AOD-9604: The hGH Fragment That Never Cleared Its Human Trials

AOD-9604 corresponds to residues 176-191 at the C-terminus of human growth hormone, with a tyrosine appended at the N-terminus. Its molecular formula is C78H123N23O23S2 and its molecular weight is approximately 1,815 g/mol; it resolves in PubChem under CID 71300630 and carries CAS 221231-10-3. The two sulfur atoms in the formula reflect a pair of cysteines that form an internal disulfide bond, so the peptide carries a small loop rather than being a linear chain.

The compound is therefore an excerpt, not an invention. Human growth hormone is a 191-residue protein with several functionally distinct regions, and AOD-9604 reproduces only the last sixteen of them plus one addition. The name reflects its history as a development candidate - AOD for 'anti-obesity drug' - rather than any structural feature.

The added N-terminal tyrosine is a synthetic-chemistry convenience with a real consequence: it provides an aromatic residue that supports radioiodination and simplifies detection, which is why the analytical literature refers to the compound as a tyrosine-modified fragment rather than the bare 176-191 sequence.

Growth hormone does several things that are hard to separate in a clinical setting. It promotes longitudinal growth and lean tissue accretion, largely through IGF-1; it antagonises insulin action; and it mobilises fat. The first two are what make growth hormone unattractive as an approach to adiposity - insulin resistance and unwanted growth signalling arrive with the effect being sought.

The premise of AOD-9604, set out by Ng and colleagues, is that these activities are structurally separable. If the lipolytic activity lives in a discrete C-terminal domain, then a fragment containing only that domain should reproduce fat mobilisation without engaging the growth hormone receptor in the way that drives IGF-1 production and insulin antagonism. The literature reports that the fragment does not produce the growth-promoting or insulin-antagonising effects of the intact hormone.

This is an elegant idea and it has never been confirmed in humans in a published controlled trial. That combination - a clean hypothesis with strong rodent support and no human confirmation - describes the compound accurately and should frame everything that follows.

The substantive efficacy work is a pair of 2001 papers from Heffernan and colleagues, both in mice. The first, in the International Journal of Obesity, reported that chronic administration of either human growth hormone or the modified C-terminal fragment increased fat oxidation and reduced body weight in obese mice. Running both arms in the same experiment is what makes this study useful: it tests the dissociation claim by comparison rather than assuming it.

The second, in Endocrinology, went after mechanism using β3-adrenergic receptor knockout mice. The β3-adrenergic receptor is the principal adrenergic receptor mediating lipolysis in rodent adipose tissue, and the knockout comparison implicates it in the response to the fragment. This is the only defined pathway the AOD-9604 literature offers.

Two limitations deserve emphasis rather than a footnote. First, β3-adrenergic receptor biology differs substantially between rodents and humans - the receptor's role in human adipose tissue is considerably more modest, and a mechanism that runs through it in mice does not transfer automatically. Second, both studies used obese mouse models, in which almost any intervention that increases energy expenditure produces measurable weight change. Neither observation invalidates the findings; both constrain how far they can be carried.

AOD-9604 was a genuine obesity development candidate. The 2004 review by Wilding in Current Opinion in Investigational Drugs describes it in exactly those terms, at a point when the programme was active and the expectation was that clinical evidence would follow.

It did not follow into the indexed literature. Searching PubMed for a completed randomised controlled trial of AOD-9604 in humans returns nothing of that kind. What exists instead are early metabolic studies, rodent work, review articles from the development period, and - increasingly, after roughly 2014 - analytical chemistry papers concerned with detecting the compound in biological samples.

That shift in the character of the literature is itself informative. The work by Cox and colleagues on detection methods and in vitro metabolism was motivated by anti-doping testing requirements, and a separate strand of publications concerns identification of the peptide in seized or unlabelled preparations. When the most recent substantial research on a compound is about how to detect it rather than what it does, the therapeutic programme has effectively concluded without producing a positive result. Absence of a published trial is not proof of failure, but it is not neutral either.

Human efficacy is the largest gap and it is close to total. Whether the fat-oxidation effect observed in obese mice occurs in humans at any exposure is not established in the published literature, and the mouse-to-human step is particularly uncertain here because the implicated receptor has a different weight in human adipose physiology than in rodent.

The receptor interaction is itself incompletely defined. The literature reports what AOD-9604 does not do - it does not appear to reproduce growth hormone receptor-mediated growth signalling - more clearly than it reports what it does. There is no well-characterised binding partner in the sense that VEGFR2 is characterised for BPC-157 or the incretin receptors are for tirzepatide; the β3-adrenergic receptor involvement is inferred from a knockout phenotype rather than from direct binding.

Human pharmacokinetics are unpublished, long-term toxicology in humans is absent, and no controlled human safety dataset of meaningful duration exists. The identification literature adds a practical concern: preparations circulating outside controlled channels have been analysed and found to be of variable and sometimes incorrect composition, which makes reported experience impossible to attribute to the compound at all.