AOD-9604 Research

AOD-9604 was tested in around 900 randomized adults across six trials ^[3]. Its pivotal phase 2b obesity trial, METAOD006, missed its primary endpoint, and the sponsor terminated the obesity program in February 2007 ^[1]. FDA's 2024 conclusion is that AOD-9604 failed to show benefit for weight reduction against placebo in most of the studies it identified ^[1]. The single positive weight-loss claim in the program rests on a conference abstract with no full publication ^[7]. This article describes research findings, not outcomes anyone should expect.

Research-grade AOD-9604 is supplied for laboratory research use only.

The trial that decided it (human)

METAOD006, called the OPTIONS study, was randomized, double-blind, placebo-controlled and parallel-group, and it ran across 16 Australian hospitals and medical centers ^[1][3]. A four-week single-blind placebo run-in preceded 24 weeks of oral dosing ^[1]. Participants were adults aged 18 to 65 with a body mass index between 30 and 45 ^[1]. Three dose arms were compared against placebo, on top of a dietician-supervised diet and exercise program ^[1]. FDA and the sponsor's filing record 536 enrolled and 502 randomized, while the published safety paper gives 534 enrolled ^[1][3].

The first primary endpoint was significant weight loss at 12 weeks in any one dose arm, and the second was safety ^[1]. The trial was powered for an 80 percent chance of significance if placebo-adjusted weight loss reached 1.8 kg ^[1]. There was no significant difference versus placebo at 12 weeks or at 24 weeks, at any dose ^[1]. The sponsor's own filing states that the difference was "too low to reach statistical significance" ^[1]. On 21 February 2007 the company announced that development for obesity was terminated, and it never restarted ^[1][2].

Nineteen years later no peer-reviewed publication of the trial's methods or results exists, and no ClinicalTrials.gov record for AOD-9604 exists at all ^[1][15]. Three sources agree on the failure, one of them the sponsor itself, while the underlying trial data sit outside the scientific record ^[1][2][5].

The rest of the human program (human)

Six randomized, double-blind, placebo-controlled trials were run in approximately 900 randomized adults ^[3]. A drug-development review covering three of the earlier ones reports no significant weight loss in any ^[6]. Non-esterified fatty acids rose after dosing in both an intravenous and an oral study, while weight did not follow ^[6]. In a one-week study in 36 men the lowest dose arm lost 1.0 kg against 0.6 kg on placebo, and higher doses appeared less effective ^[6].

The only positive human weight-loss result is a 2005 conference abstract that was never published in full ^[7]. It describes 300 adults over 12 weeks, with weight loss above placebo at all doses but a non-linear dose response ^[7]. The abstract claimed the weight result was significant in females, and the weekly rate of waist-circumference reduction significant at all doses ^[7]. Both are claims made inside a conference abstract with no correction for multiplicity, and the weight claim is a subgroup finding ^[7]. The largest effect was at the lowest dose ^[7], and FDA assessed the claim as resting on small differences of unclear therapeutic meaning ^[1]. A dose response in which the lowest dose performs best is a recognized signature of noise. That reading is this page's, not FDA's.

Safety, and what the safety claim rests on (human and animal)

The one peer-reviewed publication reporting pooled primary data covers safety alone, and gives no efficacy result for either phase 2b trial ^[3]. Across the six trials it found no effect on serum IGF-1 and no adverse effect on oral glucose tolerance ^[3]. No anti-AOD9604 antibodies were detected up to 24 weeks, and the authors called tolerability "indistinguishable from placebo" ^[3]. That paper is widely cited as validation, but it never reports whether the compound worked ^[3].

In the 12-week trial in 300 adults, five cancers occurred, all in active arms and none on placebo ^[3]. The authors dismissed these on two grounds, that none arose in the top dose group and that Australia has high background skin-cancer rates ^[3]. Neither was a prespecified analysis ^[3]. FDA's separate reading is that the reports held too little information to assess relatedness to AOD-9604 ^[1][8].

FDA also re-read the sponsor's animal toxicology and reached different conclusions from it ^[8]. A 26-week oral rat study showed dose-dependent changes in serum osteocalcin, and a nine-month monkey study showed slight periportal hepatocyte vacuolation ^[8]. That liver change also appeared in one control animal ^[8]. Genotoxicity was judged inconclusive rather than clean, with equivocal signals across three assay types ^[8]. These are animal signals of uncertain human relevance, and FDA separately concluded that the molecular target and mechanism of action remain unknown ^[8].

Identity, and the peptides it gets merged with

AOD-9604 is Tyr-hGH(177-191), PubChem CID 71300630, CAS 221231-10-3, molecular weight 1815.1, while unmodified hGH fragment 176-191 is a separate record, CID 16131230, CAS 66004-57-7, molecular weight 1799.1 ^[9]. Mature growth hormone residue 176 is phenylalanine, so AOD-9604 substitutes an N-terminal tyrosine and the two formulas differ by one oxygen atom ^[9]. A third peptide, AOD-9401, is hGH 177-191 without that added tyrosine ^[10].

What the preclinical work showed (animal and in vitro)

In genetically obese ob/ob mice, 14 days of AOD-9604 reduced body-weight gain and raised fat oxidation and plasma glycerol ^[11]. Unlike full-length growth hormone it did not induce hyperglycemia, and in cells it did not compete for the growth hormone receptor ^[11]. In obese Zucker rats, 19 days of oral AOD-9604 reduced body-weight gain by more than half, which is reduced gain rather than weight loss ^[11].

A companion study treated obese mice and beta-3 adrenergic receptor knockout mice ^[11]. Chronic dosing failed to produce in the knockouts the changes seen in wild-type controls ^[11]. In an acute experiment, however, AOD-9604 still raised energy expenditure and fat oxidation in those same knockout mice ^[11]. The authors concluded that the lipolytic action is not directly mediated through that receptor ^[11]. FDA reads the same data as showing dependence at least in part on intact beta-3 signaling ^[8].

Two of the most-cited lipolysis papers studied AOD-9401 rather than AOD-9604, and FDA treated AOD-9401 as out of scope for its evaluation ^[8][10]. AOD-9401 increased lipolytic activity in isolated adipose tissue from obese rodents and humans, which is tissue in a dish rather than a person ^[10]. That result belongs to a different peptide from the one sold here ^[10].

The published hGH fragment 176-191 article states that this lipolysis literature belongs to AOD-9604, and that it does not transfer to the unmodified fragment. That holds for the papers that studied AOD-9604 itself, and it is favorable to neither product. The unmodified fragment has no human efficacy evidence, and AOD-9604's human efficacy evidence is negative ^[1][5].

After 2007 the compound was redirected toward cartilage and joint repair, supported mainly by one rabbit osteoarthritis study ^[2][12]. There is no human trial of AOD-9604 in any joint indication, and the sponsor called this work preclinical only ^[2][8]. No human pharmacokinetic data exist for any route, and FDA states that oral bioavailability is unclear even in animals ^[8].

Regulatory status, and the GRAS confusion

AOD-9604 has no drug approval anywhere, and field-qualified openFDA queries return zero records on both the label and Drugs@FDA endpoints, while control queries return records ^[4]. FDA states that neither the free base nor the acetate is a component of an approved drug, and the EMA assessment-report dataset holds no entry ^[4]. The sponsor itself stated in 2013 that no health authority anywhere had approved it as a pharmaceutical product ^[4].

A separate claim circulates about GRAS status, and it is not a drug approval. In June 2012 the sponsor said AOD-9604 had GRAS recognition as a food ingredient, determined by a privately convened expert panel ^[14]. The FDA GRAS Notice Inventory returns zero records, so no notice was ever submitted and FDA never reviewed it ^[14]. A self-affirmed food-ingredient conclusion says nothing about efficacy. Conflating it with a drug approval is the most common error made about this compound.

In December 2024 FDA concluded that the evidence does not support effectiveness for obesity by any route ^[15]. Its advisory committee voted 0 to 12 against compounding inclusion ^[15]. FDA also recorded that no human data exist for the subcutaneous or transdermal routes ^[15].

Detection, funding and what remains untested

AOD-9604 is named on the 2026 WADA Prohibited List within Section S2, where hGH fragment 176-191 is named separately ^[16]. A validated urine method exists, and one metabolite identified in vitro is significantly more stable than the parent compound and the other metabolites ^[17]. The method's authors say that metabolite may potentially extend the detection window ^[17]. Claims of undetectability are therefore unsupported.

Almost the entire positive literature on AOD-9604 comes from one source. Both safety papers were funded by Metabolic Pharmaceuticals and co-authored by its chief executive, who held roughly 7 percent of the parent company ^[3][13]. They exist because the food-ingredient panel's conclusion was made conditional on publication of the existing safety data in peer-reviewed journals ^[2][13]. One co-author was the former medical director who had run the trials he was reporting ^[3]. The rodent work came from the university laboratory that originated the compound, and independent replication is essentially absent ^[11].

AOD-9604 is not approved for human therapeutic use anywhere, and it is not for human consumption.

Frequently Asked Questions

Did AOD-9604 produce weight loss in human trials?

No, with one unpublished abstract as the only exception. METAOD006 randomized 502 adults and found no significant difference in weight loss versus placebo at 12 or 24 weeks, at any dose ^[1]. A drug-development review covering three of the earlier trials reports no significant weight loss in any of them ^[6]. The exception is a 2005 conference abstract that was never published in full ^[7]. The sponsor terminated the obesity program in February 2007 ^[1][2].

Is AOD-9604 approved for anything, and what does its GRAS status mean?

It has no drug approval anywhere, confirmed by field-qualified openFDA queries returning zero records while control queries return records ^[4]. The GRAS claim is a self-affirmed food-ingredient determination made by a private expert panel ^[14]. No GRAS notice was ever submitted to FDA, so FDA never reviewed it ^[14]. A food-ingredient conclusion is not a drug approval and says nothing about efficacy.

Is AOD-9604 the same molecule as HGH Fragment 176-191?

No. AOD-9604 is Tyr-hGH(177-191), and the two formulas differ by one oxygen atom because tyrosine replaces the native phenylalanine ^[9]. They are separate PubChem records with separate CAS numbers, and WADA names them separately ^[9][16]. A third peptide, AOD-9401, appears in two of the most-cited lipolysis papers and is neither of them ^[10].

What does the safety literature actually establish?

The pooled safety paper reports no IGF-1 effect, no adverse glucose-tolerance effect and no antibody formation ^[3]. That same paper records five cancers in the 12-week trial, all in active arms and none on placebo, dismissed post hoc by its authors ^[3]. FDA judged those reports too limited to assess relatedness ^[8]. FDA's own reading of the animal toxicology found bone-turnover, liver and inconclusive genotoxicity signals ^[8].

References

1. 1

   US Food and Drug Administration. Briefing Document, Pharmacy Compounding Advisory Committee meeting, 4 December 2024: Evaluation of AOD-9604-related Bulk Drug Substances for Inclusion on the 503A Bulk Drug Substances List (memo dated 5 November 2024; signatories Mathew B, Amaechi C, Asante K, Albuquerque E, Benedict A, Kneeream E, Lopez L), pp. 24-27. https://www.fda.gov/media/183584/download. Underlying sponsor disclosure of 21 February 2007, lodged by Metabolic Pharmaceuticals Limited (Australia; ASX-listed at the time, later restructured under Calzada Ltd), SEC-filed at https://www.sec.gov/Archives/edgar/vprr/0702/07021963.pdf. No PMID or DOI; both documents retrieved and read in full.
2. 2

   Calzada Limited / Metabolic Pharmaceuticals Pty Ltd (David Kenley, CEO). "AOD9604 - Important Clarifications", ASX company announcement, 26 April 2013. No PMID or DOI. A company disclosure whose evidential value lies in its admissions against interest.
3. 3

   Stier H, Vos E, Kenley D. "Safety and Tolerability of the Hexadecapeptide AOD9604 in Humans". Journal of Endocrinology and Metabolism. 2013;3(1-2):7-15. DOI: 10.4021/jem157w. Not indexed in PubMed; no PMID, so the DOI was resolved on Crossref on 28 August 2026 and the title and author line confirmed there. The only peer-reviewed publication reporting primary pooled data from the AOD-9604 human program, and it reports safety only, as its own title says.
4. 4

   openFDA drug/label and drug/drugsfda endpoints, field-qualified on openfda.substance_name, openfda.generic_name and openfda.brand_name, queried 27 August 2026, with control queries returning records for tesamorelin via generic_name and for ibuprofen via substance_name, confirming the fields are populated; EMA European Public Assessment Report dataset, same date; FDA Briefing Document 4 December 2024, p. 7; Calzada Limited ASX announcement, 26 April 2013.
5. 5

   Valentino MA, Lin JE, Waldman SA. Clinical Pharmacology and Therapeutics. 2010;87(6):652-62. PMID 20445536. DOI: 10.1038/clpt.2010.57. A narrative review of antiobesity drug targets documenting the 2007 termination. PubMed indexes it as a journal article carrying NIH extramural and non-US government research support, not under the Review publication type, and that funding record is what makes it the independent member of the three agreeing sources. Verified in the PMC full text (PMC3136748), because the abstract does not mention AOD-9604. This entry no longer asserts what Valentino's own source for the termination was, since none of the verification material establishes it.
6. 6

   Wilding J. "AOD-9604 Metabolic". Current Opinion in Investigational Drugs. 2004;5(4):436-40. PMID 15134286. No DOI. The title is reproduced as PubMed indexes it; this journal's drug-profile titles pair the compound with the developing company, which here is Metabolic Pharmaceuticals. PubMed publication type is Review, not primary human data; trial specifics are taken from FDA's rendering of the full text at reference 1, p. 24. FDA notes that Wilding's underlying references were company press releases that are no longer available. Wilding gives 16 participants for the oral single-dose study where reference 3 gives 17 for the same trial.
7. 7

   Herd C, Wittert G, Caterson L, Proietto J, Strauss B, Prias J, Stocks A, Vos E, Belyea C. Obesity Research. 2005;13 (annual scientific meeting abstract supplement). No PMID and no DOI; returns zero hits on PubMed and Crossref. Cited here only because FDA cited and assessed it at reference 1, p. 25. Co-author Evert Vos was the sponsor's medical director.
8. 8

   US Food and Drug Administration, Nonclinical Assessment section (Albuquerque E, Benedict A), Briefing Document, Pharmacy Compounding Advisory Committee, 4 December 2024. https://www.fda.gov/media/183584/download. Same document as reference 1, different section. Where the body pairs it with reference 1, the claim is FDA's own reading of reported adverse events, which concerns the human reports rather than the animal toxicology, so it is cited to the proceeding rather than to this section alone.
9. 9

   PubChem CID 71300630 (AOD-9604; CAS 221231-10-3; UNII 7UP768IP4M; C78H123N23O23S2; MW 1815.1) and PubChem CID 16131230 (hGH fragment 176-191; CAS 66004-57-7; C78H123N23O22S2; MW 1799.1). NIH/NLM compound records, queried 27 August 2026, with each CAS round-tripped to its CID. Structure independently corroborated in prose by Cox HD 2015 (PMID 25208511.). The somatostatin synonyms on CID 71300630 are a database annotation error and should never be reproduced.
10. 10

    Heffernan MA, et al. American Journal of Physiology, Endocrinology and Metabolism. 2000;279(3):E501-7. PMID 10950816. DOI: 10.1152/ajpendo.2000.279.3.E501. Ng FM, et al. Journal of Molecular Endocrinology. 2000;25(3):287-98. PMID 11116208. DOI: 10.1677/jme.0.0250287. Compound studied in both: AOD-9401, not AOD-9604. Both contain in vivo rodent work alongside the isolated-tissue work.
11. 11

    Heffernan MA, et al. International Journal of Obesity and Related Metabolic Disorders. 2001;25(10):1442-9. PMID 11673763. DOI: 10.1038/sj.ijo.0801740. Heffernan M, et al. Endocrinology. 2001;142(12):5182-9. PMID 11713213. DOI: 10.1210/endo.142.12.8522. Ng FM, et al. Hormone Research. 2000;53(6):274-8. PMID 11146367. DOI: 10.1159/000053183. All rodent. The beta-3 conclusion reported here is the authors' own.
12. 12

    Kwon DR, Park GY. Annals of Clinical and Laboratory Science. 2015;45(4):426-32. PMID 26275694. No DOI assigned. Rabbit model of collagenase-induced knee osteoarthritis, n=32, eight per group, intra-articular route; no blinding described in the abstract.
13. 13

    Moré MI, Kenley D. "Safety and Metabolism of AOD9604, a Novel Nutraceutical Ingredient for Improved Metabolic Health". Journal of Endocrinology and Metabolism. 2014;4(3):64-77. DOI: 10.14740/jem213w. Not indexed in PubMed; no PMID, so the DOI was resolved on Crossref on 28 August 2026 and the title and author line confirmed there. Nonclinical original research reporting toxicology and pharmacokinetics in rats, pigs and monkeys. Its title states the nutraceutical framing outright, which is consistent with the food-ingredient motive described at reference 14. The oral bioavailability comparison in this paper is not reproduced here, because FDA's dose-adjusted recalculation implies a figure above 100 percent and FDA concluded the true magnitude is unclear (reference 8).
14. 14

    US FDA GRAS Notice Inventory, searched 27 August 2026: 0 records for "AOD9604", 0 for "AOD-9604" and 0 for "oligopeptide", against a control search for lutein that returns 10 records and proves the search works. Sponsor characterization of the panel determination from the Calzada Limited ASX announcement, 26 April 2013, repeated at references 3 and 13.
15. 15

    US Food and Drug Administration, Pharmacy Compounding Advisory Committee, 4 December 2024: Briefing Document (https://www.fda.gov/media/183584/download), voting questions (https://www.fda.gov/media/183945/download) and verbatim meeting transcript (https://www.fda.gov/media/185641/download), vote at Topic 2, Question 2A, where an initial 1-yes tally was corrected on the record to 0 yes, 12 no, 0 abstentions. The briefing document also states that a search of ClinicalTrials.gov listed no studies on AOD-9604; the ClinicalTrials.gov v2 API was re-queried independently on 28 August 2026, returning zero studies for both "AOD9604" and "AOD-9604" while a tesamorelin control returned 24. An advisory committee vote is a recommendation, not a final rule.
16. 16

    World Anti-Doping Agency. The 2026 Prohibited List, Section S2.2.3, Peptide Hormones, Growth Factors, Related Substances, and Mimetics. In force 1 January 2026. AOD-9604 and hGH 176-191 are indexed separately. No PMID or DOI; the list PDF was retrieved and searched directly.
17. 17

    Cox HD, Smeal SJ, Hughes CM, Cox JE, Eichner D. Detection and in vitro metabolism of AOD9604. Drug Testing and Analysis. 2015;7(1):31-8. PMID 25208511. DOI: 10.1002/dta.1715 (validated urine method, limit of detection 50 pg/mL; six candidate metabolites). Orlovius AK, Thomas A, Schänzer W, et al. Drug Testing and Analysis. 2013;5(11-12):850-2. PMID 24124033. DOI: 10.1002/dta.1557 (AOD-9604 does not influence the WADA hGH isoform immunoassay). Indexed as a Letter, so it is a short communication rather than a full method paper.