Menin is required in neural crest for the palate to close - and both menin inhibitors cause cleft palate in animals
also called menin neural crest palatogenesis · ziftomenib cleft palate · menin rib patterning
Deleting menin in the cells that build the face kills mouse pups at birth with a cleft palate and skull defects. Both approved menin drugs cause cleft palate and skeletal malformations in animal pregnancies, at doses below the human dose.
The same thing, for a scientist
A mechanistic link this graph has not drawn, and the two halves are both checkable. Genetically: tissue-specific inactivation of menin in Pax3- or Wnt1-expressing neural crest cells causes perinatal death, cleft palate and other cranial bone defects, associated with decreased p27(kip1); deletion in Pax3-expressing somite precursors additionally produces rib patterning defects [E1]. Pharmacologically: the KOMZIFTI label reports that ziftomenib in pregnant mice during organogenesis caused embryo-fetal mortality, structural abnormalities and altered fetal growth at approximately **0.3x** the clinical AUC, with the malformations enumerated as *cleft palate, fused cervical arches, absent lumbar vertebrae, fused and/or misaligned costal cartilage, absent or short rib, supernumerary suture bone or split palatine of the skull, and fused, unossified or incompletely ossified sternebrae* [E2]. The REVUFORJ label reports embryo-fetal mortality, malformations and altered fetal growth in rats at approximately **0.5x** the human AUC [E3].
The concordance is specific, not generic
Developmental toxicity in a pregnant animal is common and usually uninformative about mechanism. This one is different: the named malformations in the ziftomenib label are the malformations of the neural-crest menin knockout. Cleft palate, split palatine, supernumerary suture bone, costal and rib defects, vertebral fusion - that is the Pax3/Wnt1 neural-crest and somite territory 1, lesion for lesion. The label reports the phenotype and the genetics explain it, and no published source connects them.
This matters beyond pregnancy because it establishes that menin loss has an on-target developmental consequence in a skeletal and craniofacial programme, at sub-clinical exposure multiples (0.3x and 0.5x), in two species and two chemotypes. Both labels handle it only as an embryo-fetal contraindication.
Why a paediatric beat should care about an embryo-fetal finding
The standard reading is that organogenesis is over by birth, so an embryo-fetal finding is a pregnancy-prevention issue and nothing more. That reading is weaker than it looks for this population:
- Revumenib is approved from 1 year of age and the safety population included 25 infants under 2 3. Craniofacial sutures, the secondary palate's postnatal growth, vertebral ossification centres and the sternebrae are not finished at 1 year.
- The p27(kip1) route implicated in 1 is the same cyclin-dependent kinase inhibitor arm implicated in the beta cell and in Men1 neuroendocrine tumorigenesis, so it is a shared, dose-responsive pharmacology rather than a pregnancy-specific idiosyncrasy.
- Adolescents treated with a menin inhibitor may later become pregnant; the labels handle this with contraception advice for 3-6 months, which is appropriate but is not the same as knowing whether ovarian reserve or a future pregnancy is affected. Both labels state findings in animals indicate impaired fertility, irreversible for ziftomenib after a 4-week recovery period 2.
| kind | toxicity |
|---|---|
| severity | embryo-fetal lethality and skeletal/craniofacial malformation at 0.3-0.5x clinical AUC in two species |
| mitigation | contraception during treatment and for 3-6 months after (on both labels); no postnatal craniofacial or vertebral monitoring is required or performed |
| mechanistic link | Pax3/Wnt1 neural-crest menin deletion reproduces the same lesion set (cleft palate, palatine, suture, rib, vertebral), via decreased p27(kip1) |
| link published | false |
affects
[ "drug-revumenib", "drug-ziftomenib" ]
- Both approved menin inhibitors are skeletal and craniofacial teratogens below the clinical exposure, in two species, and the genetics say it is on-target rather than off-target chemistry.
- Approval from 1 year of age places the drug in a window where craniofacial sutures, vertebral ossification and palatal growth are still incomplete, and nobody follows any of them.
- Irreversible animal fertility findings in a class given to children make gonadal function a late effect this beat has not costed, and no menin-inhibitor trial collects a single reproductive endpoint.
- Is the lesion-level concordance between the ziftomenib embryo-fetal findings and the neural-crest Men1 phenotype real? Comparing the sponsor's fetal skeletal examination data against the knockout phenotype would settle it and needs no new animals.
- Do infants dosed from 1 year of age have any craniofacial, suture or vertebral follow-up? Nothing in either label or any registered trial requires it.
- Is the fertility effect - irreversible in ziftomenib animals after recovery - measured in any adolescent who has received a menin inhibitor?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin inactivation in Pax3- or Wnt1-expressing neural crest causes perinatal death, cleft palate and cranial bone defects with decreased p27(kip1); deletion in Pax3+ somite precursors causes rib patterning defects.
Menin is required in cranial neural crest for palatogenesis and perinatal viability. · Engleka KA, Wu M, Zhang M, Antonucci NB, Epstein JA · Developmental Biology 2007What this evidence cannot show
Mouse conditional genetics, complete deletion from early development; no pharmacological arm; the p27 association is correlative within the paper.
Evidence that argues against this (1)
- Complete embryonic deletion is not comparable to postnatal partial pocket occupancy, and no postnatal craniofacial phenotype has been reported in any menin-inhibitor-treated animal or child. Syndax Pharmaceuticals et al. 2026
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
Ziftomenib caused cleft palate, split palatine, supernumerary suture bone, rib, costal-cartilage, vertebral and sternebral defects plus embryo-fetal lethality in pregnant mice at ~0.3x the clinical AUC, and irreversible fertility findings after a 4-week recovery.
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 summary of a gestation-day 6-15 study at 6, 20 and 60 mg/kg/day; the dose at which each malformation appeared is not broken out; the fertility statement is a label summary without the underlying study.
Evidence that argues against this (1)
- Cleft palate is among the commonest non-specific findings in mouse developmental toxicology and can follow maternal toxicity alone; decreased maternal body-weight gain is reported alongside. Syndax Pharmaceuticals et al. 2026
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
Revumenib caused embryo-fetal mortality, malformations and altered fetal growth in rats at ~0.5x the human AUC; the paediatric safety population included 25 infants under 2 years and revumenib is approved from 1 year of age.
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; the malformations are not enumerated in the same detail as in the ziftomenib label, so the lesion-level concordance with the neural-crest phenotype can only be asserted for ziftomenib.
Evidence that argues against this (1)
- No craniofacial or skeletal malformation has been reported in any treated child; the paediatric experience is 132 children with a median exposure of 2.5 months, which is short relative to a developmental readout. Syndax Pharmaceuticals et al. 2026
The enumerated embryo-fetal malformations on the KOMZIFTI label (cleft palate, split palatine, suture bone, rib, costal cartilage, vertebrae, sternebrae) are lesion-for-lesion the Pax3/Wnt1 neural-crest Men1 knockout phenotype; the link is unpublished
Embryo-fetal mortality, malformations and altered fetal growth at ~0.5x human AUC, in a drug approved from 1 year of age with 25 infants under 2 in the safety population
- ← related toWhat the labels measure in a treated child: bone growth, in one sentence, and nothing else endocrine
Irreversible animal fertility findings and skeletal teratogenicity below clinical exposure, with no reproductive or craniofacial endpoint collected in any treated child