ACE-031

ACE-031 is the ligand-binding ectodomain of ActRIIB stitched to an IgG1 Fc domain, which gives it a dimeric structure and a long circulating half-life. Because it is the receptor itself, it captures everything ActRIIB normally binds: myostatin, activin A and B, GDF-11 and several BMPs. Blocking that whole set removes Smad2/3-mediated inhibition of muscle growth far more comprehensively than a myostatin-selective antibody would, which explains both its potency - healthy postmenopausal women gained lean mass on a single dose - and its downfall. Activin and BMP signalling maintain vascular integrity, and the Duchenne trials were stopped in 2011 after dose-dependent epistaxis, gum bleeding and telangiectasia appeared. That safety signal is the reason the entire field moved toward selective myostatin blockade.

Targets: Activin receptor type IIB (as decoy), Myostatin (GDF-8), Activin A, Activin B, GDF-11, BMP9/BMP10

• · Trials used 1-3 mg/kg. The microgram figures here are per kilogram of body weight, not absolute doses. This programme was discontinued; there is no current legitimate supply.

Work out your exact syringe units →

• commonEpistaxis (nosebleeds)— Dose-dependent and one of the two findings that ended the Duchenne programme.
• commonTelangiectasia (dilated surface capillaries)— The other programme-ending finding, reflecting BMP9/BMP10 pathway blockade in the vasculature.
• commonGum bleeding— Part of the same bleeding-diathesis picture.
• commonHeadache
• uncommonReduced FSH and reproductive hormone disturbance— Predictable from activin blockade.
• uncommonImmunogenicity / anti-drug antibodies— Standard risk with Fc-fusion biologics.

• Any bleeding disorder, anticoagulant use or hereditary haemorrhagic telangiectasia - the vascular signal is the defining hazard of this molecule.
• Active malignancy.
• Pregnancy and breastfeeding.
• Realistically: there is no legitimate way to obtain this compound, and grey-market material is not what it claims to be.

• redundantfollistatin-344 — Same axis, compounded off-target activin blockade.
• redundantbimagrumab — Both act at activin type II receptors, one as a decoy and one as an antibody.
• redundantapitegromab — Apitegromab exists specifically to hit myostatin selectively and avoid ACE-031's vascular problem; combining them defeats that.

• · Any bleeding, nosebleeds or new spider veins would be the signal to stop immediately.
• · Full blood count and coagulation screen.
• · FSH, LH and sex hormones if used for any period.

Never approved; development discontinued. Not legally available. Myostatin-pathway inhibitors are banned at all times in sport by WADA.

• Campbell et al. 2017, Muscle & Nerve - myostatin inhibitor ACE-031 in boys with Duchenne muscular dystrophy (trial)
• Phase 1 study of ACE-031 in healthy postmenopausal women showing increased lean mass (trial)
• Reviews of activin type II receptor biology and the vascular consequences of broad ligand trapping (review)

ACE-031 is the decoy-receptor version of the myostatin story, and it is the compound that taught the field why hitting this pathway broadly is dangerous. The extracellular domain of ActRIIB is fused to an IgG1 Fc, which does two things: it makes the trap bivalent and it gives it FcRn recycling, so a protein that would otherwise be filtered within hours persists for a mean of ten to fifteen days. Circulating ACE-031 binds anything that would normally dock at ActRIIB - myostatin, activin A and activin B, GDF11, and several BMPs - and prevents receptor engagement. No ActRIIB engagement means no ALK4/ALK5 transphosphorylation, no Smad2/3 signalling, no transcription of the FoxO-linked atrophy programme, and disinhibition of Akt-mTORC1. The result in humans was fast and unambiguous: a single subcutaneous dose at 3 mg/kg produced 3.3% lean body mass and 5.1% thigh muscle volume by day 29. That is a larger effect from one injection than most compounds in this class produce in a cycle. Then the Duchenne trial found the problem. Epistaxis and telangiectasia appeared, and the study was stopped after the second dosing regimen. The vascular signal is generally attributed to the breadth of the trap: ActRIIB ligands include BMP9 and BMP10, which are central to endothelial quiescence through ALK1 signalling, and disrupting that axis produces exactly the mucocutaneous telangiectatic phenotype seen in hereditary haemorrhagic telangiectasia. That is the mechanistic lesson the entire field took from ACE-031, and it is the reason bimagrumab targets the receptor with an antibody and apitegromab targets pro-myostatin selectively.

The half-life is the thing people fail to account for. Ten to fifteen days means a single injection commits you for six weeks of meaningful exposure, and if telangiectasia or epistaxis appears at week two there is nothing to do but wait. Grey-market ACE-031 is also a fusion protein produced in mammalian cells - about the hardest thing in this catalogue to manufacture correctly and the easiest to fake, which is why anti-doping laboratories have published detection methods for black-market material. The third failure is dose reasoning by analogy: the healthy-volunteer study used a single dose up to 3 mg/kg, which for an 90 kg person is 270 mg, and the Duchenne trial dosed every two to four weeks. Protocols circulating that use small fixed milligram doses weekly are not derived from anything. The fourth is ignoring the specific warning signs, because nosebleeds feel minor and telangiectasia looks cosmetic, and they were serious enough to stop a pharmaceutical programme in children who had a lot to gain.

• Myostatin (GDF-8) — High-affinity ligand capture by the ActRIIB ectodomain; exact Kd not published for the fusion

  Prevents ActRIIB engagement, abolishing Smad2/3-driven atrophy signalling.

• Activin A and activin B — High - activins are the native high-affinity ligands for ActRIIB

  Contributes substantially to the muscle effect and to the endocrine consequences, since activin regulates FSH.

• GDF11 — High; GDF11 and myostatin share receptor binding surfaces

  Blocked, with unclear consequences for the physiology GDF11 regulates.

• BMP9 and BMP10 — Meaningful binding to the ActRIIB ectodomain

  The likely explanation for the epistaxis and telangiectasia that stopped the Duchenne programme. BMP9/BMP10 signalling through ALK1 maintains endothelial quiescence.

• A single ascending-dose study of muscle regulator ACE-031 in healthy volunteers Phase 1 · n=48 · 2013

  Safety, tolerability and pharmacodynamics of a single subcutaneous dose of ACE-031 at 0.02-3 mg/kg versus placebo in healthy postmenopausal women, randomised 3:1. Mean half-life 10-15 days with dose-linear exposure. At 3 mg/kg, lean body mass rose 3.3% and thigh muscle volume 5.1% by day 29. Generally well tolerated, with injection-site erythema the main adverse effect.

• Myostatin inhibitor ACE-031 treatment of ambulatory boys with Duchenne muscular dystrophy (NCT01099761 and extension NCT01239758) Phase 2 · n=24 · 2017

  Randomised, double-blind, placebo-controlled ascending-dose study of subcutaneous ACE-031 every 2-4 weeks. Stopped after the second dosing regimen because of epistaxis and telangiectasia. Trends toward maintained six-minute walk distance, increased lean mass and bone density, and reduced fat mass were seen but the study was not completed. Both the main study and its extension are listed as terminated on ClinicalTrials.gov.

Measurable lean mass and thigh muscle volume change by day 29 from a single dose - unusually fast for anything in this class. Exposure persists for six weeks or more after one injection. The vascular signal in the Duchenne study emerged during the second dosing regimen, so weeks rather than months.

There is no sensible stacking argument for ACE-031 with any other ActRII-pathway agent - follistatin, bimagrumab, apitegromab, trevogrumab all converge on the same node, and combining them multiplies the vascular and endocrine risk without a rationale. The combination that is worth thinking about is with GLP-1 receptor agonists, because that is precisely the design the pharmaceutical industry is now testing with the safer members of this family: an incretin drives weight loss with an unwanted lean-mass cost, and an ActRII blocker offsets it. The BELIEVE and COURAGE trials are running that experiment properly, with a decoy-receptor drug that does not cause telangiectasia. Doing it yourself with ACE-031 means adopting the one mechanism the field abandoned. With growth hormone or IGF-1 analogues the mechanisms are genuinely independent - Smad2/3 removal versus Akt-mTORC1 activation - but with a ten-to-fifteen-day half-life you cannot back out of a bad interaction quickly.

Against bimagrumab: both block ActRII signalling, but bimagrumab is an antibody against the receptors rather than a soluble ligand trap, and it has not produced the telangiectasia and epistaxis signal across hundreds of subjects including a 507-participant obesity trial. That is the whole argument for the antibody approach. Against apitegromab: apitegromab binds only pro- and latent myostatin, leaving activin and BMP signalling entirely alone, which is why it has a clean safety profile across a phase 3 trial and ACE-031 does not. Against follistatin: similar breadth of ligand blockade, but ACE-031 has real human efficacy data - 3.3% lean mass from a single dose - and follistatin protein has none. Against every other compound in this class, ACE-031's phase 1 result is among the most impressive: one injection, 29 days, 3.3% lean mass. The reason nobody develops it is not that it failed to work.

ACE-031 was never approved and has no legitimate market price. Black-market material exists and has been characterised by anti-doping laboratories, but no reliable price could be resolved, and a vial of an uncharacterised Fc-fusion protein has no meaningful price-per-effect anyway.

• ?No amino acid sequence is given because the exact ActRIIB ectodomain boundaries and IgG1 Fc construct used in ACE-031 were not resolved to a primary source in this session.
• ?Numeric clearance and volume of distribution were never published; only the 10-15 day half-life and dose-linearity are available.
• ?Time to steady state is entered as approximately 60 days on the basis of four to five half-lives at the upper end of the reported range, not from measured accumulation data.
• ?The BMP9/BMP10 explanation for telangiectasia is the field's working hypothesis and is mechanistically well supported, but was not proven in the trial itself.
• ?Binding affinities of the fusion protein for individual ligands were not published in the sources resolved here.
• ?Whether black-market ACE-031 is the same construct as the clinical material is unknown; detection work confirms that material exists, not that it is authentic.

• A single ascending-dose study of muscle regulator ACE-031 in healthy volunteers Attie KM, Borgstein NG, Yang Y, Condon CH, Wilson DM, Pearsall AE, Kumar R, Willins DA, Seehra JS, Sherman ML, Muscle & Nerve, 2013 · PMID 23169607

  Source of the only human pharmacokinetics for this compound - 10-15 day half-life, dose-linear exposure - and of the 3.3% lean mass and 5.1% thigh muscle volume figures from a single injection.

• Myostatin inhibitor ACE-031 treatment of ambulatory boys with Duchenne muscular dystrophy Campbell C, McMillan HJ, Mah JK, Tarnopolsky M, Selby K, McClure T, Wilson DM, Sherman ML, Escolar D, Attie KM, Muscle & Nerve, 2017 · PMID 27462804

  The trial that ended the programme, and the reason to take epistaxis and telangiectasia seriously rather than as curiosities.

• Gel Electrophoretic Detection of Black Market ACE-031 Reichel C, Drug Testing and Analysis, 2025 · PMID 40312924

  Direct evidence that ACE-031 circulates on the black market and that anti-doping laboratories have had to build assays for it. Worth reading before assuming a vial is what it says.

• ACE-031, a soluble activin type IIB receptor, increases muscle mass and strength in the common marmoset Cadena SM, et al., PLoS ONE, 2026 · PMID 41686840

  Non-human primate pharmacology confirming the muscle mass and strength effect, published long after the clinical programme ended.