Semax won an FDA advisory vote. FDA's own reviewers had found two usable human studies.

Why this note exists

An advisory vote is the loudest regulatory event a research-market peptide ever gets, and it is the one most likely to be reported as though something had been proven. Semax is the clearest case in the July 2026 batch, because FDA published its own analysis six weeks beforehand and the analysis says, in plain language, how thin the human record is.[1] Both things are true at once: the committee voted yes, and the file it voted on is nearly empty of the evidence people assume is behind it.

In the index generated on September 3, 2026, 92 of 104 vendors carried a semax listing — 142 listings, eighth of 83 compounds by vendor coverage, ahead of every other nootropic peptide except selank.[17][18] Thirteen of those 142 listings name a nasal, spray or intranasal product; the rest are powder or solution sold by the vial, most commonly 10 mg.[17] That ratio is most of this note in one line, and the reason for it is in the next two sections.

A note on what kind of evidence this is: the human figures below carry their design and participant count, and every one of them is small. One section reports rodent pharmacology and is labeled as such throughout. Where FDA declined to consider a study, this note says why FDA declined and does not treat the exclusion as a verdict on the study. Doses appear only as their sources published them, with route and population attached.

The nomination

What was actually asked for, and by whom

Two compounders nominated semax-related substances for the 503A Bulks List: Wells Pharmacy Network and LDT Health Solutions.[1] FDA records that "[t]he nominators provided inconsistent information in the nomination packages regarding the specific BDS proposed" — it was unclear in both packages whether the nomination was for semax acetate or semax free base, which are different active pharmaceutical ingredients and therefore different substances.[1] Both nominations were later withdrawn, and semax still appears on FDA's public list of substances that may present significant safety risks under the heading "Bulk drug substances nominated but withdrawn."[1][7] FDA evaluated the substances anyway, "at its discretion" and "on its own initiative."[1]

The proposed products matter more than they look. FDA states: "[t]he semax-related drug products proposed in the nominations are intranasal spray or subcutaneous injection in 7,500 µg/mL or 1,000 µg/mL dosage strengths."[1] The nominated uses were cerebral ischemia, migraine and trigeminal neuralgia. Two other nominated uses did not survive triage: ADHD "was not evaluated because supporting literature for this use was not found," and "nootropic" was folded into the cerebral-ischemia evaluation because it "does not have an ICD-10 code and no professional society treatment guidelines could be found for this use."[1]

For orientation: there is no USP or NF monograph for either substance, neither is a component of an FDA-approved drug, and FDA found no monograph in the European, Japanese or International Pharmacopoeias.[1] Semax "is a registered drug in Russia and is available as 0.1% and 1% nasal drops."[1] Nothing about that Russian registration transfers to the United States, and nothing about it makes a vial an approved product anywhere.

The human record

One 1996 study of 37 people, and a meeting abstract

FDA searched PubMed, Embase, the Cochrane Database of Systematic Reviews and ClinicalTrials.gov, plus the references the nominators supplied.[1] What it found for effectiveness was two items.

The first is Koroleva and colleagues, 1996, which supplied both the migraine and the trigeminal-neuralgia evidence. In the migraine group, 12 subjects — 10 women and 2 men, aged 19 to 56 — received one intranasal dose of 0.5 mg/kg.[1] FDA records that the authors reported a rating-scale score falling from 78% to 32%, that "[f]our subjects out of twelve (33%) reported cessation of headache pain 90-120 minutes after semax administration," and that in the other 8 "the effect of semax was weaker; 'pain was not eliminated but became less severe.'"[1] FDA's listed limitations are the ones that decide the question: "the small sample size, no blinding, no control group, the absence of reporting of the actual scores or standard deviations for the scores of the rating scales, and there were insufficient details provided on the design and the conduct of this study."[1]

The trigeminal-neuralgia arm of the same study is the more pointed result, because it is negative. Twenty-five adults were studied: 16 with typical trigeminal neuralgia and 9 with dental plexalgia.[1] In the typical trigeminal neuralgia subjects, FDA reports, "there were no changes in the characteristics of pain in the MPST nor changes in sensation or pain thresholds and TSEP after administration of semax," and the authors themselves concluded that "no significant changes in TSEP were observed after a single intranasal administration of semax. This observation indicates that semax does not exhibit analgesic activity by itself."[1]

The second reference is Cherkasova and colleagues, 2002 — and it is a meeting abstract, not a paper. FDA states that "[n]o full published article could be located for this reference," that "[a]n unknown number of human subjects with chronic ischemic brain disease received intranasal semax 600 mg daily for 10 days," and that "the authors do not provide full results for these lab values, nor do they discuss clinical function before and after receiving semax."[1] That abstract is the entire cerebral-ischemia effectiveness file that FDA was able to consider.[1]

Counts and design details as recorded in FDA's May 2026 evaluation; registry count from a ClinicalTrials.gov search run for this note.[1][15]

FDA's summary sentence is worth reading twice, because it is stronger than "we don't know": "There is insufficient evidence to support the effectiveness for semax (free base) or semax acetate for cerebral ischemia, migraine, or trigeminal neuralgia. Only two available references were available with insufficient details on the design and conduct of the studies, and the available references demonstrated lack of effectiveness and were limited by small sample size and lack of information about some relevant clinical endpoints."[1]

The exclusion

The Russian literature was set aside on a filing rule, not on the merits

The obvious objection to everything above is that semax has been studied in Russia for decades. FDA's review answers that directly, and the answer is procedural. It names four Russian-language references on cerebral ischemia — Miasoedova 1999, Gusev 1997, Gusev 2005 and Gusev 2018 — and states that "[b]ecause the full references were not available in English, these references were not considered as part of this evaluation," citing the regulation that requires foreign-language material submitted to FDA to arrive with a verified, complete translation.[1][9] It adds that the review articles proposing semax for cerebral ischemia rest on animal data, and that the human references in them "are written in Russian" or are "in healthy adults."[1]

That is a rule about what FDA may formally consider. It is not a finding that the excluded work is unsound, and it is equally not a reason to assume the excluded work would have carried the day. The largest of those papers is Gusev and colleagues, 2018, in a Russian neurology journal: 110 patients after ischemic stroke, split into early and late rehabilitation groups, each subdivided into semax-treated and untreated subgroups, with plasma BDNF, a motor scale and the Barthel index as outcomes.[12] The English abstract available through PubMed describes subgroups rather than randomization and does not describe blinding or a placebo, and the full text is in Russian.[12] pepmg's own semax dose record comes from that paper and is labeled with exactly those limits — a published human regimen, reported as published, not a recommendation.[12]

What has appeared in the English-language literature more recently is not clinical. Two small imaging studies from a Moscow group gave healthy volunteers intranasal semax and measured resting-state fMRI: 24 volunteers, of whom 14 received semax and 10 placebo, reporting a difference in the extent of a default-mode-network subcomponent (2018); and 52 healthy participants across semax, selank and placebo, reporting differences in amygdala connectivity (2020).[10][11] Both measured brain imaging signals in healthy people, not symptoms in patients. The other recent English-language work is a rat transcriptome study of cerebral ischaemia-reperfusion (2020) and an artificial-membrane study of copper-induced amyloid aggregation (2022) — animal and in vitro, respectively.[13][14]

Animal data, labeled as such

The rodent file is substantial, and one of its findings is about route

Everything in this section is rodent pharmacology reported in FDA's review; none of it is human evidence, and none of it establishes an effect in people.[1] In a rat model of focal photothrombotic ischemia, infarcted brain areas were about 20% smaller in rats given intranasal semax at 250 µg/kg than in vehicle-treated rats, and treated rats did not show the memory-retention deficit the vehicle group did.[1] In rats with global cerebral ischemia from carotid occlusion, semax at 100 µg/kg intraperitoneally reduced histological ischemic damage.[1] Other rat work reports increases in nerve growth factor and brain-derived neurotrophic factor expression after intranasal dosing.[1]

FDA then notes the limitation that matters for a market selling vials: those studies are "generally restricted to a fixed dose of semax (100 µg/kg, IP, in some studies or 250 µg/kg, IN, in others). As such, dose-response relationships for semax delivered via different ROAs have not been established."[1] The sharpest single line in the whole animal section is on analgesia: intraperitoneal semax raised pain thresholds in rats dose-dependently, and "the same semax doses delivered intranasally were unable to induce analgesia."[1] In rodents, the route changed the result entirely. No human study has compared the two routes, because no human study of the injected route exists.[1]

Two rodent findings run in the risk direction rather than the benefit direction. Rats given intranasal semax at 1 mg/kg/day for five days showed increased anticoagulant and fibrinolytic activity and developed smaller thrombi, which FDA carries forward as a bleeding concern.[1] And in mice, "semax also potentiated amphetamine-induced dopamine release in the striatum," which FDA calls concerning "because increased dopaminergic tone in the striatum is a response typically induced by drugs of abuse such as cocaine," while recording that no published nonclinical toxicity studies exist for the nominated subcutaneous route.[1]

Safety

No human pharmacokinetics, and no data at all on the route being sold

Two sentences from the review carry this section. "We were not able to find pharmacokinetic studies in humans following exposure to semax (free base) or semax acetate via any ROA."[1] And: "We were unable to find literature that discussed administration via subcutaneous injection ROA. The only ROA discussed in the literature was intranasal. Therefore, assessment of safety for semax was limited to the intranasal ROA."[1]

The total identified human exposure, across every reference FDA and the nominators could find, is "33-47 healthy adults, 69 adults with medical conditions (including chronic ischemic brain disease, pain due to migraine and trigeminal neuralgia, and peptic ulcer), and 451 children with either depression or tics/Tourette Syndrome."[1] Three of those references are meeting abstracts, the pediatric full studies were unavailable, and in several the peptide was given alongside other drugs, "making it difficult to determine the contribution of semax to any effects seen."[1] Adverse events were simply not discussed in most of them; one reference reported that none were observed.[1] The doses FDA tabulates, all intranasal, are "0.5 mg/kg once, 1 mg daily for 2 days, 600 mg daily for 10 days, or 2-4 drops per nostril three times a day for 10 days."[1]

A single spontaneous report is not an incidence rate and does not establish cause, and FDA's own footnote explains why the number is uninformative in either direction: compounders under section 503A generally do not report adverse events to FDA, so "[u]nless an adverse event report is submitted to FDA, the Agency may not be aware of adverse events associated with a product compounded under section 503A."[1] An empty safety database is not a safety record.

The rest of the safety discussion is about the injectable form specifically. FDA writes that it "did not identify clinical studies assessing immunogenicity or aggregation of semax (free base) or semax acetate," and that it "is concerned about their potential for immunogenicity when administered by an injection ROA as proposed due to the potential for aggregation as well as potential peptide-related impurities."[1] Its final proposal cites, among other things, "lack of endotoxin data for injectable routes of administration."[1] FDA's public risk table states the same conclusion in one line: compounded semax "may pose risk for immunogenicity for certain routes of administration due to the potential for aggregation and peptide-related impurities. FDA has no, or limited, safety-related information for proposed routes of administration. Therefore, the agency lacks sufficient information to know whether the drug would cause harm if administered to humans."[7]

The market

One common name, several different molecules

FDA devotes its characterization section to a problem that is visible in any peptide catalogue: "Semax is a common name and not a United States Adopted Name (USAN). FDA has encountered multiple salts, and derivatives, including different active moieties, sold commercially under the same common name for similarly situated products. Inconsistent naming conventions that do not follow established chemical nomenclature standards … represent a safety risk for patients as they may be dosed with a different BDS than the physician ordered."[1]

pepmg's index shows the shape of that. Of 142 semax listings from 92 vendors in the index generated September 3, 2026, 21 listings from 19 vendors are sold under a modified name — most often N-Acetyl Semax Amidate or NA-Semax — while FDA's evaluation covers exactly two substances, semax free base and semax acetate.[17][1] Listed prices run from $0.84 to $11 per mg, a thirteen-fold spread for a name.[17] pepmg makes no claim about what is in any vendor's vial; the point is narrower and comes from FDA — the name on the label does not identify the substance.[1]

The other mismatch is the route. The entire human literature FDA could find is intranasal; the Russian registered product is nasal drops; 13 of the 142 listings in pepmg's index name a nasal or spray product, and the rest are sold by the vial.[1][17] FDA's survey of the US market found the same split from the other side: websites of "health/wellness clinics, medical concierge services, functional and regenerative medicine clinics, and online retailers" marketing semax injection and intranasal products, one retailer listing "nasal spray and subcutaneous injection as common methods of administration," some selling a "50:50" combination of semax and selank, and "[n]one of the websites searched indicate whether their marketed products contain the base or a salt or ester of semax."[1]

Selank is the compound to watch next for the same reason, and it is bigger here: 98 vendors, third of 83 compounds by coverage in the same index.[18] FDA's semax review explicitly set it aside — a nominator-submitted article on selank's analgesic effects "is out of the scope of this evaluation" — and selank was not among the seven substances the committee voted on in July.[1][3]

What the vote changes

Advice, then rulemaking, then a list of what pharmacies may compound

The committee's recommendation is the beginning of a process, not the end of one. FDA describes advisory committees as bodies that "provide independent advice" and states that the agency "is not bound to follow" their recommendations.[8] For this list specifically, FDA says it "will continue to evaluate bulk drug substances that have been nominated with sufficient supporting information and address those substances on a rolling basis through notice-and-comment rulemaking," and that it is "consulting with the Pharmacy Compounding Advisory Committee … for input as it evaluates them."[6]

Three things the vote does not do. It does not make semax an approved drug: nothing changed in Drugs@FDA, and the 503A list governs what a compounding pharmacy may use, not whether a substance is safe and effective.[16][6] It does not make a research vial legal, priced or pure — a vial bought online was never a compounded prescription in the first place. And it is not a finding of efficacy; FDA's file for the meeting says the opposite in writing, and the committee's own vote count records five members who agreed with it.[1][4]

It is also not nothing. A recommendation from this committee is one input to FDA’s compounding review, and it happened for six of the seven peptides on the agenda; emideltide, the one FDA-reviewed substance the committee rejected, went down 6–7.[4] pepmg covered the full batch of votes and what they legally do in a separate note.

Three things this note is not saying

First, it is not saying semax does not work. Nobody has run the study that would answer that. The honest statement is that the evidence FDA could evaluate does not support effectiveness for the three nominated uses, and that a compound with two uncontrolled human references is untested rather than disproven.[1]

Second, it is not saying the Russian research is worthless. It is saying that FDA did not consider it because it was not filed in English, that the largest of those papers describes subgroups rather than a randomized comparison in its English abstract, and that a body of work no US regulator has been able to assess cannot carry the weight the market puts on it.[1][12]

Third, it is not saying the committee was wrong. Members were asked whether a substance should be available to compounding pharmacies under supervision — a different question from whether it works, and one where "patients are already buying this unsupervised" is a legitimate argument. The vote was close, and the record it rests on is the one described above.[4]

What it is saying is narrow and checkable: as of September 3, 2026, the published human record for semax that FDA was able to evaluate consists of one 1996 study of 37 people given a single intranasal dose and one meeting abstract with no reported sample size, there are no registered trials, there is no human pharmacokinetic data by any route, and there is no human safety data for the injected route that most of this market sells.[1][15][17]

Questions people are asking