Skip to content
type1.science

Icovamenib (BMF-219) — oral menin inhibitor

Biomea Fusion

The strongest human regeneration signal so far — from five people.

A once-daily pill that blocks menin, a protein that acts as a brake on beta-cell growth. It has produced the largest and most durable C-peptide signal yet from a drug whose stated mechanism is beta-cell regeneration: in the Phase 2 COVALENT-112 trial, 12 weeks of dosing raised stimulated C-peptide by about 52% in people diagnosed within three years; 40 weeks after the drug stopped that gain itself was gone, but C-peptide was still only about 7% below baseline — against the roughly 47% annual decline reported in historical placebo cohorts. The catch is size and rigour — five people at that dose, open-label, no placebo arm ever completed, and a programme that spent nearly four months on a full FDA clinical hold over possible drug-induced liver injury.

Years awayEarly evidenceregenerationmeninicovamenibbmf-219biomeaoralsmall-moleculebeta-cellc-peptidephase-2immunotherapy-neededOfficial site ↗

The scorecard

Regrowth efficacy35

The largest human C-peptide gain reported for a regeneration-mechanism drug — mean stimulated C-peptide AUC about 52% above baseline at Week 12 on 200 mg in people diagnosed within 3 years — but n=5, open-label, with no concurrent placebo arm and no peer-reviewed publication.[1]

Durability30

At Week 52 — 40 weeks after the last dose — mean C-peptide AUC was only about 7% below baseline, against the roughly 47% annual decline reported in historical placebo cohorts. A striking hint of durability, but from tiny, uncontrolled numbers against a historical comparator, not a control arm.[2]

Safety40

This drug has a liver-safety history. On 6 June 2024 the FDA placed the entire BMF-219 diabetes programme — type 2 and type 1 — on a full clinical hold over possible drug-induced hepatotoxicity seen in the completed type 2 dose-escalation stage. The hold was lifted on 26 September 2024. Dosing in T1D was otherwise tolerated through 52 weeks with no evidence of systemic immune activation, but long-term liver safety in T1D rests on tiny numbers.[5]

Eligibility breadth55

An oral pill, so in principle deliverable to many people — but it acts on residual beta cells, and the trial required detectable C-peptide (≥0.2 nmol/L if diagnosed within 3 years; ≥0.08 nmol/L if 3–15 years) in adults aged 18–70.[3]

Maturity35

Further along in people than the rest of this category — a Phase 2 proof-of-concept with 52-week human data — but the trial is registered as terminated, its placebo-controlled part was never completed, results are only in company releases and a conference presentation, and the planned larger Phase 2 was not registered as of 14 July 2026.[3]

The full picture

Menin is a protein that sits inside beta cells and acts as a brake: it holds the cells out of the cell cycle and damps down some of the machinery they use to make insulin. Icovamenib (also called BMF-219) is a once-daily pill that binds menin covalently and takes that brake off for a while. The idea is the same one behind the rest of this category — get the body to rebuild its own insulin-producing capacity rather than importing cells from a donor or a stem-cell line — but the drug got further into humans than the others have.

What the trial did

COVALENT-112 (NCT06152042) was an open-label Phase 2 study in adults aged 18–70 who had T1D and still had some measurable insulin production of their own: entry required a C-peptide of at least 0.2 nmol/L for people diagnosed within three years, or 0.08 nmol/L for those diagnosed three to fifteen years ago.1 Participants took icovamenib for just 12 weeks, then were followed drug-free for another 40 weeks.

What it found

In the group diagnosed within the previous three years and dosed at 200 mg, mean stimulated C-peptide AUC — the standard measure of how much insulin the pancreas can still produce on demand — was about 52% above baseline at Week 12.2 Then the drug stopped. Forty weeks later, at Week 52, mean C-peptide AUC was still only about 7% below baseline, against the roughly 47% annual decline reported in historical placebo cohorts.3 In other words, three months of a pill appeared to buy a year of preserved function.

At the American Diabetes Association's 86th Scientific Sessions (5–8 June 2026), the company added translational data: cytokine profiling showed no evidence of systemic immune activation, with inflammatory markers stable or reduced through Week 52.3 (An HbA1c benefit was also reported at those sessions — but in the separate type 2 diabetes study, not in T1D. It says nothing about this drug in T1D.)

Now the honest part

The 200 mg cohort was five people. The 100 mg cohort was six; the longer-diagnosed cohort was nine. The study was open-label. The randomized, placebo-controlled Part 2 was never completed, so roughly half the intended population was never studied.2 The 47% comparator is a historical figure, not a control arm in this trial. Every number above comes from company press releases and a conference presentation — there is no peer-reviewed publication and no results posted on ClinicalTrials.gov.1 And the registry lists the trial as terminated, with the reason given as a sponsor business decision about portfolio priorities.1

The liver-safety history

This is the part a person with T1D should read before anything else on this page. On 6 June 2024 the FDA placed the entire BMF-219 diabetes programme on a full clinical hold — COVALENT-111 in type 2 diabetes and COVALENT-112, the type 1 trial described above. The stated reason was possible drug-induced hepatotoxicity: liver injury signals seen in the completed dose-escalation phase of the type 2 study.4 Dosing and enrolment in the T1D trial stopped.24

The hold was lifted on 26 September 2024, roughly sixteen weeks later, and the diabetes trials resumed.5 We have no peer-reviewed account of what the liver findings actually were, how many people had them, or how severe they got — only the company's announcements of the hold and its lift. A resolved hold is not a disqualifying one, but "the FDA halted this drug over possible liver toxicity" is a permanent part of its record, and any future trial in T1D will have to clear that bar again. A drug whose whole mechanism is telling cells to divide is one to watch closely.

The unsolved problem is the same one

Even if icovamenib regrows beta cells, the immune system that destroyed the originals is still there. Biomea's own plan concedes the point: the announced next study includes an arm combining icovamenib with an immunosuppressive agent, alongside longer dosing (6–12 months at 200 mg) at the Barbara Davis Center, Joslin, UT Health San Antonio and the University of Miami.2 That trial was targeted for the second half of 2026 and, as of 14 July 2026, is not registered on ClinicalTrials.gov — so we are not putting it on the calendar.

How we rank it

Icovamenib sits at the top of the regeneration category, but not by much and not for the reason a press release would suggest. Other entries here have human C-peptide data too — Diamyd's GABA/remygen work produced a randomized trial (which was negative), and the semaglutide new-onset series showed C-peptide rising. What icovamenib has is the largest effect size and the longest apparent durability of any regeneration-mechanism drug tested in people, from a trial small enough that a single unusual responder could move the average. That is a genuine signal and a genuinely weak evidence base at the same time, and both facts belong in the score: efficacy and durability move up, maturity stays low. If the confirmatory Phase 2 opens, enrols properly, and includes a placebo arm, this record will change quickly — in one direction or the other.

References

  1. Phase 2 Randomized Trial of BMF-219 in Participants With Type 1 Diabetes Mellitus (COVALENT-112, NCT06152042) — registered as terminated; no results posted as of 14 July 2026. https://clinicaltrials.gov/study/NCT06152042 ↩ ↩2 ↩3

  2. Biomea Fusion. Positive 52-Week Results from Phase 2 COVALENT-112 Trial in Type 1 Diabetes (2026-04-27) — company press release. https://www.globenewswire.com/news-release/2026/04/27/3282109/0/en/Biomea-Fusion-Announces-Positive-52-Week-Results-from-Phase-2-COVALENT-112-Trial-in-Type-1-Diabetes-Showing-C-Peptide-Improvement-and-Durability-Following-12-Weeks-of-Icovamenib-Tr.html ↩ ↩2 ↩3 ↩4

  3. Biomea Fusion. New Clinical and Translational Data for Icovamenib at the ADA 86th Scientific Sessions (2026-06-05). https://www.globenewswire.com/news-release/2026/06/05/3307662/0/en/Biomea-Fusion-Presents-New-Clinical-and-Translational-Data-for-Icovamenib-at-the-American-Diabetes-Association-ADA-86th-Scientific-Sessions-and-Announces-Expansion-of-Ongoing-Phase.html ↩ ↩2

  4. Biomea Fusion Announces BMF-219 in Diabetes Placed on Clinical Hold (2024-06-06) — full hold on COVALENT-111 (type 2) and COVALENT-112 (type 1), citing possible drug-induced hepatotoxicity in the completed COVALENT-111 dose-escalation phase. https://investors.biomeafusion.com/news-releases/news-release-details/biomea-fusion-announces-bmf-219-diabetes-placed-clinical-hold ↩ ↩2

  5. FDA Lifts Clinical Hold on BMF-219 in Type 2 and Type 1 Diabetes Trials (2024-09-26). https://www.globenewswire.com/news-release/2024/09/26/2954087/0/en/FDA-Lifts-Clinical-Hold-on-BMF-219-in-Type-2-and-Type-1-Diabetes-Trials.html ↩

Coming soon

ETA · Phase 2 proof-of-concept complete (52-week data, April 2026), but the trial is registered as terminated and no confirmatory trial is open. A larger Phase 2 in recently diagnosed T1D was announced for the second half of 2026; it was not registered on ClinicalTrials.gov as of 14 July 2026, so we are not putting a date on it.

  • →Announced larger Phase 2 in recently diagnosed T1D at the Barbara Davis Center, Joslin, UT Health San Antonio and the University of Miami — extended dosing (6–12 months) at 200 mg · Targeted for H2 2026; not yet registered as of 14 July 2026
  • →Announced combination arm pairing icovamenib with an immunosuppressive agent, to protect regrown cells from the ongoing autoimmune attack
  • →Peer-reviewed publication and posted registry results — neither exists yet for the T1D data

Sources

  1. [1]Biomea Fusion Announces Positive 52-Week Results from Phase 2 COVALENT-112 Trial in Type 1 Diabetes · manufacturer · 2026-04-27 — Company press release, not a peer-reviewed publication. Reports the 52% Week-12 C-peptide rise (200 mg, n=5) and the ~7% decline from baseline at Week 52; also states the placebo-controlled Part 2 was not completed.
  2. [2]Biomea Fusion Presents New Clinical and Translational Data for Icovamenib at the ADA 86th Scientific Sessions · conference · 2026-06-05 — Source of the ~47% historical placebo decline comparator, and of the cytokine data showing no systemic immune activation through Week 52. The HbA1c result announced at the same sessions is from the separate type 2 diabetes study, not from T1D.
  3. [3]Phase 2 Randomized Trial of BMF-219 in Participants With Type 1 Diabetes Mellitus (COVALENT-112, NCT06152042) · registry — Registered as TERMINATED (sponsor business decision on portfolio prioritization). Phase 2, 37 participants, ages 18–70, entry required detectable C-peptide. No results posted as of 14 July 2026.
  4. [4]Biomea Fusion Announces BMF-219 in Diabetes Placed on Clinical Hold · manufacturer · 2024-06-06 — Source for the hold itself. The FDA placed a full clinical hold on both COVALENT-111 (type 2) and COVALENT-112 (type 1), citing possible drug-induced hepatotoxicity observed in the completed dose-escalation phase of COVALENT-111. Announced 6 June 2024; it says nothing about the later lift.
  5. [5]FDA Lifts Clinical Hold on BMF-219 in Type 2 and Type 1 Diabetes Trials · manufacturer · 2024-09-26 — Source for the lift, announced 26 September 2024 — the only citation on this record that supports the claim that the hold was lifted.