The same pocket suppresses JUND: menin inhibition's unmeasured endocrine liability in infants

When this entry was written it asserted that endocrine parameters were not monitored and that "nobody has looked". Both claims are false. The endocrine beat found the primary documents and the coordinator verified each one directly — DailyMed and the FDA approval letter, read in full, not from a summary 34.

The REVUFORJ label already reports the signal. Among 241 patients: intact PTH increased 34%, phosphate increased 51% and decreased 25%, corrected calcium increased 15% (0% grade 3–4), alkaline phosphatase increased 33%. "Hyperparathyroidism" is listed under Endocrine disorders among clinically relevant adverse reactions occurring in under 20% 3.

The FDA already compelled the study, in these words. Postmarketing requirement 4730-1 under FDCA 505(o)(3) states that spontaneous reporting "will not be sufficient to assess ... signals of serious risks of elevated parathyroid hormone", and requires a clinical trial to determine whether long-term use carries "serious potential risk of endocrine-related disorders due to chronic menin inhibition". All patients must have at least 3 years of treatment. Final report due 05/2028 4.

So the regulator reached this conclusion before this project did, named the mechanism in almost the same terms, and the answer arrives in 2028.

What remains true, and is narrower. The PMR has no paediatric endpoint, no defined analyte panel and no off-treatment follow-up, and there is no equivalent requirement for ziftomenib or bleximenib. And the signal reached the label but not the literature: the pivotal publication's adverse-event table contains no parathyroid, calcium or phosphate entry at all, so a clinician reading the published trial would not know any of this.

One paediatric anchor is stronger than anything this entry originally carried. Label §8.4 reports an irreversible increase in femur growth plate closure in juvenile rats dosed 13 weeks, at approximately 2 times the human AUC, and instructs prescribers to monitor bone growth and development in paediatric patients 3. Reversible exposure does not imply reversible effect.

And a fact this project should not overlook: the same approval letter waives the paediatric study requirement for patients under 1 year of age, on the grounds that such studies are "impossible or highly impracticable" given the small number and geographical dispersion of those patients 4. Infants under one are precisely this project's population, and they are the group formally exempted from study.

Menin is genuinely two-faced. In the endocrine lineage it is a tumour suppressor: biallelic loss causes multiple endocrine neoplasia type 1, a hereditary syndrome of parathyroid, pancreatic islet and pituitary tumours. In the haematopoietic lineage it acts as an oncogenic cofactor, holding KMT2A fusions on chromatin. The whole therapeutic argument of this project runs through the second role.

The crystal structures settle how both work, and the answer is uncomfortable: there is one pocket, not two. Menin binds short peptides of MLL1 and of JUND in the same manner, in the same deep pocket, with opposite effects on transcription. The menin-JUND interaction blocks JNK-mediated JUND phosphorylation and suppresses JUND-induced transcription; the menin-MLL1 interaction promotes transcription. Several MEN1 missense mutations found in patients disrupt the menin-JUND interaction, which is what ties that interaction to the tumour-suppressor function 1.

A menin inhibitor is a pocket occupant. On the structural evidence it should displace JUND as readily as it displaces MLL1. That is not a criticism of the drugs — it is how they work — but it means the endocrine tumour-suppressor function and the leukemic cofactor function are being disrupted by the same binding event, and cannot be separated by dose or by schedule.

The standard reassurance is that menin and MLL1 are genetically separable, so blocking their interaction should not obliterate normal blood formation. That result is real: menin and MLL1 regulate largely distinct pathways in haematopoietic stem cells and developing B cells, with only a small set of co-regulated genes, and the authors explicitly concluded it supports developing menin-MLL1-disrupting drugs as safe and selective 2.

But that experiment was done in haematopoiesis. It establishes a window in blood. It says nothing about whether occupying menin's pocket for months is safe in parathyroid, islet or pituitary tissue, where menin's job is the opposite one. The graph has been treating a haematopoietic safety result as though it were a general one, and it is not.

This is the same shape as the anthracycline problem this project already tracks: an acceptable acute profile in a population whose defining feature is that they will live for decades afterwards.

• Endocrine surveillance in treated survivors. Serum calcium, PTH, prolactin, IGF-1 and fasting gastrin are cheap, and none of them appears in the published toxicity tables, which report febrile neutropenia, differentiation syndrome and QTc.
• Duration and reversibility. MEN1 syndrome arises from lifelong biallelic loss followed by somatic second hits over years. Months of reversible pharmacological occupancy is a very different exposure, and the honest position is that nobody knows where the threshold is — or whether there is one.
• A JUND-sparing ligand. If a chemotype could occupy the pocket in the MLL1-blocking orientation while leaving JUND displacement weaker, that would be a genuine selectivity axis. The structures suggest the peptides bind in the same manner, so this may be impossible, and demonstrating that it is impossible would also be worth knowing.
• The relevant natural experiment already exists: MEN1 mutation carriers are heterozygous and tumour-free for decades before a somatic second hit. Their trajectory bounds how much menin function must be lost, and for how long, before endocrine neoplasia follows.