anaemia, alopecia, dysgeusia, neutropenia, lymphopenia, leukopenia, decreased appetite and pyrexia, and grade 3 or above events included thrombocytopenia, anaemia, lymphopenia, leukopenia and neutropenia. NCT04102150.
An open-label, single-arm phase 2 trial of valemetostat for relapsed or refractory adult T-cell leukemia/lymphoma. · Izutsu K, Makita S, Nosaka K, Yoshimitsu M, Utsunomiya A, Kusumoto S, et al. · Blood 2023What this evidence cannot show
Abstract read from the eutils record; full text not read. 25 patients, single arm, no control, median age 69 - an adult T-cell malignancy, not a paediatric or B-lineage disease and not KMT2A-rearranged. The safety profile establishes that the class is tolerable in heavily pretreated adults; it does not establish tolerability in an infant receiving concurrent intensive chemotherapy, and every grade 3+ event listed is a cytopenia.
Evidence that argues against this (1)
- Every grade 3 or above event reported is haematological, which is the expected consequence of removing a complex that is essential for haematopoietic stem cell maintenance. In a leukaemia the target tissue and the dose-limiting tissue are the same, and the mouse therapeutic-window claim that PRC2 suppression spares non-transformed haematopoietic cells was measured against immortalised lines rather than primary HSPCs. Shi J et al. 2013
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
DS-3201 (valemetostat) inhibited growth of B-cell ALL cell lines more significantly and strongly than the EZH2-specific inhibitor EPZ-6438, induced cell-cycle arrest and apoptosis in vitro, up-regulated CDKN1C and TP53INP1 which CRISPR/Cas9 knockout confirmed as direct EZH1/2 targets responsible for the antitumour effect, and significantly inhibited the growth of a B-ALL patient-derived xenograft harbouring MLL-AF4.
Dual inhibition of EZH1/2 induces cell cycle arrest of B cell acute lymphoblastic leukemia cells through upregulation of CDKN1C and TP53INP1. · Ito J, Yamagata K, Shinohara H, Shima Y, Katsumoto T, Aikawa Y, et al. · Int J Hematol 2023What this evidence cannot show
Abstract read from the eutils record; full text not read. **This single result carries the entire index-fusion case for this drug.** One PDX, with no n, no dose, no patient age and no statement of infant status in the abstract; tazemetostat was not run in the PDX so the head-to-head is cell-line only and is not matched for exposure or target engagement; no H3K27me3 pharmacodynamic is described; and the CDKN1C/TP53INP1 mechanism was established in B-ALL lines generally rather than in the MLL-AF4 PDX.
Evidence that argues against this (1)
- The de-repressed mediators differ from those reported for PRC2 loss in mouse MLL-AF9 (Cdkn2a, Cdkn2b, Gata2, Egr1), so there is no single biomarker that could be used to select patients or to confirm mechanism across models - and CDKN1C is an imprinted locus whose regulation involves DNA methylation as well as H3K27me3, making it a confounded pharmacodynamic for a drug whose reported resistance route is compensatory DNA methylation. Danis E et al. 2015
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.review
EZH2-selective inhibitor efficacy is limited by functional redundancy with EZH1, which the dual EZH1/2 inhibitor valemetostat overcomes by reactivating tumour-suppressor genes, achieving durable responses in adult T-cell leukaemia/lymphoma and peripheral T-cell lymphoma; therapeutic resistance can nonetheless emerge through PRC2 gatekeeper mutations and compensatory DNA methylation.
Polycomb repressive complexes as therapeutic targets in hematologic malignancies. · Yamagishi M, Iwama A, Kitabayashi I · Exp Hematol 2026What this evidence cannot show
A 2026 review, cited for the field's own account of the redundancy problem and of the resistance routes rather than for a primary result. The resistance statement is unquantified - no frequency, no time to resistance and no patient series are given in the abstract - and the review does not discuss KMT2A-rearranged leukaemia.
Evidence that argues against this (1)
- Compensatory DNA methylation as a resistance route is specifically worrying in this disease rather than generically: t(4;11) and t(11;19) infant ALL are characterised as extensively promoter-hypermethylated, with the degree of hypermethylation influencing relapse-free survival, while t(9;11) and MLL-wild-type infant ALL resemble normal bone marrow. The index disease may therefore already possess the escape mechanism. Stumpel DJ et al. 2009
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Orally bioavailable EZH1/2 dual inhibitors from the programme that produced valemetostat suppressed cellular H3K27 trimethylation more than EZH2-selective inhibitors, showed greater antitumour efficacy in vitro and in vivo, and in a haematological cancer panel assay were effective against leukaemia carrying fusion genes including MLL-AF9, MLL-AF4 and AML1-ETO, with no severe toxicity in rats treated for 14 days above the antitumour exposure.
Novel orally bioavailable EZH1/2 dual inhibitors with greater antitumor efficacy than an EZH2 selective inhibitor. · Honma D, Kanno O, Watanabe J, Kinoshita J, Hirasawa M, Nosaka E, et al. · Cancer Sci 2017What this evidence cannot show
Abstract read from the eutils record; full text not read. Sponsor-authored panel assay with no n, no potency values and no fusion-versus-non-fusion comparison in the abstract; the headline efficacy is EZH2-mutant DLBCL. The compounds described are the programme's leads and the abstract does not state which, if any, is valemetostat itself. The 14-day rat toxicology statement carries no information about the months of exposure a paediatric leukaemia protocol would require, and is contradicted in direction by the human trial's haematological dose-limiting toxicities.
Evidence that argues against this (1)
- The clinical experience overrides the rat statement: in 25 ATL patients every grade 3 or above adverse event was a cytopenia, so 'no severe toxicity in rats' did not predict the human dose-limiting profile. A preclinical safety statement from a 14-day rodent study should not be carried forward into a paediatric argument. Izutsu K et al. 2023
- E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
PubMed `valemetostat AND (KMT2A OR MLL)` returns 1 record while `valemetostat AND leukemia` returns 37 and `valemetostat AND (approval OR "adult T-cell")` returns 39 - so the drug has a substantial leukaemia literature that is almost entirely adult T-cell leukaemia/lymphoma, and essentially none of it touches KMT2A-rearranged disease.
PubMed esearch counts for valemetostat queries, run 2026-09-20 · NCBI PubMed 2026What this evidence cannot show
The development code **DS-3201 was not searched separately**, and the one paper carrying the index-fusion result (PMID 36280659) uses 'DS-3201' in its abstract alongside 'valemetostat' - so the INN-only count happens to capture it here but would not be reliable for a different paper. The 37 and 39 counts were not title-checked. This is an absence of published KMT2A-rearranged study, not an absence of experiment.
Evidence that argues against this (1)
- This graph's own alias rule says a count keyed to one name can under-count by an order of magnitude, and the DS-3201 code is the one used throughout the preclinical literature from the originating programme. The absence recorded here should be re-run over DS-3201 and DS-3201b before anyone relies on it. Ito J et al. 2023
dual EZH1/EZH2 inhibitor, approved in Japan for R/R ATL
the second half of the dual mechanism, and the reason it outperforms tazemetostat here
- treated by →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
significantly inhibited growth of a B-ALL PDX harbouring MLL-AF4 - the only index-fusion result for any PRC2 drug
no paediatric formulation, dose or exposure data exists; the case rests on one xenograft
- ← analogous toTazemetostat - an approved EZH2 inhibitor, and the drug this disease's own biology predicts will not work
same target family, different coverage - a drug-class difference, not a potency difference
same hazard, more completely, because dual inhibition depletes the mark further