the research, exhibit by exhibit

Before the exhibits

AOD-9604's research record breaks into two clean stories that should not be conflated. The first is the mechanism: the C-terminal fragment of growth hormone shuts down acetyl-CoA carboxylase in fat cells (inhibiting new-fat synthesis) and, in chronically treated obese mice, raises beta-3 adrenergic receptor expression toward lean-mouse levels. The second story is the human trial: across roughly 900 obese adults and six placebo-controlled studies, the safety profile was indistinguishable from placebo, and the pivotal 24-week Phase IIb trial did not produce statistically significant weight loss. A third thread — intra-articular rabbit cartilage data from 2015 — is one preclinical study with no human follow-up. The exhibits below walk each thread in turn, with the species, route, and preclinical-vs-human distinction kept explicit throughout, because for AOD-9604 that distinction is the point.

Mechanism: a deliberately narrow slice of growth-hormone biology

AOD-9604's design hypothesis is unusually clean. Intact human growth hormone does several things at once — it elevates IGF-1, modulates glucose tolerance, supports linear growth, and mobilises fat — and Frank Ng's group at Monash set out to isolate the last activity from the others by synthesising just the C-terminal fragment that the lipolytic work in the 1970s had pointed to ^[1][20]. The resulting peptide, with an extra N-terminal tyrosine for receptor and stability considerations and a disulfide bridge between Cys-7 and Cys-14 that closes the cystine loop, was designed to deliver lipolysis without engaging the growth hormone receptor ^[20].

Mechanistically, the published rodent work places AOD-9604 upstream of the β3-adrenergic receptor on adipocytes. Heffernan and colleagues showed that chronic treatment in obese mice restored repressed β3-AR mRNA in white adipose tissue back toward lean-control levels, and that the chronic lipolytic phenotype disappeared in β3-AR knockout mice ^[3]. The acute increase in fat oxidation, interestingly, did not disappear in the knockouts — so AOD-9604 likely engages at least two routes, with β3-AR signalling responsible for the durable weight-loss effect in chronic dosing and a second, slower-to-fade acute pathway operating independently of β3-AR ^[3].

Downstream of β3-AR, the cascade follows the standard adipocyte lipolytic pathway: cAMP rises, protein kinase A activates, hormone-sensitive lipase is phosphorylated, and intracellular triglyceride is hydrolysed to glycerol and free fatty acids ^[21]. Acetyl-CoA carboxylase is concurrently inhibited, suppressing de novo lipogenesis ^[21]. The upstream cell-surface binding partner of AOD-9604 itself — the receptor or receptor complex that triggers β3-AR upregulation — has not been definitively identified in published work, and a 2014 review note flags this as the major unresolved mechanistic question ^[21].

Preclinical efficacy: Zucker rats, ob/ob mice, and a credible separation of mechanism

The foundational preclinical efficacy study (Ng et al., Hormone Research 2000) treated genetically obese Zucker rats with oral AOD-9604 at 500 μg/kg/day for 19 days ^[1]. Body-weight gain in the treated arm was 15.8 ± 0.6 g, compared to 35.6 ± 0.8 g in controls — a reduction of more than 50%. Adipose tissue lipolytic activity rose, and insulin sensitivity was preserved — the latter point being central to the design hypothesis, since intact hGH would have been expected to push glucose tolerance in the opposite direction ^[1].

Heffernan's 2001 ob/ob mouse work, using a 14-day continuous infusion delivered through subcutaneous mini-osmotic pumps, replicated the weight-loss effect and added two important measurements: plasma glycerol — a direct in-vivo lipolysis marker — rose, and fat oxidation increased on indirect calorimetry, while hyperglycaemia and impaired insulin secretion did not appear ^[2]. Taken together, these two studies framed AOD-9604 as a lipolytic-but-not-diabetogenic candidate, distinct from intact hGH in exactly the way the design hypothesis predicted.

The Endocrinology paper that followed in the same year added the mechanism-of-action half: β3-AR knockout mice failed to lose weight on chronic AOD-9604, while wild-type obese mice both lost weight and showed restored β3-AR mRNA expression in adipose tissue ^[3]. That study did the work of converting a phenotype observation into a pathway hypothesis.

The six-trial human programme: a 12-week signal that did not survive 24 weeks

Between 2001 and 2007, Metabolic Pharmaceuticals ran six sequential human trials of AOD-9604 — METAOD001 through METAOD006 — comprising roughly 900 subjects in total, mostly obese adults ^[6][21]. The first four were dose-escalation, single-dose IV PK, and short oral repeat-dose studies. The two pivotal trials were METAOD005 (12-week Phase IIb dose-finding, n=300, oral 1-30 mg daily) and METAOD006 (24-week Phase IIb pivotal, n=502, oral 0.25-1 mg daily) ^[4][5].

METAOD005 produced the result that justified moving forward. In the 12-week dose-finding study, the 1 mg arm lost an average of 2.6 kg compared with 0.8 kg for placebo — a statistically significant short-term separation that, on its own and at that timepoint, looked like the awaited human translation of the rodent preclinical work ^[4]. An interim look at the first 100 subjects in the larger Phase IIb had even shown the 1 mg arm losing roughly 2.8 kg, more than triple placebo ^[18]. The decision to extend the same dose range across 24 weeks at the 502-subject scale was conventional for the era and well-justified by the interim data.

METAOD006 did not replicate the result. At both the 12-week and 24-week timepoints, the 0.25, 0.5, and 1 mg oral arms did not separate from placebo by a statistically significant margin on the primary weight-loss endpoint ^[5]. Metabolic Pharmaceuticals discontinued the obesity development programme in February 2007 ^[5]. The most parsimonious read on the trial — which is the read the published programme summary effectively endorses — is that the early signal at 12 weeks did not persist at 24 weeks and did not generalise to a larger population ^[5][6]. The pooled safety result, by contrast, was unambiguously clean: across all six trials, safety and tolerability were indistinguishable from placebo, no clinically significant IGF-1 elevation appeared, oral glucose tolerance did not shift, no anti-AOD-9604 antibodies were detected, and a small number of serious adverse events in long-term arms (skin cancers, lipoma, breast cancer) were judged unrelated to study drug by investigators ^[6][21].

The osteoarthritis pivot: one rabbit study, no human trial

After the obesity programme closed, Metabolic Pharmaceuticals' corporate successor Calzada Ltd. announced in 2012-2013 that AOD-9604 had shown signals in cartilage-repair preclinical work, and the molecule's development was repositioned away from systemic anti-obesity dosing and toward intra-articular use for osteoarthritis ^[17]. The published cartilage evidence is thinner than the obesity record but worth reading on its own terms.

The principal in-vivo study is Kwon & Park 2015 (Annals of Clinical and Laboratory Science), a collagenase-induced knee osteoarthritis model in mature New Zealand white rabbits ^[9]. Thirty-two animals were randomised across four arms: saline control, hyaluronic acid alone (6 mg), AOD-9604 alone (0.25 mg), and AOD-9604 + hyaluronic acid combination. Injections were ultrasound-guided and delivered weekly for four to seven weeks. On gross morphological and histopathological scoring, the combination arm outperformed every other arm. The lameness-recovery period — the time-to-return-to-normal-gait endpoint — was 11 ± 4 days in the combination arm versus 15 ± 3 days for hyaluronic acid alone, 16 ± 2 days for AOD-9604 alone, and 25 ± 2 days for saline (all p<0.05 versus saline) ^[9].

A 2024 review in Cartilage (Liao, Chen, Chang) cataloguing peptides with in-vivo efficacy for chondrogenic induction and cartilage regeneration includes AOD-9604 in its peptide table, citing the Kwon & Park rabbit study as the supporting evidence ^[10]. No human cartilage trial of AOD-9604 is referenced in the review, and as of mid-2025 no such trial appears in ClinicalTrials.gov, EudraCT, or the Australian New Zealand Clinical Trials Registry ^[10][16]. The cartilage-repair claim for AOD-9604 is, in the strictest reading, a one-rabbit-study claim with supporting in-vitro chondrocyte work ^[11]. That is enough to motivate further investigation; it is not enough to convert into a clinical recommendation.

The 2024 regulatory record: FDA PCAC, FAERS/CAERS, and the GRAS distinction

Three regulatory artefacts now sit on top of the older clinical record and shape the 2026 review picture.

The first is the December 4, 2024 FDA Pharmacy Compounding Advisory Committee meeting on bulk drug substances nominated for the 503A Bulks List. PCAC voted that AOD-9604 — both the free base and the acetate forms — should not be added to the list, citing limited long-term safety data, peptide-impurity profile concerns, and immunogenicity risk ^[14]. The practical effect: 503A pharmacies in the United States can no longer use AOD-9604 as a bulk substance for patient-specific compounded preparations. A pharmacy-compounder public testimony submitted to the same docket ('Saving AOD9604') argued the opposite case based on observational clinical use and the GRAS nutraceutical pathway; the committee did not adopt it ^[22].

The second is the pharmacovigilance dossier prepared for that meeting. Searches of the FDA Adverse Event Reporting System (FAERS) through January 2024 and the CFSAN Adverse Event Reporting System (CAERS) from January 2004 through April 2024 returned no AOD-9604-attributed adverse-event reports ^[13]. That is a striking number, although it is also the expected number for a compound that is not an approved drug, is not legally compoundable through 503A, and has been sold mostly through research-peptide channels that do not feed FAERS.

The third is the GRAS self-affirmation that is sometimes conflated with FDA approval and is not the same thing. AOD-9604 received GRAS self-affirmation for use as an ingredient in food, beverages, and dietary supplements at approximately 1 mg per serving — an industry-led safety conclusion based largely on the rat 6-month and monkey 9-month chronic toxicology (NOAELs of 100 and 50 mg/kg/day respectively) and the human Phase I/II safety pool ^[7][15]. GRAS self-affirmation is a U.S. food-ingredient regulatory status, not a drug approval; it applies to low-mg oral nutraceutical use only, and it does not authorise injectable, compounded, or higher-dose clinical use ^[15]. Conflating the two is one of the more common errors in the popular literature about AOD-9604.

Pharmacokinetics and detection

AOD-9604's plasma half-life is short: approximately 3-4 minutes after intravenous administration in pigs, with the intact molecule undetectable roughly 56 minutes after an IV bolus ^[8]. The principal metabolic pathway is rapid sequential proteolysis from the N-terminus, generating -1aa, -2aa, and -3aa fragments as the dominant in-vivo degradation products ^[8]. Oral bioavailability in rats was reported at approximately 40% — the rationale for the oral capsule format used throughout the METAOD003-006 human programme ^[8]. Tissue distribution studies in rats showed preferential uptake in pancreas, pineal gland, thyroid, liver, and kidney cortex ^[8].

A validated LC-MS/MS detection assay for AOD-9604 in human urine was published in Drug Testing and Analysis (Cox et al., 2015), developed as part of the analytical infrastructure underpinning WADA's prohibition of growth-hormone-related fragments under Section S2 of the Prohibited List ^[12][16]. AOD-9604, as a synthetic fragment derived from hGH residues 177-191, falls squarely inside the S2 category and is banned at all times — both in and out of competition — for athletes subject to the WADA Code ^[16].