Eytan Stein (MSKCC), Ghayas Issa (MD Anderson) and the clinical menin-inhibitor investigator network
also called Stein EM · Issa GC · AUGMENT-101 investigators · KOMET investigators · clinical menin programme
A small number of adult-leukemia doctors at big American cancer centres run essentially every menin-inhibitor trial. They decide which combinations get tested and in whom, and almost none of their studies enrol infants.
The same thing, for a scientist
Eytan M. Stein (Leukemia Service, Memorial Sloan Kettering Cancer Center) and Ghayas C. Issa (Department of Leukemia, MD Anderson Cancer Center) were the corresponding authors of the first-in-human revumenib report. Around them sits a recurring investigator set - Eunice Wang (Roswell Park), Amir Fathi (Massachusetts General Hospital), Richard Stone (Dana-Farber), Ibrahim Aldoss (City of Hope), Joshua Zeidner (UNC), Michael Thirman (Chicago) - that runs the menin-inhibitor trial portfolio across compounds and sponsors. The programme has moved rapidly from monotherapy to combinations, and the combination era is where the strategic problem for an infant-focused laboratory becomes explicit.
What they control2 deeper
Which combinations get tested, in whom, and in what order. The monotherapy questions are answered: revumenib gave 30% CR/CRh in relapsed/refractory KMT2A-r or NPM1-mutant acute leukaemia 1, and the registration-enabling phase 2 of KOMET-001 gave ziftomenib a CR/CRh of 22% in 92 patients with relapsed/refractory NPM1-mutant AML, median age 69 2. The field has now moved to combinations: ziftomenib with venetoclax and azacitidine in 67 patients with relapsed/refractory NPM1-mutant AML, with Issa as senior author 3, and a new wave of menin-inhibitor plus cereblon-modulator studies 4.
On-treatment and relapse correlative samples. Every mechanistic question about how menin inhibitors fail in patients depends on material that flows through these investigators' institutions.
The structural problem: this is an adult AML programme
This is the most important strategic observation in the whole landscape survey and it is easy to miss because the word 'KMT2A' appears everywhere.
Look at the populations. The KOMET-001 phase 2 cohort had a median age of 69 years (range 33-84) and was NPM1-mutant, not KMT2A-rearranged 2. The ziftomenib-venetoclax-azacitidine combination cohort had a median age of 66 and was again NPM1-mutant 3. The current combination studies recruiting at these centres are titled for adults or 'adolescents and adults' 4.
NPM1-mutant AML is an adult disease and is roughly three times more common than KMT2A rearrangement in AML. The commercial and academic centre of gravity of the menin field is therefore adult NPM1-mutant AML, not infant KMT2A-rearranged ALL. Infants are a small, difficult, low-revenue tail of the same mechanism.
The consequences for the lab are concrete:
- Combination regimens will be optimised against adult AML biology and adult-tolerable toxicity, then handed down. Venetoclax-azacitidine is not the infant ALL backbone.
- Resistance data will be generated overwhelmingly in NPM1-mutant adults. Whether KMT2A::AFF1 infants escape the same way is an assumption, not a finding.
- No amount of goodwill from these investigators gives an infant-focused lab access to an infant population, because their trials do not have one.
The partnering route for infants therefore runs through Interfant and the paediatric cooperative groups, not through this network. This network is where you go for the adult comparator data and the resistance samples.
Collaborator or competitor?
Neither, strictly. They are gatekeepers. They do not do interface discovery and will not compete with the lab on mechanism, but nothing the lab discovers reaches a patient without passing through a trial network of this kind.
The realistic ask is narrow and worth making: paired diagnosis-relapse samples from KMT2A-rearranged patients treated with a menin inhibitor, analysed for something they are not already doing. The MEN1 interface mutations are taken; the open question is what happens in the majority of patients who relapse without a MEN1 mutation. That is a defined, non-duplicative correlative project.
| group | Clinical menin-inhibitor investigator network: Eytan M. Stein (MSKCC), Ghayas C. Issa (MD Anderson), Eunice S. Wang (Roswell Park), Amir T. Fathi (Massachusetts General Hospital), Richard M. Stone (Dana-Farber), Ibrahim Aldoss (City of Hope), Joshua F. Zeidner (UNC Lineberger), Michael J. Thirman (University of Chicago) |
|---|---|
| institution | Memorial Sloan Kettering Cancer Center, New York; The University of Texas MD Anderson Cancer Center, Houston; plus the recurring US academic trial-site network listed above |
| focus | Early-phase clinical development of menin inhibitors: monotherapy dose-finding, registration cohorts, and combinations with venetoclax/azacitidine, intensive chemotherapy, and cereblon modulators |
| partnering angle | Gatekeepers, not competitors. Do not expect infant access from this network - its centre of gravity is adult NPM1-mutant AML. The one defined, non-duplicative ask is paired diagnosis-relapse material from KMT2A-rearranged patients who relapsed on a menin inhibitor WITHOUT a MEN1 interface mutation, which is the majority mechanism and is unexplained. For infants, route through Interfant and the paediatric cooperative groups instead. |
| relationship | gatekeeper; neither collaborator nor competitor on mechanism |
| population centre of gravity | adult relapsed/refractory NPM1-mutant AML (median age 66-69), not infant KMT2A-rearranged ALL |
key papers — the numbers behind the chart
[
{
"title": "The menin inhibitor revumenib in KMT2A-rearranged or NPM1-mutant leukaemia",
"year": 2023,
"venue": "Nature",
"pmid": "36922593",
"doi": "10.1038/s41586-023-05812-3"
},
{
"title": "Ziftomenib in relapsed or refractory AML (KOMET-001 phase 1)",
"year": 2024,
"venue": "Lancet Oncology",
"pmid": "39362248",
"doi": "10.1016/S1470-2045(24)00386-3"
},
{
"title": "Ziftomenib in Relapsed or Refractory NPM1-Mutated AML (KOMET-001 phase 2)",
"year": 2025,
"venue": "J Clin Oncol",
"pmid": "40997296",
"doi": "10.1200/JCO-25-01694"
},
{
"title": "Ziftomenib with venetoclax and azacitidine in relapsed/refractory NPM1-mutated AML (KOMET-007)",
"year": 2026,
"venue": "Blood",
"pmid": "42227701",
"doi": "10.1182/blood.2026034043"
}
]trials observed
[ "NCT04065399", "NCT07355335", "NCT07356154", "NCT05886049", "NCT07751991" ]
not verified
[ "minimum eligible age for NCT07355335, NCT07356154, NCT05886049 and NCT07751991", "how many infants under one year have received a menin inhibitor on any trial to date", "current recruitment status of each listed trial" ]
- The whole clinical programme is optimised for adult NPM1-mutant AML; regimens, doses, combination partners and toxicity thresholds handed down to infants may be wrong in all four respects.
- A 4.6-month median duration of response in the registration cohort is the single most sobering number for the obligate-partner-interface hypothesis: the mechanism works, and then it stops working.
- Because the resistance literature is being written in adult NPM1-mutant disease, the field may standardise on a resistance model that does not describe infant KMT2A-rearranged ALL at all.
- Dependence on a small investigator network means the questions that get asked are the ones sponsors will fund.
- How do KMT2A-rearranged patients relapse on menin inhibitors when they do NOT acquire a MEN1 interface mutation? That is the majority of relapses and the mechanism is unpublished as far as this survey found.
- How many infants under one year have ever received a menin inhibitor, on any trial, anywhere? If the answer is in single digits, then every claim about menin inhibition in infant ALL is an extrapolation from adults.
- A median duration of response of 4.6 months in the ziftomenib registration cohort: is that ceiling a property of a heavily pretreated adult population, or a property of interface disruption as a strategy?
- Will the combination era make menin inhibition durable, or merely change which escape route the disease takes?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Issa (MD Anderson) and Stein (MSKCC) were the corresponding authors of the first-in-human phase 1 revumenib trial (NCT04065399), which reported a 30% CR/CRh rate in relapsed/refractory KMT2A-rearranged or NPM1-mutant acute leukaemia across a multi-centre US investigator network.
The menin inhibitor revumenib in KMT2A-rearranged or NPM1-mutant leukaemia · Issa GC, Aldoss I, DiPersio J, Cuglievan B, Stone R, Arellano M, Thirman MJ, Patel MR, Dickens DS, Shenoy S, Shukla N, Kantarjian H, Armstrong SA, et al., Stein EM · Nature 2023What this evidence cannot show
Single-arm phase 1. Several authors are Syndax employees. Paediatric sites appear in the author list (Dickens at Iowa, Shukla at MSK, Shenoy at Washington University), so the trial was not adults-only, but the number of infants enrolled is not stated in the abstract and was not verified here.
Evidence that argues against this (1)
- The same structural point at the other end of the class, from the revumenib resistance paper's competing-interests statement. E.M. Stein declares 'research support to his institution from Agios, Amgen, Astellas, Bayer, Biotheryx, Bristol Myers Squibb, Eisai Foghorn, Servier, Syndax, Syros and consulting fees from' a list that includes both Syndax and Kura, and 'also holds equity interest in Auron Therapeutics'. Read together with Issa's declarations on the revumenib and ziftomenib reports, the two corresponding authors of the first-in-human revumenib paper are each funded by, and consult for, both of the companies whose approved menin inhibitors are now being compared in the clinic. Two consequences for this entry's strategic advice. (1) It supports, on firmer grounds than the entry gives, the recommendation to enter through a trial investigator rather than a sponsor -- but it also means the investigator is not a neutral channel and the sponsor will effectively be in the room. (2) The entry's second open question, how many infants under one year have ever received a menin inhibitor, is exactly the kind of question this network is not structured to answer: its financial and trial relationships are with adult programmes, and the infant question sits with COG AALL2121 and TINI 2 instead. Caveat: disclosures are declarations, not conduct findings; they are also not complete, since institutional research support does not disclose amounts or direction. Read: the competing-interests section of the PubMed record; abstract read, not full text. Perner F et al. 2023
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In the registration-enabling phase 2 of KOMET-001, 92 patients with relapsed/refractory NPM1-mutant AML, median age 69 years (range 33-84), received ziftomenib 600 mg once daily and met the primary endpoint with a CR/CRh rate of 22%, overall response rate 33%, median duration of response 4.6 months.
Ziftomenib in Relapsed or Refractory NPM1-Mutated AML · Wang ES, et al. · Journal of Clinical Oncology 2025What this evidence cannot show
This cohort is NPM1-mutant AML, not KMT2A-rearranged, and the median age of 69 makes it maximally distant from the infant population. Single-arm registration design without a randomised comparator. The 4.6-month median duration of response is the number that should temper enthusiasm about durability of menin-interface disruption.
Evidence that argues against this (1)
- A disclosure fact that bears on how this investigator network's comparative statements should be read, taken verbatim from the paper's declaration of interests and nothing else. In the KOMET-001 report, G.C. Issa declares 'research funding to their institution from Astex, Cullinan Oncology, Kura Oncology, Merck, Novartis, and Syndax Pharmaceuticals; consultancy fees from AbbVie, Kura Oncology, Novartis, and Syndax Pharmaceuticals; support for attending meetings and travel from Kura Oncology; steering committee role with Kura Oncology and Novartis', and E.S. Wang declares advisory board participation for both Sumitomo and Syndax. The same declaration records, for several authors including Issa and Wang, 'medical writing support for this work from Kura Oncology' -- that is, the sponsor provided writing support for the primary clinical report of its own compound. The point is not that anyone behaved improperly; these are standard, properly declared arrangements. The point is structural and it is decision-relevant for this project: the same small investigator set holds research funding and consultancy from Syndax, Kura, Sumitomo and Janssen simultaneously, so the field has no financially disinterested clinical voice comparing menin inhibitors, and cross-compound claims that circulate in reviews and at meetings trace back to investigators paid by more than one of the sponsors. A lab seeking an independent read on which compound to pursue in infant disease will not find one here and should weight registry and preclinical data accordingly. Caveat: declared interests are disclosures, not findings; they vary between papers and none of this speaks to the validity of the trial results. Read: the declaration-of-interests section of the PubMed record; abstract read, not full text. Wang ES et al. 2024
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The KOMET-007 phase 1 trial combined ziftomenib with venetoclax and azacitidine in 67 adults with relapsed/refractory NPM1-mutant AML (median age 66, 55% with prior venetoclax), with Issa as senior author, marking the field's move from monotherapy to combination.
Ziftomenib with venetoclax and azacitidine in relapsed/refractory NPM1-mutated acute myeloid leukemia · Wang ES, ... Issa GC · Blood 2026What this evidence cannot show
Adults only, NPM1-mutant only, phase 1. Venetoclax-azacitidine is an older-adult AML backbone with no equivalent in infant ALL, so this combination result does not transfer to the population the lab cares about.
Evidence that argues against this (1)
- Sixty-seven adults with a median age of 66 in relapsed or refractory NPM1-mutant AML, 55% venetoclax-pretreated. Every axis of that population differs from the beat's — age, lesion, prior therapy — so the triplet's tolerability and activity there does not transfer to a KMT2A-rearranged infant, in whom the same combination has not been given. Journal of Clinical Oncology 2025
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.trial_registry
The current menin combination portfolio at these institutions includes ziftomenib plus mezigdomide in adolescents and adults with relapsed/refractory AML (NCT07355335, Fathi at Massachusetts General and Stone at Dana-Farber), revumenib plus mezigdomide in leukaemia (NCT07356154, Memorial Sloan Kettering with Issa at MD Anderson and Aldoss at City of Hope), revumenib with daunorubicin and cytarabine in newly diagnosed AML (NCT05886049), and a further menin-class agent CHM-029 (NCT07751991).
ClinicalTrials.gov investigator and site records for menin-inhibitor trials in KMT2A-rearranged acute leukaemia, retrieved via the ClinicalTrials.gov API v2 · US National Library of Medicine · ClinicalTrials.gov 2026What this evidence cannot show
Retrieved from a registry search of investigators, not from reading each protocol. Minimum eligible ages for these specific trials were NOT individually verified in this pass; the 'adolescents and adults' framing comes from the registered study titles, which is weaker than the eligibility record. Registry entries are sponsor-submitted and can lag reality. Site lists change constantly.
Evidence that argues against this (1)
- A registry field is a snapshot with a date on it. Status, site count, target enrolment and eligibility are edited by sponsors between checks, and this graph has already recorded one trial whose status string changed under it. Anything planned on these numbers has to re-query them rather than read them here. A combination portfolio is also the most volatile thing in a registry: studies are added, suspended and withdrawn between quarters, and an investigator named on a record is not a commitment to collaborate. U.S. National Library of Medicine 2026
ran the first-in-human and registration trials
- related to →Kura Oncology — ziftomenib (KO-539): an adults-only sponsor that quietly supplies infant trials
Counter-evidence (declaration of interests, PMID 39362248): the same investigators hold research funding, consultancy and steering-committee roles with both Kura and Syndax, and the KOMET-001 report acknowledges sponsor medical-writing support -- no financially disinterested cross-compound clinical voice exists
- related to →Syndax Pharmaceuticals — revumenib (SNDX-5613), the only menin inhibitor licensed down to infancy
Counter-evidence (competing interests, PMID 36922589): Stein reports institutional research support from Syndax and consulting fees from Syndax and Kura, plus equity in Auron Therapeutics
corresponding authors of the first-in-human revumenib trial (NCT04065399)
run the KOMET programme including the phase 2 registration cohort and the venetoclax/azacitidine combination
clinical programme centred on adult relapsed/refractory AML, predominantly NPM1-mutant, median age 66-69
gatekeeper for menin-inhibitor on-treatment and relapse material, the hardest resource to access