AOD-9604 — the human growth hormone fragment Australian biotech designed to skip the anabolic side

Ng's question was whether you could separate them. Whether the fat-metabolism signal and the growth signal lived in different parts of the molecule.

It wasn't a naive question. By the early 1990s, researchers had been mapping the functional regions of human growth hormone for years, and there were suggestive findings. The hGH molecule is 191 amino acids long, a relatively large peptide hormone, and different segments of it appeared to engage different receptors and signaling pathways. The growth-promoting, IGF-1-stimulating effects seemed to cluster around a different region of the molecule than the lipolytic — fat-breakdown — signaling. Ng's group at Monash focused on the C-terminal end of the molecule: specifically, amino acids 176 through 191, the last sixteen residues of the chain. The fragment was synthesized with an additional tyrosine residue at the N-terminus for stability and practical synthesis purposes, making it a 15-amino-acid fragment plus the added tyrosine — a compound that would eventually be called AOD-9604, with AOD standing for anti-obesity drug.

The logic was elegant and, at the time, genuinely exciting. If the lipolytic activity of growth hormone was concentrated in this C-terminal fragment, you might be able to activate fat metabolism without engaging the GH receptor in the growth-promoting, IGF-1-driving, glucose-dysregulating way. You'd have a molecule that mimics the part of GH that burns fat while leaving the parts that cause insulin resistance and anabolic side effects behind. In animal research conducted through the 1990s, this hypothesis had enough support to attract serious investment. Obese mice given the fragment showed fat loss. Fat cell activity in vitro responded in ways consistent with the lipolysis hypothesis. The preclinical picture was compelling.

Metabolic Pharmaceuticals, an Australian biotech company, licensed and developed the compound through the late 1990s and into the 2000s. This was the era when the obesity epidemic was beginning to register as a genuine pharmaceutical market — not just a public health concern but a commercial opportunity of significant scale. If you could develop a drug that reduced fat without the side-effect baggage of exogenous HGH or the cardiovascular risks associated with stimulant-based weight loss compounds, you had a potential blockbuster. Metabolic Pharmaceuticals moved AOD-9604 into clinical development as an anti-obesity therapeutic with that framing: a clean, targeted, growth-hormone-derived fat metabolism agent.

The Phase I safety work was encouraging. AOD-9604 appeared to be well-tolerated, didn't show the glucose-dysregulating effects of full-length GH, and didn't elevate IGF-1 — which was exactly the point. The receptor-selectivity hypothesis, at least in terms of what the compound didn't do, seemed to be holding. This was enough to move into Phase II trials for obesity.

The Phase IIa and IIb studies ran across the mid-2000s and involved patients with obesity receiving AOD-9604 orally — this was an oral formulation, which was itself a notable choice, since most peptides are degraded in the gut and require injection — at varying doses over twelve-week periods. The results were real but modest. Weight loss in the AOD-9604 groups exceeded placebo by a few kilograms across the trial duration. It was statistically meaningful in some analyses and less convincingly meaningful in others. The effect existed. It was small.

This is where the drug development story becomes instructive rather than triumphant. The modest delta over placebo in a twelve-week window was not going to justify the Phase III trials, the regulatory pathway, and the commercial infrastructure required to launch an obesity drug. The pharmaceutical industry had learned the hard way that obesity drugs needed to show substantial and durable weight loss — not because a few kilograms wasn't real, but because competing against diet and exercise counseling, and eventually against more potent pharmacological alternatives, required a larger effect size. Metabolic Pharmaceuticals could not make the numbers work commercially. The obesity drug program didn't advance to Phase III.

What survived, oddly, was the research interest. The compound didn't disappear — it shifted direction. A separate research thread at Phosphagenics, another Australian company, began exploring whether the fragment had activity in cartilage and joint tissue, based on findings suggesting it might interact with fat-precursor cells in ways relevant to osteoarthritis. This was a completely different therapeutic category, and the science there was more preclinical than clinical, but it kept the compound visible in research literature and extended the patent and intellectual property conversations around it.

Meanwhile, the compound's pharmacological profile — the verified absence of IGF-1 stimulation, the apparent lack of glucose disruption, the animal data showing fat metabolism activity — made it a natural candidate for the compounding pharmacy ecosystem that was growing in parallel with mainstream pharmaceutical development. As GH peptide compounding expanded through the 2010s, AOD-9604 became part of the available repertoire: a compound with a documented research history, a plausible mechanism, and a safety profile that looked clean relative to the full-length GH alternatives.

The regulatory picture shifted in Australia when the Therapeutic Goods Administration moved AOD-9604 from a compounding-permitted compound to a schedule that placed tighter controls on its preparation and prescription. This happened in the early 2020s and reflected the TGA's evolving view of the peptide landscape more broadly. In the United States, AOD-9604 existed in a different regulatory space — not FDA-approved for any indication, accessible through compounding pharmacies under provider prescription, and sitting in the ambiguous territory that many research-grade peptides occupy.

The story of AOD-9604 from the 1990s Monash laboratory to its current presence in compounding menus is a specific kind of pharmaceutical narrative, and it's worth understanding what kind. This is not a story of a compound that was found to be dangerous. It's not a story of fraud or of science being misrepresented. It's a story of a compound whose science was real enough to attract serious investment and move through meaningful clinical trials, but whose effect size in the primary intended indication was insufficient for the commercial and regulatory demands of the pharmaceutical market. That's a different story than most people assume when they hear "this never became an approved drug."

What it means for the compound's current use is nuanced. The clinical trials exist and are accessible. The mechanism is characterized and understood at a reasonable level, with some significant remaining uncertainties. The safety signals that emerged from the clinical program were reassuring. The efficacy signals were real but modest. That combination — honest safety data, honest efficacy data, legitimate mechanistic research — is more solid footing than a great deal of what circulates in the compounding peptide space, where compounds sometimes have minimal human data at all.

And yet the gap between what the trials showed and how AOD-9604 is currently positioned in consumer markets is real and worth naming. The clinical research was conducted on obese patients in a twelve-week pharmaceutical trial. Its applicability to the way the compound is now discussed — as a body recomposition tool, a fat loss accelerant, a targeted lipolytic for people who are already relatively lean and training — has never been tested in that population, in those conditions, with those goals. The science that exists supports a modest and real lipolytic effect in an obesity-treatment context. It does not support the broader enthusiasm that has accumulated around it.

Sometimes drug development reveals that the science works and the commercial framing doesn't. Modest effect sizes that don't justify a blockbuster investment can still represent genuine biology. The fat-metabolism signal in the C-terminal fragment of growth hormone is real. The question Frank Ng started with in the early 1990s — whether you can separate the lipolytic signal from the anabolic and glucose-dysregulating signals — yielded a partial yes. Not a large yes, not a commercially sufficient yes, but a yes that has kept researchers and compounding providers interested for thirty years. That is its own kind of answer, and it deserves to be understood clearly: as what it is, not as what the marketing around it sometimes implies it might be.