Bimagrumab (BYM338, later LY3985863) is a fully human IgG monoclonal antibody directed against the activin type II receptors (ActRIIA and ActRIIB). It was selected from the HuCAL phage-display library in a Novartis-MorphoSys collaboration. This is worth stating up front: although the compound is routinely listed alongside research peptides, it is not a peptide. It is an immunoglobulin of roughly 150 kDa, about a hundred times larger than a typical short peptide, with different pharmacokinetics, different stability and different handling requirements. It has no "amino-acid sequence" in the sense that peptide reagents list one.

Receptor-level blockade instead of ligand-level

This is the key difference between bimagrumab and the other compounds studied in a muscle-mass context. Skeletal muscle is held in check by a family of negative regulators from the TGF-β superfamily: myostatin (GDF-8), activin A and GDF-11. All of them signal through shared receptors, ActRIIA and ActRIIB.

Most experimental myostatin inhibitors target the ligand itself. Bimagrumab acts one level down: it binds the receptor with high affinity and physically occludes the docking site, so that none of those ligands can initiate a signal. In the preclinical work of Lach-Trifilieff and colleagues (2014), receptor-level blockade produced pronounced skeletal-muscle hypertrophy and protection from atrophy, an effect markedly stronger than neutralising myostatin alone.[1] A subsequent commentary in the literature specifically emphasised that dual ActRIIA/IIB blockade, not blockade of one receptor, is what is critical for maximal hypertrophy.

That is why bimagrumab became a reference tool for studying the ActRII axis: it switches off the whole node rather than one branch.

Clinical trajectory: three attempts at muscle pathology

Bimagrumab is one of very few compounds in this category with a large peer-reviewed clinical record. Novartis tested it systematically wherever muscle is lost:

  • Sporadic inclusion body myositis (sIBM). The pilot study by Amato and colleagues (2014) showed an increase in muscle volume after a single dose[2], and the FDA (US Food and Drug Administration) granted breakthrough therapy designation in 2013. However, the confirmatory phase 2b trial RESILIENT (Hanna et al., 2019; n=251) missed its primary endpoint: change in 6-minute walking distance at week 52 did not differ from placebo at any of the three doses (p=0.22 for 10 mg/kg).[3] The programme was discontinued.
  • COPD (chronic obstructive pulmonary disease) with low body mass index. In the randomized trial by Polkey and colleagues (2019), bimagrumab substantially increased thigh muscle volume and total lean body mass, but this did not translate into improved muscle function or functional performance.[4]
  • Sarcopenia. The interim study by Rooks et al. (2017, n=40) looked encouraging: thigh muscle volume rose 4.80% versus −1.01% on placebo (p=0.002), and a subgroup with slow gait showed gains in walking speed.[5] But the larger study (Rooks et al., 2020) compared bimagrumab against optimized standard of care, that is, adequate nutrition plus exercise, and missed its primary endpoint: SPPB improved 1.34 points versus 1.03 (p=0.13). Function improved in both groups; the antibody added nothing beyond what nutrition and exercise delivered.[6]

The recurring theme: mass without function

If one thing should be taken from bimagrumab's clinical record, it is this. Across three independent programmes, in three different pathologies, the compound consistently and reproducibly increased muscle mass, and consistently failed to improve what people could actually do. This is not an isolated failure and not a question of dose; it is the single most important empirical result of the whole programme.

It bears directly on how the compound should be read: mass gain measured by DXA (dual-energy X-ray absorptiometry) or MRI (magnetic resonance imaging) is not a substitute for strength, endurance or mobility. Bimagrumab is a rare case where that dissociation has been demonstrated in well-controlled trials across hundreds of participants rather than assumed.

The metabolic pivot

The second chapter is the one that rescued the compound. In a 48-week phase 2 trial (Heymsfield et al., 2021; n=75) in adults with type 2 diabetes and obesity, bimagrumab 10 mg/kg IV every 4 weeks produced:

  • fat mass: −20.5% (−7.5 kg) versus −0.5% on placebo (p<0.001);
  • lean mass: +3.6% (+1.70 kg) versus −0.8% (p<0.001);
  • body weight: −6.5% versus −0.8% (p<0.001);
  • waist circumference: −9.0 cm versus +0.5 cm (p<0.001);
  • HbA1c: −0.76 percentage points versus −0.04 (p=0.005).

Simultaneous fat loss with lean-mass gain is an atypical combination, and it is what changed the molecule's fate.[8] In 2023 Eli Lilly acquired Versanis Bio, which was developing it, and the compound became LY3985863.

Next came the phase 2b BELIEVE trial (Heymsfield et al., 2026; n=507, nine arms, 48 weeks plus a 24-week extension), studying bimagrumab alone and combined with semaglutide. At week 72 in the maximum-dose arms, the share of weight loss accounted for by fat was roughly 72% for semaglutide alone, 100% for bimagrumab alone and about 93% for the combination; lean mass changed by −7.4% on semaglutide, +2.5% on bimagrumab and −2.9% on the combination. That is the hypothesis keeping the compound in development: not weight loss as such, but the composition of the weight lost.[9]

The trajectory nonetheless remains uncertain. In 2025 Lilly withdrew one planned phase 2b combination trial with tirzepatide before enrolment began, citing "strategic business reasons"; other combination studies continue.

Safety and the limits of the evidence

Across the reported trials the most frequently mentioned adverse events are muscle spasms, diarrhoea and acne. In RESILIENT, muscle spasms occurred in 40–68% of treatment-arm participants versus 21% on placebo; serious adverse events in 17–33% across arms; two deaths were recorded, neither deemed treatment-related by the investigators. The safety profile in these trials was described as acceptable, but that applies to the specific populations studied and to intravenous dosing regimens, not to any other use.[7]

What can honestly be said about the limits of current knowledge:

  • There is no approved regulatory indication. As of 2026 the compound remains investigational.
  • The largest trials are phase 2 and 2b; there are no confirmatory phase 3 trials with clinical endpoints.
  • The metabolic data come predominantly from adults with obesity and type 2 diabetes; there is no basis for extrapolating them to healthy individuals or to an athletic context.
  • The long-term consequences of chronic ActRII blockade, including what years of silencing GDF-11 and activin A signalling outside skeletal muscle would mean, are not established.
  • All data relate to intravenous administration in a clinical setting.

Where it sits in the research landscape

Among compounds studied in a skeletal-muscle context, bimagrumab occupies an unusual position: it acts neither through the growth-hormone axis nor through IGF-1 signalling, but by releasing a brake, blocking the inhibitory TGF-β branch. Its head-to-head comparison with semaglutide in BELIEVE additionally made it a reference tool for the question of what exactly is lost during weight loss, a question that concerns the entire incretin-agonist class.

Reagent status. This monograph is reference material. As a reagent, bimagrumab is for laboratory research use only (RUO): it is not a medicine, not a supplement, not intended for human or veterinary consumption, and not for use in diagnosing, treating or preventing any condition. The figures above are published clinical trial results, not usage guidance.