AOD-9604: the weight-loss peptide that almost got approved

AOD-9604 is the peptide that comes the closest to a clean pitch. The pitch is the fat-burning fragment of growth hormone, separated from the cell-growing, blood-sugar-raising, joint-thickening bits the full hormone drags along. If that fragment did what the marketing says it does, this would be one of the better-designed peptide drugs on the menu.

The thing nobody walking the cosmetic-side peptide forums tells you is the next sentence. Somebody actually built a drug company around exactly that idea, ran the human trials, and the human trials disappointed. Then the molecule pivoted to joints. Now, in the era of GLP-1 weight-loss drugs that take muscle alongside fat, the same molecule is being rediscovered for muscle preservation.

That arc is the entire AOD-9604 story. Let’s walk it.

What it actually is

AOD-9604 — the name is short for Anti-Obesity Drug 9604 — is the C-terminal piece of human growth hormone, slightly modified for stability. Most of full-length hGH’s effects come from binding the hGH receptor, which kicks off cell proliferation, raises blood sugar, and over time thickens connective tissue. The fragment was designed to drop most of those activities and keep just one: the lipolytic signal that tells fat cells to release stored triglyceride.

Heffernan and colleagues at Monash University, in a 2001 paper in the International Journal of Obesity, nailed the mechanism case. Obese mice given chronic AOD-9604 lost weight relative to controls. The mechanism was clean: fat oxidation up, plasma glycerol up (the marker of lipolysis), in-vivo energy expenditure shifted. The control experiments mattered as much as the headline. AOD-9604 did not bind the hGH receptor in a cell-based assay. It did not induce cell proliferation. Unlike full hGH given to the same mice, AOD-9604 did not cause hyperglycaemia or suppress insulin secretion.

That is the kind of preclinical profile that gets venture capital out of the chair. A molecule that does the desirable bit of growth hormone without the metabolic and proliferative downsides should, at minimum, be worth taking into humans.

The Australian drug story

Metabolic Pharmaceuticals — an Australian biotech that came out of the same Monash research environment — picked AOD-9604 up and ran it through clinical development. Wilding’s 2004 brief in Current Opinion in Investigational Drugs captures the snapshot: Phase IIa trials were underway by February 2002. Metabolic Pharmaceuticals was the sponsor. Obesity was the target indication.

The part the supplement scene almost never narrates is what happened after that 2004 development brief. Metabolic Pharmaceuticals ran later-stage obesity trials in humans, and the publicly available record — published peer-reviewed accounts plus the company’s own contemporaneous communications — converged on the same answer: the placebo-corrected weight loss in the active arm was real but small, and not enough to justify carrying the molecule into a registration trial. Metabolic shelved the obesity programme. The author has to flag that this account of the Phase II failure sits in conference proceedings and company communications more than in a single clean peer-reviewed Phase IIb publication, but the downstream evidence is consistent: no Phase III obesity trial of AOD-9604 was ever registered, no obesity approval was ever pursued, and Metabolic divested the molecule.

That ends the AOD-9604 became an approved fat-loss drug arc. The molecule was tested in humans, for the thing it was designed for, and the human result did not match the mouse result tightly enough to support continued development.

The pivot to joints

When the obesity programme stalled, the molecule did not disappear. The biology Heffernan and the Monash group had built — Metabolic licensed the compound for a different indication entirely, and a small body of musculoskeletal work began.

Kwon and Park, writing in the 2015 Annals of Clinical and Laboratory Science, are the cleanest published example of that pivot. They ran a collagenase-induced rabbit knee osteoarthritis model, gave thirty-two rabbits weekly intra-articular injections of either saline, hyaluronic acid, AOD-9604, or AOD-9604 combined with hyaluronic acid, and tracked cartilage degeneration. Saline rabbits looked the worst, by both gross morphology and histopathology. The hyaluronic acid, AOD-9604, and combination arms were all significantly better.

That is a rabbit study. The collagenase model is well-characterised in the OA literature. The intra-articular delivery route is exactly the one a compounding pharmacy or a sports-medicine practice would use. A real animal result.

What it is not is a published human osteoarthritis trial of AOD-9604. The indexed literature does not contain one. The pivot story is therefore: positive animal model in 2015, then quiet on the human-trial front since. The reason there is not a Phase 2 human OA trial of AOD-9604 in the peer-reviewed record is presumably the same reason there is not a Phase 3 obesity trial: somebody would have had to fund it, and the regulatory and market case for a peptide with a stalled obesity programme is hard to make.

The third life: the GLP-1 muscle-loss problem

This is where the molecule has found its third audience.

The arrival of semaglutide and tirzepatide as obesity drugs changed the math. Patients on GLP-1 receptor agonists or dual GLP-1/GIP agonists lose a lot of weight. They also lose, depending on the trial, between roughly twenty and forty percent of that weight as lean mass — muscle and bone, not fat. That number is uncomfortable. A fifty-year-old who loses fifteen kilos on tirzepatide and gives back five kilos of muscle along the way has a body composition outcome very different from a fifty-year-old who loses fifteen kilos through resistance training and protein adequacy.

Arora and colleagues, in a January 2026 review in the Journal of Clinical Medicine, document the resulting research landscape — pharmacological strategies for preserving lean body mass during weight loss. The review categorises the candidate classes being developed for this problem, including peptides that target fat oxidation without engaging muscle-suppressing pathways. AOD-9604 sits naturally in that conversation because Heffernan’s original mouse data showed exactly the mechanism the GLP-1 era needs: fat loss without the cell-proliferative downsides that full hGH would carry. Whether the human trial data ever supports that promise is a separate question. The mechanistic case for renewed interest is defensible.

The supplement scene is, predictably, ahead of the formal trial machinery. AOD-9604 is being sold off research-chemical sites as a stack-on adjunct to GLP-1 use, on a mix of the original Monash mouse data, the rabbit OA study, and the GLP-1-muscle-loss framing. None of those three things is a published human trial of AOD-9604 for muscle preservation during weight loss. The use is well ahead of the evidence.

Where the FDA settled it — December 4, 2024

AOD-9604 is not on the July 23–24, 2026 FDA Pharmacy Compounding Advisory Committee docket — and the reason is sharper than the standard not on the docket line. The FDA already reviewed AOD-9604 in 2024. After it was removed from the Category 2 list in September 2024 and referred to PCAC, the committee evaluated it at the December 4, 2024 meeting and voted against inclusion on the 503A bulk drug substances list. The agency’s stated reasoning followed the standard pattern for the four peptides reviewed in that round: insufficient human safety data, mechanistic concerns, and lack of reproducible efficacy data outside small or open-label trials.

That has a different meaning here than it does for some other peptides on this menu. AOD-9604 was once a serious drug candidate. It went into clinical development under a regulated sponsor — Metabolic Pharmaceuticals — and the human trials that stalled were proper trials. The peptide has, structurally, more drug-development history behind it than BPC-157 or GHK-Cu. What it does not have is a current path to US approval. Metabolic divested the compound. No sponsor is currently advancing a new drug application. PCAC’s December 2024 vote against 503A inclusion closed the compounding-pathway question alongside.

That leaves anyone using AOD-9604 today buying it through research-chemical channels, with the regulatory review already concluded against compounding access and no near-term prospect of legitimate US prescribing access. Mendias and Awan’s 2026 Sports Medicine review names AOD-9604 specifically among the twelve approved and unapproved peptides being marketed direct to patients, frames the broader unapproved-peptide market as operating largely outside regulatory oversight, and flags that animal-model results have not translated to rigorous human safety data — exactly the position AOD-9604 sits in today.

For WADA, the picture is more interesting. AOD-9604 is structurally derived from hGH and is on the prohibited list as a substance with a similar chemical structure or biological effect to hGH; competitive athletes should treat it as banned and confirm against the current code on the day of testing rather than relying on a peptide-by-peptide intuition.

What would actually settle the AOD-9604 question

If you wanted AOD-9604 to be a credible muscle-preserving adjunct to GLP-1 weight loss — the use case that is actually driving demand today — here is the evidence that does not yet exist.

A randomised controlled trial of AOD-9604 added to a GLP-1 receptor agonist in adults with obesity, with body composition endpoints — DEXA or MRI — measured rather than estimated. None has been published.

A pharmacokinetic study of subcutaneously injected AOD-9604 in adults on GLP-1 therapy. Not in the literature.

A revisit of the failed Phase II obesity programme with the benefit of twenty-plus years of obesity-trial methodology improvements — better endpoint definition, better placebo design, better population selection. Nobody has run it.

A published Phase II osteoarthritis trial in humans following the Kwon rabbit work. The indexed literature does not contain one.

These are not exotic asks. The reason they are not in the literature is that AOD-9604 lost its commercial sponsor twenty years ago, and the resurgent interest from the GLP-1 era has not yet translated into a regulatory or trial pipeline.

Where AOD-9604 lands

The mechanism case is genuinely clean. A fragment of growth hormone that drives fat oxidation without binding the hGH receptor and without raising blood sugar is exactly the kind of designed molecule the field needs more of. The Monash preclinical work is rigorous.

The clinical track record is honest about its disappointment. The early-phase obesity programme did not progress to a registration trial. The rabbit OA work pivoted the molecule but did not advance into humans. Twenty-plus years on, there is no published human trial of AOD-9604 with a strongly positive primary endpoint.

The current use case — adjunct to GLP-1 weight loss for muscle preservation — is mechanistically attractive, has zero direct human trial evidence today, and is being sold through unregulated channels into a population (people on tirzepatide or semaglutide losing significant lean mass) who genuinely have a real problem the mechanism could plausibly address. The plausibility is real. The data is absent.

AOD-9604 is the peptide on this menu with the clearest mechanistic case and the bleakest commercial-trial history. Whether it becomes a real drug again depends on somebody funding the GLP-1-era trial it actually deserves. Today, what you can buy off a research-chemical site is the same molecule whose Phase 2 obesity programme stalled twenty years ago, sold for a use it has never been tested in.

That is what we are building. Wolverine Health is a physician-supervised peptide service — real prescriptions, US-licensed compounding pharmacies, every batch third-party tested. It isn’t live yet, because the regulation isn’t either, and we’re not going to sell you ahead of the science. If you want to know the moment somebody finally funds the GLP-1-era AOD-9604 trial — the one that turns the mechanism case into either a real drug or an honest no — join the waitlist.