Twelve weeks of a menin inhibitor in diabetic mice: more beta cells, lower fasting glucose, and nobody looked for a tumour
also called MI-463 in vivo beta cell · menin inhibitor glucose tolerance mouse · menin inhibitor islet phenotype
Diabetic mice given a menin-blocking drug every day for twelve weeks ended up with more insulin-making cells, lower fasting blood sugar and better sugar handling, with no change in weight or appetite. The study never examined their pancreases for tumours, so its silence on that point is not reassurance.
The same thing, for a scientist
This is the only published in-vivo pharmacological menin-inhibition experiment with an islet endpoint, and it is the strongest single fact in the inversion. MI-463 at 70 mg/kg/day by oral gavage: in 8-12-week-old chow-fed mice for 8 weeks it improved glucose tolerance and glucose-stimulated insulin secretion and raised the Ki67-positive beta-cell fraction; in 18-week-old high-fat-diet diabetic mice for 12 weeks it lowered fasting blood glucose, progressively improved IPGTT from week 6 through week 12, and raised BrdU-positive beta cells and beta-cell mass, with no difference in body weight, food intake, adiposity index, and improved insulin tolerance. Every effect was abolished in Pbk kinase-dead knock-in mice, establishing the route [E1]. Group sizes are n=4 to n=5. **No histopathology, tumour incidence or long-term follow-up is reported**, and the animals were not followed off drug.
The numbers, read from the full text
From Europe PMC PMC8103087, figures 7 and 8 and their legends 1:
| Arm | Model | Dose | Duration | n | Result |
|---|---|---|---|---|---|
| Chow | Pbk WT/WT vs KI/KI mice, 8-12 wk old | MI-463 70 mg/kg/day gavage | 8 weeks | 4/group | GTT AUC improved (p=0.0005); GSIS enhanced (p=0.0042); Ki67+ beta cells up (p=0.0026). All absent in Pbk KI/KI |
| HFD/DIO | 18-wk-old HFD-fed hyperglycaemic mice | MI-463 70 mg/kg/day gavage | 12 weeks | 5/group | Fasting glucose down (p=0.0026); IPGTT AUC improved (p=0.0024) from week 6; BrdU+ beta cells (p=0.0059) and beta-cell mass (p=0.0103) up; body weight, food intake, adiposity unchanged; ITT improved (p=0.0029) |
| HFD + genotype | Pbk WT/WT vs KI/KI on HFD | MI-463 | 8 weeks | 3/group | IGT improvement present in WT, completely abolished in Pbk KI/KI |
What this does and does not license saying
It licenses: a menin inhibitor, given daily for three months at a dose that works, produces a beneficial endocrine phenotype in the tissue the MEN1 syndrome damages most reliably, with no weight or appetite confound and with a clean genetic epistasis to the proposed effector.
It does not license: "no tumour occurred". The paper does not report pancreatic histopathology for neoplasia, does not state a tumour incidence, and does not follow animals after drug withdrawal. Absence of a reported tumour in a study that did not look for one is absence of measurement, not a negative result. The discussion itself only speculates that limiting Pbk induction to a subset of beta cells "could be a safe way to drive modest cell proliferation, while not increasing beta cell tumorigenesis" - a hope, not a finding.
It also does not license extrapolating to children. The animals are adult mice, the compound is a tool molecule, the model is diet-induced diabetes, and n is 4-5.
Set against the nonclinical package of the approved drugs
risk-menin-inhibitor-class-multiorgan-hyperplasia-nonclinical records that 13-week rat studies of revumenib produced hyperplasia in testes, mammary gland, uterus, pancreas and kidney, irreversible in three of those, at roughly 2x human AUC 2, and that ziftomenib produced hyperplasia in multiple organs in mice and dogs at 0.05-0.3x human AUC 3. So the regulatory package and this paper report the same class of tissue response - proliferation in endocrine and hormone-responsive organs - and grade it oppositely. The honest reading is that increased proliferation is the on-target consequence, and whether it reads as therapy or as toxicity depends on the tissue, the duration and whether it reverses. That is exactly the question nobody has run.
| compound | MI-463 (tool menin inhibitor) |
|---|---|
| dose | 70 mg/kg/day oral gavage |
duration weeks
[ 8, 12 ]
models
[ "chow-fed C57BL/6 Pbk WT/KI", "high-fat-diet-induced obese diabetic" ]
n per group
[ 3, 4, 5 ]
endpoints improved
[ "fasting blood glucose", "IPGTT AUC", "GSIS", "insulin tolerance", "BrdU+ beta cells", "Ki67+ beta cells", "beta-cell mass" ]
endpoints not measured
[ "pancreatic histopathology for neoplasia", "off-drug follow-up", "parathyroid", "bone", "growth" ]
- The study that most supports menin inhibition being safe for the islet never looked for the harm it is cited as excluding.
- If a menin inhibitor genuinely raises beta-cell mass, a child on it for months may have an insulin-secretory phenotype on withdrawal that nobody is watching for, including hypoglycaemia.
- The drug being beneficial in one endocrine tissue is not protective in another; nothing about this result argues against the parathyroid, pituitary or growth-plate risks in this beat.
- Does the islet phenotype reverse, persist or progress after MI withdrawal? No arm of this study addresses it and it is the paediatric question.
- Would revumenib at its approved paediatric exposure reproduce any part of this phenotype?
- In the same animals, what happens to the parathyroid, the growth plate and the pituitary? Nobody has measured the endocrine system as a system under a menin inhibitor.
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
MI-463 at 70 mg/kg/day for 8-12 weeks increased beta-cell proliferation and mass, lowered fasting glucose and improved glucose tolerance in mice, Pbk-dependently.
Menin-regulated Pbk controls high fat diet-induced compensatory beta cell proliferation. · Ma J, Xing B, Cao Y, He X, et al. · EMBO Molecular Medicine 2021What this evidence cannot show
n=4-5 per group; tool compound, not a clinical menin inhibitor; adult mice; diet-induced and genetic diabetes models; no pancreatic histopathology for neoplasia, no off-drug follow-up, no dose-response in vivo, and no measurement of the parathyroid, pituitary or bone in the same animals.
Evidence that argues against this (2)
- The revumenib prescribing information reports hyperplasia in multiple organs including the pancreas in 13-week rat studies at approximately 2x human AUC, irreversible in testes, mammary gland and pancreas - a proliferative response in the same organ, graded as toxicity. Syndax Pharmaceuticals et al. 2026
- Conditional Men1 deletion in the beta cell does not stop at adaptive proliferation: it produces glucagon-expressing tumours and loss of beta-cell transcription-factor identity in older mice. Li F et al. 2015
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
Revumenib caused hyperplasia in pancreas and other organs in 13-week rat toxicology, irreversible in three organs, at ~2x human AUC.
REVUFORJ (revumenib) tablets, for oral use - US Prescribing Information (SPL version 5, published 20 Feb 2026) · Syndax Pharmaceuticals, Inc. · DailyMed / FDA Structured Product Label, setid 6eb3cdbc-0e74-477d-82d6-3bb172d3f63f 2026What this evidence cannot show
Label summary of a study report; the primary study is not public, so lesion severity, incidence and the exact pancreatic compartment involved cannot be checked.
Evidence that argues against this (1)
- Hyperplasia at supratherapeutic exposure in a 13-week rodent study is a common and often reversible finding; the label itself distinguishes the reversible from the irreversible organs, and the beta-cell literature treats islet hyperplasia as adaptive rather than pre-neoplastic when bounded. Ma J et al. 2021
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
Ziftomenib produced hyperplasia in multiple organs in 13-week mouse and dog studies at exposures as low as 0.05-0.3x human AUC.
KOMZIFTI (ziftomenib) capsules, for oral use - US Prescribing Information (SPL version 1, published 16 Dec 2025) · Kura Oncology, Inc. · DailyMed / FDA Structured Product Label, setid b650f696-3391-4274-8b55-a5f5e9d04769 2025What this evidence cannot show
Label text is a summary; the organs are not individually enumerated for dogs, and no reversibility statement is given.
Evidence that argues against this (1)
- The same label reports no parathyroid, calcium, phosphate or PTH abnormality in patients, so the animal proliferative signal has not translated into a detected human endocrine signal for this compound. Kura Oncology et al. 2025
- depends on →How menin holds the beta cell still: JunD recruits it, with HDAC3, onto the Pbk promoter
The in-vivo glucose and proliferation phenotype is abolished in Pbk kinase-dead knock-in mice, tying it to this axis
Same on-target proliferative response in the same organ (pancreas), graded as benefit rather than toxicity; the entry's inference from 'hyperplasia' to 'tumour-suppressor loss phenotype' does not survive unqualified
Provides the pharmacological arm that experiment nominates, and leaves its central question - does it reverse off drug - entirely open, since no withdrawal arm was run
Provides the receptor-level mechanism for the in-vivo glucose benefit, and the human-islet Ki67 result the in-vivo paper is often miscredited with
Genetic counterpart of the pharmacological phenotype; deletion is the deeper exposure and still produces benefit at the durations studied
Human-tissue counterpart of the mouse in-vivo phenotype: MI-2 plus TGF-beta inhibition replicates adult human beta cells without loss of insulin production
- ← part ofThe same drug target, pursued for benefit: menin inhibition is a diabetes strategy in the literature
The strongest single result in the benefit literature: 12 weeks of MI-463 lowers fasting glucose and raises beta-cell mass
- ← depends onHypothesis: the same drug is protective in the islet and harmful in the parathyroid and the growth plate
The islet-benefit arm rests entirely on this one in-vivo tool-compound experiment