Follistatin-344: the human trials used gene therapy, not an injected peptide

Follistatin blocks myostatin, the protein that limits muscle growth, and blocking myostatin produces some of the most striking muscle results in biology. But the human work on follistatin-344 is not about injecting a peptide. It is about delivering the FS344 gene into thigh muscle with a virus, in two small trials totalling twelve patients, one of which drew a published rebuttal [5,6,7]. The vials sold online contain a protein with no published human data at all.

What it is

Follistatin is a natural protein, first sequenced from human tissue in 1988 [1]. That paper also explains the number: the gene produces a 344-amino-acid precursor, and alternative splicing produces a shorter 317-residue version [1]. FS344 is the longer form.

Its job is to mop up members of the TGF-beta family. It was identified in 1990 as the binding protein for activin [2], and later work showed it also binds myostatin, the muscle-limiting protein [3]. Myostatin normally signals through activin type II receptors to restrain muscle growth; follistatin intercepts it before it gets there [3].

The 2001 experiment that made follistatin famous was genetic. Transgenic mice engineered to make high levels of follistatin in skeletal muscle showed dramatic increases in muscle mass, comparable to mice with no myostatin at all [3]. The authors suggested follistatin and similar blockers "may be useful agents for enhancing muscle growth" for human and agricultural use [3]. That sentence has been doing marketing work ever since.

Note what the experiment was: mice whose muscles continuously produced follistatin from birth, from their own DNA. Not an injection.

The human evidence

Both human studies came from the same group at Nationwide Children’s Hospital in Ohio, and both used an adeno-associated virus vector, AAV1, carrying the FS344 gene, injected directly into the quadriceps.

Becker muscular dystrophy

Six men with Becker muscular dystrophy received bilateral injections into the quadriceps, three at a lower dose and three at a higher one [5]. The measure was the distance walked in six minutes. Two men in the first group improved by 58 and 125 metres, and one did not change; in the second group two improved by 108 and 29 metres and one did not [5]. Muscle biopsies showed less scarring and a more normal spread of fibre sizes. No adverse effects were reported [5].

Inclusion body myositis, and the objection

The second trial gave the same vector to six people with sporadic inclusion body myositis, a progressive inflammatory muscle disease, alongside an exercise programme [6]. The treated patients gained a median 56.0 metres a year on the walk test, while eight untreated comparison patients declined by 25.8 metres a year (p = 0.01) [6]. Four of the six improved by 58 to 153 metres; two barely changed [6].

That comparison group is where the trouble started. A neurologist at Brigham and Women’s Hospital published a critique in the same journal under the title "Unfounded claims of improved functional outcomes attributed to follistatin gene therapy in inclusion body myositis" [7]. Readers should look at that exchange before taking the headline number at face value. Six patients, no randomisation, an exercise co-intervention and a non-randomised comparison group is a design that can produce an apparent effect on its own.

What the primate work added

Before the human trials, the same vector was injected into the quadriceps of cynomolgus macaques, producing durable increases in muscle size and strength with no abnormal changes in major organs over long-term expression [4]. That study is the reason the human trials went ahead, and it is also gene delivery, not a peptide.

What the animal work shows, and its limits

The mouse results are the reason anyone cares about follistatin, so they deserve to be stated precisely.

Mice born without working myostatin have roughly twice the muscle mass of normal mice, and the same pattern appears in cattle, sheep, dogs and in one person with a natural loss-of-function mutation [8]. Adding a follistatin transgene to mice that already lacked myostatin pushed muscle mass to about four times normal [8]. That second result is the interesting one: it means follistatin blocks growth-limiting signals beyond myostatin alone, which is both why it is more powerful and why it is less selective.

A single injection of a myostatin-inhibitor gene, follistatin among them, raised muscle mass and strength in normal and dystrophic mice for more than two years, including in animals that were already old when treated [9].

All of these are genetic manipulations: transgenes present from conception, or one-off gene delivery. None involves repeated injections of follistatin protein, and none tells you what such injections would do.

The distinction that matters

There is a second problem with the injected version. Follistatin does not only block myostatin. It binds activins, which act across the reproductive system, the pituitary, inflammation and tissue repair [2]. Confining follistatin to muscle by injecting the gene into muscle sidesteps that. Circulating protein does not.

What happened to other myostatin blockers

The wider field is a caution in itself. ACE-031, a decoy activin receptor designed to soak up myostatin, was tested in boys with Duchenne muscular dystrophy in a randomised, placebo-controlled, ascending-dose trial [10]. There were trends towards more lean mass, less fat and better walking distance, none statistically significant. The study was stopped after the second dosing regimen because of nosebleeds and small dilated blood vessels in the skin, side effects that had nothing to do with muscle [10]. Those are the vascular consequences of interfering with this signalling family, and they showed up in a properly run trial rather than in anecdote.

Why "344" is on the label

Vendors use the number because the trials did. The Ohio group chose the FS344 splice form deliberately: they wrote that they used it "to avoid potential binding to off target sites", since the shorter FS317 form sticks to cell surfaces more readily [5]. That reasoning applies to a gene expressed inside muscle cells. Whether it means anything for a protein in a vial, nobody has tested. The number on the label is borrowed from a gene-therapy design decision.

Safety

There is no human safety data for injected follistatin protein. For the gene therapy, the two small trials reported no adverse effects [5,6], but twelve patients cannot establish safety, and gene delivery carries its own long-term unknowns.

The theoretical concerns follow from what follistatin does. Blocking activin signalling affects FSH release and reproductive function [2], and the ACE-031 experience shows this pathway has vascular effects [10]. Anyone selling follistatin for muscle growth is proposing to interfere with a system whose other jobs are not optional.

Legal and regulatory status

• US and UK: no follistatin product is approved as a medicine for any indication. The human work has been experimental gene therapy.
• Sport: prohibited at all times, twice over. Follistatin is named as a myostatin-binding protein under S4.3, agents preventing activin receptor IIB activation, on the 2026 WADA Prohibited List [11]. Delivering it as a gene would also fall under M3, gene and cell doping, which prohibits the use of nucleic acids to alter gene expression [11].

What we still don’t know

• Whether injected follistatin protein reaches muscle in useful amounts in humans, or does anything at all.
• Whether the gene-therapy walking improvements would survive a randomised, placebo-controlled trial. Neither study had one [5,6].
• What blocking activin signalling throughout the body does to fertility, hormones and blood vessels over time.
• What is actually in products sold under this name, since no regulator checks them.

Educational content only — not medical advice. Many peptides discussed on HPR are not approved for human use. Talk to a qualified clinician before making any decision about your health.