ACE-083 is an experimental, engineered protein built from human follistatin, the body’s own binding protein for myostatin and activins. Acceleron Pharma designed it to be injected directly into a single weak muscle and to act only there, as a possible treatment for neuromuscular diseases in which particular muscles waste unevenly, such as facioscapulohumeral muscular dystrophy (FSHD) and Charcot-Marie-Tooth disease (CMT) (Pearsall et al., Sci Rep 2019; Glasser et al., Muscle Nerve 2018).
The single most important fact about ACE-083 is that it worked on the scan but not on the patient. In three placebo-controlled human studies it reliably enlarged the injected muscle by roughly 9–16%, yet in FSHD and CMT it did not improve walking, arm function, falls or quality of life compared with placebo. Acceleron stopped the FSHD program in September 2019 and discontinued ACE-083 entirely in March 2020 (Acceleron, Sept 2019; Acceleron, March 2020). It was never approved anywhere.
This page summarizes what the published trials, trial registries and Acceleron’s own regulatory filings report. It is for education only and is not medical advice.
| Property | Detail |
|---|---|
| What it is | Recombinant fusion protein: a shortened human follistatin (FST291) joined to the Fc domain of human IgG2, forming a dimer |
| Drug class | Myostatin/activin pathway inhibitor (“ligand trap”); Acceleron called it a “Myostatin+” agent |
| Developer | Acceleron Pharma (now a subsidiary of Merck & Co.) |
| Route studied | Intramuscular injection into one target muscle, under EMG or ultrasound guidance |
| Doses studied in humans | 50–200 mg per muscle (healthy volunteers); 150–240 mg per muscle every 3 weeks (FSHD, CMT) |
| Half-life | Not established in humans; in the phase 1 study serum ACE-083 was detectable only during the first day after injection, mostly near the assay’s lower limit |
| Ligands neutralized (in vitro) | Myostatin (GDF8), GDF11, activin A and activin B |
| Approval status | Never approved; development discontinued March 2020 |
| US controlled-substance status | Not listed in the anabolic steroid schedule (21 CFR 1308.13) |
| WADA status | Not named, but falls within S4.3 (myostatin-binding proteins such as follistatin), prohibited at all times |
What it is
Follistatin is a natural protein that traps several members of the TGF-β family that hold back muscle growth, including myostatin and the activins. It comes in forms that differ in how strongly they stick to the matrix around cells: the long form (FST315) circulates in blood, while the short form (FST288) binds heparan sulfate on cell surfaces and tends to act locally (Pearsall et al., Sci Rep 2019). The follistatin-344 page covers the natural protein and the gene-therapy trials that used it.
ACE-083 was engineered to exploit that local-binding behavior. Compared with native FST315, almost all of the C-terminal domain was removed, which exposes the heparin-binding site, and the remaining follistatin (FST291) was fused to a human IgG2 Fc fragment. The Fc fragment pairs two follistatin chains into one molecule, which doubles the number of heparin-binding sites (Pearsall et al., Sci Rep 2019). In laboratory tests:
- ACE-083 needed a higher salt concentration to be washed off a heparin column than a full-length FST315-Fc protein (938 vs 792 mM), and it bound an extracellular-matrix extract much more strongly.
- It bound activin A, activin B, myostatin and GDF11 tightly, with dissociation constants of roughly 10–350 pM.
- In a human cell-based reporter assay it blocked signaling by all four ligands, with IC50 values of about 40–700 pM (5–80 ng/mL).
The stated design goal was a drug that stays longer in the injected muscle and is quickly inactivated once it reaches the bloodstream (Glasser et al., Muscle Nerve 2018). In other words, it was meant to grow one muscle without the whole-body effects of a systemic myostatin or activin blocker.
How it works in the body
Myostatin, GDF11 and the activins signal through activin type II receptors and the Smad2/3 pathway to restrain muscle growth. By trapping these ligands outside the cell, ACE-083 removes that brake in the tissue where it is injected. In mice, injected muscles showed less phosphorylated Smad3 and more phosphorylated Akt, a growth-promoting signal (Pearsall et al., Sci Rep 2019).
Human pharmacokinetics
Only limited human exposure data exist. In the phase 1 study, an initial immunoassay was abandoned because it picked up inactive breakdown products. A more specific mass-spectrometry assay (lower limit of quantitation 10 ng/mL) was then applied to the two tibialis anterior cohorts (Glasser et al., Muscle Nerve 2018):
| Measure | Result |
|---|---|
| Samples above the detection limit | 14 of 252 (5.6%), from 6 of 12 treated women |
| Highest measured serum level, 100 mg dose | 13.3 ng/mL |
| Highest measured serum level, 150 mg dose | 37.2 ng/mL |
| Time of highest level | 3–24 h after injection |
| Detectable after the first day | No |
The authors concluded that there was little or no systemic exposure after the first 24 hours. Because so few samples were measurable, no half-life could be calculated. The 216-hour (9-day) half-life listed in some compound databases does not correspond to any published human measurement.
In the later phase 2 trials, laboratory markers that would signal systemic exposure to this drug class (hemoglobin and serum C-terminal collagen crosslinks) did not change (Statland et al., Muscle Nerve 2022; Thomas et al., Neurology 2022).
Development history and regulatory status
| Date | Event |
|---|---|
| Oct 2014 – Apr 2016 | Phase 1 study in 58 healthy postmenopausal women (NCT02257489) |
| Nov 2016 – Oct 2019 | Phase 2 trial in FSHD, 95 participants across both parts (NCT02927080) |
| Jul 2017 – Mar 2020 | Phase 2 trial in CMT types 1 and X, 63 participants (NCT03124459) |
| Sept 16, 2019 | Acceleron announces FSHD topline results and discontinues ACE-083 in FSHD |
| March 9, 2020 | Acceleron announces CMT topline results and discontinues development of ACE-083 |
| March 2020 | Open-label extension study for both diseases terminated (NCT03943290) |
The September 2019 announcement stated that the FSHD trial met its primary muscle-volume endpoint but that the gain did not translate into statistically significant improvements in functional tests, and that Acceleron would not run further FSHD trials. The same release noted that ACE-083 was not approved for any use in any country (Acceleron, Sept 2019).
The March 2020 release reported the same pattern in CMT: a significant increase in total muscle volume, but no significant improvement in any functional or quality-of-life secondary endpoint, and Acceleron’s chief executive said the company’s myostatin-plus hypothesis had not produced the functional outcomes needed for clinically meaningful benefit (Acceleron, March 2020). The company’s next quarterly filing confirmed that development of ACE-083 was being discontinued (Acceleron 10-Q, Q1 2020). The registry now lists the sponsor as a wholly owned subsidiary of Merck & Co. (NCT03124459).
- No approval. ACE-083 never reached a marketing application in any country.
- Not a controlled substance in the US. It is a protein, not an anabolic steroid, and is not listed in Schedule III (21 CFR 1308.13).
- Prohibited in sport. The WADA list does not name ACE-083, but class S4.3 (agents preventing activin receptor IIB activation) covers myostatin-binding proteins, with follistatin given as an example, and the class is open-ended (“including, but not limited to”). S4.3 substances are prohibited at all times and are non-Specified (WADA 2026 Prohibited List).
What the research shows
Animal studies
In wild-type mice, ACE-083 injected into one calf muscle twice weekly for 4 weeks increased that muscle’s weight in a dose-dependent way, by 85% at the highest dose (100 µg), without changing the weight of the opposite calf, nearby muscles, body weight or follicle-stimulating hormone (Pearsall et al., Sci Rep 2019). The same dosing in the tibialis anterior (the shin muscle that lifts the foot) gave these results in mice:
| Mouse model | Injected-muscle weight | Force |
|---|---|---|
| Wild-type | ≈+75% vs vehicle | Twitch +52%, peak tetanic +40% vs vehicle |
| Trembler-J (CMT1A model) | +73% | Twitch +44%, tetanic +65% vs untreated side; ankle dorsiflexion torque +137% vs vehicle |
| mdx (Duchenne model) | +77% vs untreated side | Not summarized here |
In both wild-type and CMT mice, the gain in absolute force matched the gain in muscle size: force per unit of cross-sectional area (specific force) did not improve. These are animal results and did not predict the human outcome.
Healthy volunteers (phase 1, n = 58)
The first-in-human study was randomized, double-blind and placebo-controlled. Healthy postmenopausal women aged 45–75 received ACE-083 (42 women) or placebo (16) into one rectus femoris (front thigh) or one tibialis anterior, as one dose or two doses 3 weeks apart (Glasser et al., Muscle Nerve 2018).
| Muscle | Dose | Peak mean increase in injected-muscle volume |
|---|---|---|
| Rectus femoris | 200 mg × 2 | 14.5% ± 4.5% (3 weeks after last dose) |
| Tibialis anterior | 150 mg × 2 | 8.9% ± 4.7% (3 weeks after last dose) |
- The non-injected muscle on the other leg changed about as much as placebo-injected muscle, consistent with a local effect.
- Volume gains correlated with dose per gram of muscle (r = 0.77–0.78); doses over 1.0 mg/g were associated with increases over 5%.
- At 8 weeks after the last dose the increases at the highest doses were no longer statistically significant.
- Strength did not change significantly by hand-held or fixed dynamometry in either muscle.
Facioscapulohumeral muscular dystrophy (phase 2, n = 95)
FSHD is an inherited muscular dystrophy that weakens the face, shoulders, upper arms and lower legs, often asymmetrically. The trial had two parts (Statland et al., Muscle Nerve 2022):
- Part 1 (open label, 37 participants): 150, 200 or 240 mg per muscle into the biceps brachii or tibialis anterior every 3 weeks for up to 5 doses. Total muscle volume rose by more than 15% at 200–240 mg.
- Part 2 (randomized, double-blind, 58 enrolled, 55 evaluable): 240 mg per muscle or placebo into both biceps or both tibialis anterior muscles every 3 weeks for 27 weeks (9 doses), followed by 6 months of open-label ACE-083.
Results of Part 2 at day 190, ACE-083 minus placebo (least-squares mean, 90% CI):
| Group | Outcome | Difference | p |
|---|---|---|---|
| Biceps (14 vs 14) | Total muscle volume, primary endpoint | +16.4% (9.8 to 23.0) | under 0.0001 |
| Biceps | Contractile muscle volume | +23.3% (10.8 to 35.8) | 0.002 |
| Biceps | Upper-limb function score (PUL midlevel) | +2.9% (0.1 to 5.7) | 0.09 |
| Biceps | Elbow flexion strength (hand-held dynamometer) | +36.1% (11.8 to 60.5) | 0.02 |
| Tibialis anterior (13 vs 14) | Total muscle volume, primary endpoint | +9.5% (3.2 to 15.9) | 0.01 |
| Tibialis anterior | Fat fraction (absolute) | −2.7 points (−4.9 to −0.6) | 0.04 |
| Tibialis anterior | 6-minute walk distance | −5.3% (−12.0 to 1.4) | 0.20 |
| Tibialis anterior | Ankle dorsiflexion strength | −14.5% (−34.3 to 5.2) | 0.22 |
The biceps strength difference was not seen at other time points across 12 months and did not appear when placebo patients switched to ACE-083. There were no significant differences in the FSHD Health Index total score, and in the tibialis anterior group walking and stair-climbing tests did not improve. The trial and its extension were stopped early for failing the functional endpoints, and the authors state that the FSHD development program was discontinued on the basis of these results (Statland et al., Muscle Nerve 2022; NCT02927080).
Charcot-Marie-Tooth disease (phase 2, n = 63)
CMT is an inherited nerve disease; types 1 and X damage the myelin sheath of long nerves, causing weakness of the lower legs and foot drop. The trial enrolled adults with CMT1 or CMTX who had weak ankle dorsiflexion (Thomas et al., Neurology 2022):
- Part 1 (open label, 18 patients): 150, 200 or 240 mg per muscle into both tibialis anterior muscles every 3 weeks for up to 5 doses. Total muscle volume rose 12.6–14.2% across cohorts at day 106.
- Part 2 (randomized, double-blind, 44 treated: 23 ACE-083, 21 placebo): 240 mg per muscle into both tibialis anterior muscles every 3 weeks for up to 9 doses over ≈6 months, then open-label ACE-083.
Part 2 results at day 190 (per-protocol set, n = 40, ACE-083 minus placebo):
| Outcome | Difference (90% CI) | p |
|---|---|---|
| Total muscle volume, primary endpoint | +13.5% (4.9 to 22.1) | 0.0096 |
| Contractile muscle volume | +23.3% (7.2 to 39.4) | 0.02 |
| Fat fraction (absolute) | −3.1 points (−6.8 to 0.6) | 0.16 |
| Ankle dorsiflexion manual muscle test score | +0.3 (0.1 to 0.5) | 0.03 |
| Ankle dorsiflexion strength, hand-held dynamometer | Not significant | 0.13 |
| CMT Health Index total score | Not significant | 0.63 |
| Risk of first fall | HR 1.04 (0.50 to 2.18) | 0.93 |
| Risk of recurrent falls | HR 1.41 (0.72 to 2.74) | 0.40 |
The 6-minute walk and 10-meter walk/run improved similarly in both groups, which the investigators attributed to a placebo or learning effect, and placebo patients did not improve further after switching to ACE-083. The authors concluded that despite larger muscles, patients on ACE-083 did not improve functionally more than those on placebo, and that this failure led the sponsor to suspend development for CMT. Of 45 randomized patients, 23 left the study because the sponsor terminated it (Thomas et al., Neurology 2022).
Why bigger muscles did not mean better function
The investigators offered several possible explanations rather than a single answer (Statland et al., Muscle Nerve 2022; Thomas et al., Neurology 2022):
- this drug class may increase muscle volume without directly improving strength or function in diseased muscle;
- walking and arm tasks depend on many muscles, so enlarging one may not be enough;
- the trials were powered for muscle volume, not strength, and may have been too short;
- functional gains might require exercise alongside the drug;
- learning effects on repeated walk tests made both arms look better.
The mouse data point the same way: the extra force in injected muscles came only from extra size, not from better-quality muscle (Pearsall et al., Sci Rep 2019).
Side effects and risks
The three trials enrolled 216 participants in total (58, 95 and 63), including placebo recipients, all treated under medical supervision with EMG- or ultrasound-guided injections. No product has a label, so the data below are the full extent of what is known.
Injection-site and muscle effects
In the phase 1 study there were no serious adverse events, no grade 3 or higher events and no discontinuations for adverse events. Injection-site pain was nearly universal in both groups (Glasser et al., Muscle Nerve 2018):
| Related adverse event | ACE-083 (n = 42) | Placebo (n = 16) |
|---|---|---|
| Injection-site pain | 90% | 100% |
| Injection-site discomfort | 19% | 25% |
| Injection-site reaction | 14% | 6% |
| Injection-site hemorrhage | 12% | 0% |
| Pain in extremity | 43% | 44% |
| Muscle twitching | 19% | 19% |
| Myalgia (muscle pain) | 10% | 0% |
In FSHD, injection-site reactions (erythema, discomfort, swelling, itching, warmth), myalgia and peripheral or joint swelling were more common with ACE-083 than placebo; related events were grade 2 or lower and none was serious. One Part 1 participant stopped the drug because of myalgia and muscle swelling (Statland et al., Muscle Nerve 2022). In CMT, the most common related events were injection-site reactions, myalgia and pain in the extremity, and there were no drug-related serious adverse events (Thomas et al., Neurology 2022).
Nerve-related events
In the FSHD trial, one tibialis anterior patient on ACE-083 developed foot paresthesia, a partial peroneal nerve axonal injury and numbness of the top of the foot, all judged probably drug-related; they resolved with lasting, gradually improving weakness. A biceps patient stopped treatment because of tingling starting in the feet, also judged probably related (Statland et al., Muscle Nerve 2022). Repeated injections close to a peripheral nerve are an inherent part of how this drug was given.
Other discontinuations
In the CMT trial, two ACE-083 patients discontinued for events judged possibly related: one for moderate erythema and peripheral swelling, and one for newly diagnosed, moderately severe systemic lupus erythematosus during the open-label period (Thomas et al., Neurology 2022).
Anti-drug antibodies
The immune system frequently recognized ACE-083 as foreign:
| Study | Patients with anti-ACE-083 antibodies |
|---|---|
| Phase 1, healthy women | 3 confirmed positive; one with titers up to 2,560 |
| FSHD, Part 2 double-blind | 39.3% on ACE-083 vs 3.3% on placebo |
| CMT, Part 2 double-blind | 47.8% on ACE-083 vs 0% on placebo |
None of the papers linked these antibodies to clinical problems, but the trials were short and small (Glasser et al., Muscle Nerve 2018; Statland et al., Muscle Nerve 2022; Thomas et al., Neurology 2022).
What is unknown
- There are no data beyond about 12 months of treatment and no studies in athletes or in healthy men; the only healthy-volunteer data come from one or two doses in postmenopausal women. Nothing is known about combining ACE-083 with other drugs.
- Because the drug was discontinued, no regulator has reviewed its manufacturing, potency or long-term safety. Any material sold under this name outside the original trials has no verifiable link to the clinical-grade protein.
Key takeaways
- An engineered, locally acting follistatin. ACE-083 is a truncated human follistatin fused to an IgG2 Fc fragment, built to stick in the injected muscle and neutralize myostatin, GDF11 and activins A and B there.
- It reliably grew the injected muscle. Placebo-controlled trials showed volume increases of about 9–16% in healthy women, FSHD and CMT, with little measurable drug in the blood after the first day.
- It did not make patients function better. Walking, arm function, falls and quality of life did not improve versus placebo in either disease, and strength gains were inconsistent.
- The program is over. Acceleron dropped FSHD in September 2019 and discontinued ACE-083 in March 2020; it was never approved anywhere.
- Local side effects and frequent antibodies. Injection-site reactions and muscle pain were common, nerve-related events occurred in a few FSHD patients, and 39–48% of treated patients developed anti-drug antibodies.
- Banned in sport. As a follistatin-based myostatin-binding protein, it falls under WADA class S4.3, prohibited at all times.
Sources
- Pearsall RS, Davies MV, Cannell M, et al. Follistatin-based ligand trap ACE-083 induces localized hypertrophy of skeletal muscle with functional improvement in models of neuromuscular disease. Sci Rep. 2019;9(1):11392. PubMed.
- Glasser CE, Gartner MR, Wilson D, Miller B, Sherman ML, Attie KM. Locally acting ACE-083 increases muscle volume in healthy volunteers. Muscle Nerve. 2018;57(6):921–926. PubMed.
- Statland JM, Campbell C, Desai U, et al. Randomized phase 2 study of ACE-083, a muscle-promoting agent, in facioscapulohumeral muscular dystrophy. Muscle Nerve. 2022;66(1):50–62. PubMed.
- Thomas FP, Brannagan TH 3rd, Butterfield RJ, et al. Randomized phase 2 study of ACE-083 in patients with Charcot-Marie-Tooth disease. Neurology. 2022;98(23):e2356–e2367. PubMed.
- Acceleron Pharma. Acceleron announces topline results from the Phase 2 trial of ACE-083 in patients with facioscapulohumeral muscular dystrophy. Form 8-K, Exhibit 99.1. September 16, 2019. SEC.
- Acceleron Pharma. Acceleron announces topline results from the Phase 2 trial of ACE-083 in patients with Charcot-Marie-Tooth disease. Form 8-K, Exhibit 99.1. March 9, 2020. SEC.
- Acceleron Pharma. Quarterly report on Form 10-Q for the period ended March 31, 2020. SEC.
- Acceleron Pharma. Phase 1 study of ACE-083 in healthy subjects. NCT02257489. ClinicalTrials.gov.
- Acceleron Pharma. Study of ACE-083 in patients with facioscapulohumeral muscular dystrophy (FSHD). NCT02927080. ClinicalTrials.gov.
- Acceleron Pharma. Study of ACE-083 in patients with Charcot-Marie-Tooth disease. NCT03124459. ClinicalTrials.gov.
- Acceleron Pharma. Extension study to evaluate the long-term effects of ACE-083 in patients with facioscapulohumeral muscular dystrophy (FSHD) and Charcot-Marie Tooth (CMT) disease types 1 and X (CMT1 and CMTX). NCT03943290. ClinicalTrials.gov.
- World Anti-Doping Agency. Prohibited List 2026, class S4.3. WADA 2026 Prohibited List.
- Drug Enforcement Administration. Schedule III anabolic steroids. 21 CFR 1308.13. eCFR.