AOD-9604: the GH fragment meant to burn fat without touching glucose

What AOD-9604 is

Human growth hormone (hGH) has well-known lipolytic effects, but long-term use is associated with insulin resistance and hyperglycaemia⁷. At Monash University, Frank Ng's group proposed that the action on fat resided in a specific region of the molecule⁴. The result was AOD-9604: the C-terminal fragment of hGH (amino acids 177-191) with an additional tyrosine at the N-terminus¹, a 16-amino-acid peptide with a disulfide bridge⁷. The Australian company Metabolic Pharmaceuticals developed it as an oral anti-obesity candidate⁷. The compound's data sheet is at AOD-9604. Unlike the secretagogues, it does not stimulate the pituitary; for that other group, see CJC-1295, ipamorelin and MK-677.

What the laboratory showed

In obese ob/ob mice treated for 14 days with osmotic minipumps, both hGH and AOD-9604 significantly reduced weight gain, with increased fat oxidation and plasma glycerol, an index of lipolysis. Unlike hGH, AOD-9604 did not induce hyperglycaemia or reduce insulin secretion, did not compete for the GH receptor and did not induce cell proliferation². A second study linked the effect to increased expression of the β3-adrenergic receptor, although it concluded that the lipolytic action is not mediated directly through that receptor³.

The oral route was also explored in rodents. In obese Zucker rats, oral treatment for 19 days reduced weight gain by more than 50% (15.8 vs 35.6 g), with no adverse effect on insulin sensitivity measured by euglycaemic clamp⁴. In isolated adipose tissue from obese rodents and humans, a related synthetic fragment from the same group, named AOD-9401 in that paper, increased lipolysis and reduced lipogenesis in vitro; given orally it reduced weight gain in ob/ob mice⁵.

Human safety: the strongest point

The best clinical evidence for AOD-9604 is on safety. Stier and colleagues summarized six randomized, double-blind, placebo-controlled trials, with particular attention to the unwanted effects of hGH: raised IGF-1, insulin resistance and glucose intolerance. AOD-9604 did not change serum IGF-1, the oral glucose tolerance test showed no negative effect on carbohydrate metabolism, and no anti-AOD-9604 antibodies were detected in the patients tested. The authors described a safety and tolerability profile "indistinguishable from placebo"⁶.

According to the FDA's summary, the two largest trials enrolled 300 and 502 participants with obesity, treated orally for 12 and 24 weeks; in the 300-person trial, five participants had serious adverse events, including skin cancers (basal cell carcinoma, squamous cell carcinoma and melanoma) and breast cancer, which the principal investigator did not consider related to the drug; the FDA stated concern about those cancer reports⁷. The FDA also noted that the authors were employees or consultants of the company that funded the trials⁷.

The trial that ended the obesity program

A first phase 2b trial, without a formal diet and exercise program, showed greater mean weight loss than placebo in all five dose groups at 12 weeks; the best dose (1 mg) missed significance on the primary analysis (p = 0.1), though it reached it in the female subgroup⁸. The OPTIONS trial set out to confirm this: 536 adults enrolled, BMI 30 to 45 kg/m², oral doses of 0.25, 0.5 or 1 mg daily or placebo for 24 weeks, with dietitian-supervised diet and exercise, powered to detect a 1.8 kg difference⁸.

On 21 February 2007 the company announced that weight loss, after allowing for the diet and exercise program, was under 1 kg in every dose group at 12 and 24 weeks, and that development for obesity was terminated. In the same announcement it described safety and tolerability as excellent, with no difference from placebo⁸. According to the FDA, these results do not appear to have been published in medical journals⁷.

Cartilage and bone: a preclinical line

After closing the obesity program, the company announced it would continue studying AOD-9604 in osteoporosis on the basis of animal results⁸. In cartilage, the most cited published data come from a collagenase-induced knee osteoarthritis model in 32 rabbits: weekly intra-articular AOD-9604 injections were associated with better morphological and histopathological scores than saline, and the combination with hyaluronic acid was more effective than either treatment alone, with a shorter lameness period¹¹. There are no published clinical trials in osteoarthritis, and the FDA did not evaluate that use because the nomination lacked information⁷.

What the FDA concluded

In its briefing document for the December 2024 Pharmacy Compounding Advisory Committee, the FDA concluded that evidence of efficacy in obesity by any route is lacking and that it found no human data by the subcutaneous or transdermal routes and no human pharmacokinetic studies⁷. In the nonclinical review it noted dose-dependent changes in osteocalcin in rats treated for 26 weeks, minimal periportal vacuolation in monkeys treated for 9 months, and equivocal genotoxicity results⁷. A search of the FAERS adverse event database retrieved no reports⁷. The committee voted 0–12 on adding it to the 503A list⁹. The FDA's category 2 page lists it among nominations withdrawn by their nominators¹⁰. The context of the 2026 votes on other peptides is in What the FDA actually said.

Where the evidence ends

AOD-9604 fulfilled part of its design hypothesis: in animals it separated GH's lipolytic effect from its effects on the receptor, glucose and cell proliferation², and in humans it did not raise IGF-1 or alter glucose tolerance⁶. What it did not show was clinically useful weight loss in a large trial with diet and exercise⁸. The safety data come from the oral and intravenous routes, were published in summary form by company-affiliated authors, and do not cover the subcutaneous route⁶⁷. For investigators, the compound remains of interest as a laboratory tool for the biology of GH fragments and in cartilage models³¹¹.