Menin leaves AML for the islet

In Short. Menin is one chromatin scaffold with two opposite clinical asks: break its partnership with KMT2A so AML blasts differentiate, and ease its growth brake in pancreatic islets so beta cells can expand. Caspian Therapeutics' $50 million spinout around KO-7246 is the market test of that second ask under chronic metabolic dosing.

---

This piece is about two human diseases that share a nuclear protein. In acute myeloid leukemia, approved drugs already break menin's grip so immature blasts can finish differentiating. In diabetes, a Lilly-backed spinout now wants to loosen that same protein's grip in the pancreas so insulin-producing beta cells can grow back. The argument sits on the first screen: menin is a chromatin scaffold reused across tissues, and the clinical verb flips with the tissue.

What is new on the tape is capital structure around a known node. On 9 September 2026, Kura Oncology spun Caspian Therapeutics with $50 million, led by BVF Partners and including Eli Lilly and the T1D Fund, to advance KO-7246, a menin inhibitor Kura describes as purpose-designed for chronic metabolic disease. Kura keeps about half of Caspian and the oncology franchise that already owns an AML label. If Caspian's bet holds, diabetes drug development moves a step toward restoring endogenous insulin capacity, beyond glucose management alone. If it fails, MEN1 genetics already supplies a failure mode worth teaching: releasing menin can mean endocrine overgrowth.

A scaffold that partners with chromatin writers

Menin sits in the nucleus and helps organize which genes stay readable. Its best-taught partnership in cancer is with KMT2A, also called MLL, a histone methyltransferase that marks chromatin for active transcription. In certain leukemias, that partnership locks on gene programs that keep blood cells stuck in a blast-like state. The drug idea is a protein-protein interaction break. Pull menin away from KMT2A, and the leukemic program loses a required scaffold. The cells are pushed toward myeloid differentiation, out of endless self-renewal.

Think of menin as a clip that holds a gene-control machine on a stretch of DNA. In the wrong blood cell, the clip keeps a growth program fastened. Remove the clip, and the machine falls off, which lets the cell finish maturing. That is the ELI5 version of a menin-KMT2A inhibitor in AML.

Two FDA approvals made that idea clinical hardware. On 15 November 2024, the agency approved revumenib (Revuforj, Syndax) for relapsed or refractory acute leukemia with a KMT2A translocation in patients one year and older (AUGMENT-101, NCT04065399). On 13 November 2025, it approved ziftomenib (Komzifti, Kura, with Kyowa Kirin) for adults with relapsed or refractory AML carrying a susceptible NPM1 mutation and no satisfactory alternative (KO-MEN-001 / KOMET-001, NCT04067336). In the NPM1 setting, the same menin-KMT2A node supports the mutant NPM1 oncogenic program. Labels for this class carry differentiation-syndrome risk for a reason: when blasts suddenly mature, the clinic has to manage the inflammatory and fluid consequences of that maturation.

So the oncology ask is already clear in plain language. Break the scaffold partnership. Let the blast grow up.

The same brake in the islet

Pancreatic islets use menin differently. People with multiple endocrine neoplasia type 1 inherit loss of MEN1, the gene that encodes menin, and develop endocrine tumors, including in the pancreas. That genetics is the first clue that menin's job in endocrine tissue includes growth restraint.

Karnik and colleagues made the chromatin mechanism teachable in 2005 (PMID 16195383, DOI 10.1073/pnas.0503484102). In mouse islets, menin associates with the promoters of genes encoding the cyclin-dependent kinase inhibitors p27 and p18, promotes histone H3 lysine-4 methylation there, and helps keep those brakes on the cell cycle. When menin is lost, those marks and those inhibitors fall, and islet growth escapes control. Company language about "taking the brakes off" pancreatic beta cells is that biology spoken aloud. It is the MEN1 logic turned into a dosing hypothesis.

A second ELI5 picture helps. In the islet, menin helps keep the "slow down" signs posted on the beta-cell cycle. Lift the signs, and the tissue can add insulin-making cells. The risk, already written in MEN1 families, is that lifting the signs too hard or too broadly invites endocrine overgrowth of the wrong kind.

The clinical verbs therefore oppose each other while the protein stays the same. In the marrow, menin's partnership helps keep a bad program fastened, so the drug wants the partnership broken so cells differentiate. In the islet, menin helps keep growth-inhibitory programs fastened, so a metabolic drug wants that restraint eased so beta-cell mass can expand. One scaffold. Two tissues. Opposite asks.

What Caspian is actually buying

Caspian is the vehicle for asking the islet question with dedicated capital and a molecule tuned, Kura says, for chronic metabolic use. Fierce Biotech and BioPharma Dive both report the $50 million round, BVF lead, Lilly and T1D Fund participation, roughly 50% Kura ownership, and KO-7246 as the lead. A second menin program for cardiometabolic disease rides with the spinout. Troy Wilson, Kura's CEO and Caspian's executive chairman, told BioPharma Dive that blocking menin can "take the brakes off pancreatic beta islet cells," and that Caspian needed its own expertise and capital because a company of Kura's size and stage has to focus while commercializing oncology.

Kura's CSO, Francis Burrows, said in the company release that preclinical findings point at addressing loss of functional beta-cell capacity across Type 1 and Type 2 models, with translation into human islet systems. Those claims remain company-stated until the European Association for the Study of Diabetes meeting in Milan later this month, where Kura plans to show the supporting data. This draft quotes no proliferation percentages the slides have not yet put on the public record.

Biomea Fusion already sits in mid-stage clinical testing with a metabolic menin inhibitor, which gives that program a calendar lead Wilson himself acknowledges. Caspian's counter-claim is chronic-use tuning and a preclinical safety story that the molecule appears to spare other cell types from abnormal growth. That is a hypothesis to test in humans. Selectivity is still an open measurement problem. The spinout is also exploring combinations with GLP-1 drugs for patients who still fail existing control. That is a development option waiting on clinical activity, and it is still an efficacy result to earn.

Notice what the deal does not have to invent. The AML labels already prove that menin-KMT2A disruption can change cell fate in humans under oncologic dosing. The MEN1 and Karnik literature already prove that menin restrains islet growth through chromatin and CDK inhibitors. Caspian is buying the right to connect those two proofs into one chronic metabolic program, with Lilly's optionality sitting in the cap table while Kura stays concentrated on Komzifti execution.

That reuse is the teachable object. Biology often sells targets as disease-specific nouns. Menin is a better lesson: a scaffold whose partners and chromatin jobs differ by tissue, so the right drug verb differs by tissue. Oncology already chose "break the partnership." Metabolism is choosing "ease the brake." The Caspian financing is interesting because it puts money behind the second verb with a molecule claimed for chronic use, while the first verb keeps earning under FDA labels.

What would make the reuse fail cleanly

The honest limit is the gap between a reused node and a safe chronic verb. Oncology menin inhibitors showed that breaking menin-KMT2A can differentiate genetically defined blasts under short, monitored courses, with a boxed differentiation-syndrome risk the clinic knows how to watch. Diabetes asks for something harder: months or years of dosing in people whose problem is missing or failing beta cells, aiming for more functional insulin capacity, while avoiding MEN1-style endocrine overgrowth.

A clean success would look like durable gains in endogenous insulin production tied to expanded, well-behaved beta-cell mass, in both autoimmune and insulin-resistant settings, without off-islet proliferative mischief. A clean failure would also teach. The brake may refuse selective lifting under chronic exposure. Expanded cells may lose function. The oncology hazard profile may refuse to shrink into a metabolic therapeutic window. Until EASD and first-in-human data arrive, Caspian's $50 million buys the right to run that experiment.

Menin remains at work in leukemia under approved labels. What expanded is the clinical imagination: the same nuclear partnership that keeps a blast immature may be the partnership an islet uses to stay small. Caspian is betting the second sentence is actionable under chronic dosing.

---

Sources

1. Fierce Biotech (2026-09-09), James Waldron — Caspian $50M / KO-7246. https://www.fiercebiotech.com/biotech/kura-creates-lilly-backed-spinout-take-forward-next-gen-diabetes-therapy
2. BioPharma Dive (2026-09-09), Gwendolyn Wu — menin islet framing / Biomea. https://www.biopharmadive.com/news/caspian-diabetes-kura-oncology-menin-inhibitors-lilly/829860/
3. FDA — Revumenib (Revuforj) approval 2024-11-15. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-revumenib-relapsed-or-refractory-acute-leukemia-kmt2a-translocation
4. FDA — Ziftomenib (Komzifti) approval 2025-11-13. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-ziftomenib-relapsed-or-refractory-acute-myeloid-leukemia-npm1-mutation
5. Karnik et al., PNAS 2005. PMID 16195383, DOI 10.1073/pnas.0503484102.