Dihexa

ASSUMED (theoretical, foundational papers retracted): Dihexa was proposed to bind hepatocyte growth factor (HGF) with high affinity (Kd around 65 pM), stabilize its active dimeric form, and thereby amplify signaling through the c-Met receptor tyrosine kinase. This was theorized to promote synaptogenesis and dendritic spine formation in hippocampal neurons. The proposed advantage over HGF itself was that Dihexa is small enough to cross the blood-brain barrier where HGF cannot. However, both papers establishing this mechanism (Kawas 2012 and Benoist 2014) were formally retracted for data fabrication in April 2025. The mechanism is therefore currently unsupported in peer-reviewed literature. An alternative IRAP (insulin-regulated aminopeptidase) mechanism for angiotensin IV analogs exists in the literature and may partially explain some effects if any are real.

The letter counts how far testing got. Dihexa has rat data only (D); the fosgonimeton trials that failed in Alzheimer's tested a different molecule and do not count toward dihexa's letter.

• Self-experimenters report sharper recall and focus. The animal studies behind that claim were retracted for data fabrication in 2025.
• The two mechanistic papers that explained how it works were formally retracted by WSU in April 2025 after fabricated figures were confirmed. The HGF/c-Met story it was sold on is now unsupported in peer-reviewed literature.
• It activates a well-documented cancer-promoting signaling pathway and has never been tested in a single human safety trial. Every dose is a completely uncharted gamble.

Adverse events

No controlled human adverse event data exists. Anecdotal community reports include headache, mood changes, irritability, brain fog, GI upset, nausea, appetite and sleep changes, and blood pressure fluctuations. Transdermal DMSO application adds skin reaction risk at the application site.

Contraindications

Personal or family history of cancer (especially HGF/c-Met-amplified cancers: lung, gastric, renal). Active liver disease. Pregnancy or breastfeeding. Use alongside other growth factor or anabolic signaling compounds that may compound proliferative pathway load.

Theoretical concerns

The most serious theoretical risk is oncogenic. HGF/c-Met is a deeply studied oncogenic signaling axis implicated in lung, breast, colon, kidney, liver, gastric cancer, and glioblastoma. Chronic or repeated agonism of this pathway in a person with subclinical or precancerous cell populations is biologically plausible as tumor-promoting. Rodent data shows dose-dependent hepatotoxicity at higher doses; we could not verify the source for a reported rat study showing no hepatotoxicity, so no liver-safety threshold is stated. No long-term oncological endpoint data exists in any species.

What nobody knows yet

Human pharmacokinetics and bioavailability. Safe dose range in humans. Long-term safety at any dose. Drug interactions. Whether c-Met agonism from systemic dosing actually reaches the CNS at meaningful levels in humans. Whether the HGF/c-Met mechanistic hypothesis even holds, given the retraction of its foundational papers.

What else could explain the result

Community self-experimenters almost universally stack Dihexa with other nootropics (racetams, peptides, adaptogens, modafinil/armodafinil, caffeine). TRT in male self-experimenters adds a baseline cognitive and energy effect that inflates any perceived Dihexa signal. Many stack with BPC-157 or other healing peptides, further obscuring attribution. Placebo effects are particularly strong with compounds that carry heavy 'most potent nootropic ever' branding online.

Is the vial what it says?

Cross-checked against the cited sources below. The grade reflects the absence of formal human safety and efficacy data.

Half-life and storage

Half-life: Unknown in humans. Animal pharmacokinetic studies suggest extended duration relative to earlier angiotensin IV analogs due to the metabolic stabilization introduced in the PNB-0408 design. No human half-life data exists.

Once mixed: Dihexa is typically sold as a raw powder. It has reported poor water solubility; many community protocols dissolve it in DMSO, ethanol, or PEG-400 before use. Standard peptide reconstitution in bacteriostatic water is not reliable due to solubility limitations. Solubility must be confirmed in chosen vehicle before dosing.

What people report

Volume of reports: faint. Little real-world record, or mostly noise.

Moderate community discussion exists on Longecity, Reddit nootropic forums, and biohacking communities, mostly concentrated between 2015-2022. Discussion has dropped since the 2025 retractions.

Inconsistent. Positive reports of enhanced focus and verbal fluency sit alongside reports of persistent brain fog and mood disruption. No clear dose-response pattern emerges from anecdotes.

Low. No controlled reports. Heavy baseline stacking makes attribution impossible. Strong placebo priming from extreme potency marketing. The foundational science is now retracted.

• ANECDOTE: Some users report noticeable improvements in verbal fluency and word retrieval within days of starting, often described as the most striking effect.
• ANECDOTE: A subset of users report persistent brain fog, mental fatigue, or emotional blunting that outlasts a cycle and is slow to resolve after stopping.
• ANECDOTE: Users who stack with modafinil or racetams frequently report difficulty separating any Dihexa effect from the baseline stack, and many attribute effects to Dihexa based on timing alone.
• ANECDOTE: After the 2025 retraction news circulated in biohacking communities, several long-term users reported discontinuing use and noted no obvious cognitive decline on cessation, raising questions about whether earlier perceived benefits were real.

Reports are anecdote, not proof.

Can you tell if it is working?

First documented human use: No controlled human trial has been completed.

1. 2012Kawas, Wright, and Harding at WSU publish the first paper describing angiotensin IV analog binding to HGF and its role in c-Met activation (JPET, PMID 22129598). This becomes the mechanistic foundation for Dihexa's claimed effects.
2. 2013McCoy et al. (WSU) publish behavioral and in vitro synaptogenesis data showing cognitive effects of Dihexa-class molecules in rodents (PMC3533412). This paper receives an Expression of Concern in 2021 and remains published but flagged.
3. 2021WSU investigation into co-author Leen Kawas (later CEO of Athira Pharma) finds image manipulation across multiple papers. Kawas resigns as Athira CEO. Four papers earn Expressions of Concern. Athira's Phase 2 Alzheimer's trial continues under new leadership.
4. 2022Athira Pharma's ACT-AD Phase 2 trial of fosgonimeton (ATH-1017), the closest human analog to the Dihexa compound class, misses primary and secondary endpoints in mild-to-moderate Alzheimer's disease patients. Mixed exploratory signals noted in monotherapy arm.
5. 2024Athira Pharma announces topline results from the larger Phase 2/3 LIFT-AD trial of fosgonimeton (September 3, 2024; published 2025, primary analysis N=287, 26-week randomized placebo-controlled). The trial did not meet its primary endpoint (Global Statistical Test) or key secondary endpoints (ADAS-Cog11, ADCS-ADL23). This is the second consecutive failed human trial for the compound class. Results were presented at CTAD 2024 in October/November 2024.
6. 2025Both core Dihexa mechanism papers formally retracted in April 2025 (Kawas 2012 retraction PMID 40312092; Benoist 2014 retraction PMID 40312093) after WSU investigation confirms data fabrication. The biochemical foundation of Dihexa's marketed mechanism is now formally unsupported in peer-reviewed literature.

• Retraction notice for Benoist et al. 2014 (PMID 40312093), PubMed(Official JPET retraction of the primary mechanistic paper. Confirms data fabrication finding.)
• McCoy et al. 2013 PMC3533412 (Expression of Concern, not retracted)(The only unretracted primary Dihexa paper. Carries a 2021 Expression of Concern. Rodent behavioral and in vitro data only.)
• Fosgonimeton Phase 2 ACT-AD results, NeurologyLive(Human evidence for the compound class. Related HGF mimetic missed primary endpoints in Alzheimer's Phase 2.)
• Retraction Watch coverage of Athira/Kawas investigation(Establishes the research integrity timeline and which papers were flagged.)
• Fosgonimeton in mild-to-moderate Alzheimer's disease (LIFT-AD), J Alzheimers Dis Rep 2025, PMID 41393340(Second failed human trial for the compound class. Phase 2/3, primary analysis N=287, 26-week RDBPC. Did not meet primary endpoint (GST) or key secondary endpoints (ADAS-Cog11, ADCS-ADL23). Presented at CTAD 2024.)
• McCoy AT et al. J Pharmacol Exp Ther 2013;344(1):141-154. PMID 23055539. Carries a 2021 expression of concern (PMID 34551989); the same group's 2012 and 2014 mechanism papers were retracted in April 2025(dose table)
• McCoy AT et al. J Pharmacol Exp Ther 2013;344(1):141-154. PMID 23055539. Carries a 2021 expression of concern (PMID 34551989); the same group's 2012 and 2014 mechanism papers were retracted in April 2025(dose table)
• McCoy AT et al. J Pharmacol Exp Ther 2013;344(1):141-154. PMID 23055539. Carries a 2021 expression of concern (PMID 34551989); the same group's 2012 and 2014 mechanism papers were retracted in April 2025(dose table)
• FDA, Certain Bulk Drug Substances for Use in Compounding that May Present Significant Safety Risks, content current as of 04/22/2026(dose table)