What published research reports on Semax in stroke and cerebral ischemia

Published August 26, 2026 · Artemis Labs

Semax — answer capsule: Semax has been studied in stroke and cerebral ischemia only in Russia. Four PubMed records for Semax carry a clinical-trial publication type, and all four came from Russian institutions and Russian journals. The largest is a 2018 report on 110 patients after ischemic stroke from Pirogov Russian National Research Medical University in Moscow, whose abstract states no randomization, no placebo group and no blinding. The report with the strongest wording about stroke risk describes no control group at all, so its result cannot be separated from the ordinary course of the illness. PubMed indexes no meta-analysis, no systematic review and no double-blind study of Semax, and nothing in this area has been reproduced by a group outside Russia.

Key findings

• Four PubMed records for Semax carry a clinical-trial publication type — PMIDs 29798983, 18379501, 10741256 and 11517472 — and zero of them were conducted outside Russia.
• The largest stroke dataset is PMID 29798983: 110 patients after ischemic stroke at Pirogov Russian National Research Medical University in Moscow. PubMed types it Clinical Trial, not Randomized Controlled Trial, and the abstract states no randomization, no placebo and no blinding.
• The strongest efficacy wording in the entire Semax human record — significant clinical improvement, stabilization of the disease progress and reduced a risk of stroke (PMID 15792140, Russia, 2005) — comes from a report that describes no control arm at all.
• The one record PubMed tags as a randomized controlled trial (PMID 18379501, Russia, 2007) was not a stroke trial, calls itself open-label in its own abstract, and was negative on its primary endpoint.

Which Semax stroke studies exist, and who ran them?

Provenance first, because it changes how every number below should be read. Semax is a synthetic seven-amino-acid peptide developed and studied in Russia, where it is handled as a domestic pharmaceutical. A PubMed search for records carrying any clinical-trial publication type returns four — all Russian-institution or Russian-journal papers, three published in Russian, none run in the United States or Europe. That is not a criticism of Russian science. It is a structural fact about the evidence: when every study of a compound comes from one country, no independent group has ever checked the results.

What did the 110-patient 2018 ischemic stroke study report?

This is the deepest human dataset in the record. It ran at Pirogov Russian National Research Medical University in Moscow and appeared in a Russian neurology and psychiatry journal in 2018 (PMID 29798983). It is indexed as [The efficacy of semax in the tretament of patients at different stages of ischemic stroke] — the misspelling is PubMed’s own, and we quote it rather than silently fix it.

The design, in the authors’ words: One hundred and ten patients after IS (43 men, 67 women, mean age 58.0±9.7, Ме 63 years) were examined. All patients were divided into early (89±9 days) and late (214±22 days) rehabilitation groups. Each group was subdivided into semax+ and semax- subgroups. There were comparison subgroups, but no placebo, no described randomization and no described blinding. PubMed types the record Clinical Trial, and specifically not Randomized Controlled Trial.

The reported result: Administration of semax and high BDNF levels accelerated the improvement and ameliorated the final outcome of Barthel score index. BDNF is brain-derived neurotrophic factor, a protein involved in the growth and upkeep of nerve cells; the Barthel index rates how much daily activity a patient can manage without help.

Here is the limit that matters. These are correlations, and the peptide was given alongside a rehabilitation program. The authors’ own summary is joint — Early rehabilitation and administration of semax increase BDNF plasma level, speed functional recovery, and improve motor performance. That design cannot separate the peptide from the timing and intensity of rehabilitation. Anyone reporting this as “Semax improved stroke recovery” has dropped half the sentence.

What did the earlier acute-stroke comparison report?

A 1997 report in the same Russian journal (PMID 11517472) studied 30 patients in acute period of hemispherical ischemic stroke. PubMed lists no institution for this record; the country is read from the journal and the language. Its comparison group was assembled separately: Control group consisted of 80 patients with the strokes analogous in severity and location of the damages and which were treated by conventional therapy. Thirty against eighty, not randomized, not blinded, no placebo, and the peptide was added on top of conventional treatment rather than tested on its own. The finding keeps its own hedge, and so do we: it had some influence on the rate of restoration of the damaged neurological functions. That wording should never be upgraded to “improved.”

Why the strongest-sounding stroke result is the weakest evidence

One sentence in this literature gets quoted more than any other, and it needs its design attached every time. A 2005 Russian report examined One hundred and eighty-seven patients with different stages of cerebrovascular insufficiency (CI) and concluded: Semax treatment resulted in significant clinical improvement, stabilization of the disease progress and reduced a risk of stroke and transitory ischemic attacks in the disease course. That is the strongest efficacy language anywhere in the Semax human corpus (PMID 15792140). PubMed lists no institution for it either.

It sits on the weakest stated design in the corpus. No control or comparison arm is described anywhere in that abstract. PubMed does not type the record as a clinical trial of any kind; it carries Comparative Study, Evaluation Study, Journal Article. There is no randomization and no blinding. Without a control group there is no way to separate an effect of the compound from the ordinary course of the condition, from other treatment the patients received, or from the expectations of people who all knew what was being given. The same abstract’s tolerability line — a minor percent of side-effects — is uncontrolled and unquantified for the same reason. On this page that result never travels without the limitation attached in the same breath.

What happened in the only randomized controlled trial?

One Semax record in all of PubMed carries the Randomized Controlled Trial publication type, and it was not a stroke study. It examined 27 Russian patients with motor neuron disease in 2007, in the same Russian journal, with no institution listed (PMID 18379501). Two things about it matter. First, the abstract contradicts its own index tag, describing The open-label clinical trial of Semax … conducted in sequential groups of patients. Open-label means everyone knew who was getting what. Second, it was negative on its primary endpoint: Semax does not influence either the course of CPD or the dynamics of clinical estimates, in particular the terms of ensuing marked functional deficits on bulbar, cervical and lumbosacral levels of segmental innervation. A secondary quality-of-life measure did move, with the authors’ hedge preserved. The question the trial was built to answer came back negative.

What do the mechanism and animal studies add?

A 1999 Russian paper on acute ischemic stroke described A retrospective comparative clinicoimmunobiochemical analysis reporting an activating influence of Semax on antiinflammatory postischemic reactions in the brain (PMID 10358912). Retrospective means researchers looked back at records already collected, the weakest of the human designs. One caution: the next sentence of that indexed abstract labels a cytokine in a way that contradicts standard immunology, so we do not reproduce it.

The animal work does not close the gap. In preclinical models of acute oxygen deprivation, one comparison found that Neither mexidol nor semax upon single administration were effective on the models of acute hemic and histotoxic hypoxia, and reported bell-shaped reversible dose-effect relationships (PMID 20486550). A bell-shaped curve rises and then falls, so more is not simply better. For background on this compound class, see our overview of neuropeptide research.

What the research does not show

No Semax human study has ever been conducted outside Russia. That is verified, not assumed: every clinical-trial-typed record is Russian, and so is the broader human literature of roughly eighteen records. The only non-Russian primary research is not human at all — Italian laboratory work in artificial membranes and in a cell line, plus one Indonesian rat study. No independent replication of the Russian human stroke results exists in either direction.

The stroke-specific gap is the missing control arm. The report with the boldest wording about reduced stroke risk (PMID 15792140) describes none at all. The largest stroke dataset (PMID 29798983) has comparison subgroups but no placebo, no randomization and no blinding, and gave the peptide alongside rehabilitation. The 1997 comparison (PMID 11517472) used a separately assembled control group of a different size.

PubMed indexes zero meta-analyses of Semax, zero systematic reviews and zero double-blind studies. Reviews from outside Russia say the same in their own words: a 2026 review by sports-medicine and orthopaedic authors in the United States notes of this class of peptides that there is a current lack of clinical trials (PMID 41490200), and a 2026 gerontology review that searched regulatory databases through January 2026 places Semax among Non-approved peptides that lack long-term safety data and systematic validation (PMID 42021992).

One human safety observation belongs beside all of this: a 1999 Russian follow-up of 73 patients with posthypoxic encephalopathy reported that In some cases EEG showed episodes of paroxysmal activity after semax injection, and the authors attached a brain-monitoring caution (PMID 10199046). That finding and the rest of the safety record are covered on our Semax safety research page; the trial-by-trial breakdown is on our Semax human trials page. Finally, no FDA or EMA approval for Semax can be established from any primary source. Russian authors describe it in their own papers as a registered domestic drug, but no registry record was reachable, so that is their description and not our assertion.

Frequently asked questions

How many Semax stroke trials are there?

Two clinical-trial-typed records: PMID 29798983 (110 patients, Moscow, 2018) and PMID 11517472 (30 against 80 controls, Russia, 1997). A third Russian report, PMID 15792140, commented on stroke risk in 187 patients but is not typed as a trial and describes no control arm.

Was any of this research done outside Russia?

No. Zero Semax human studies of any kind were run outside Russia. The non-Russian primary research is entirely laboratory and animal work from Italy and Indonesia.

Did the randomized trial support use in stroke?

No. It studied motor neuron disease, not stroke, and reported that the compound does not influence its primary endpoint.

Why does one study sound so much more positive than the others?

Because it had the least rigorous design. PMID 15792140 pairs the strongest wording with no described control group — a pairing common enough in clinical research that design is reported here beside every result.

References

1. Efficacy of semax at different stages of ischemic stroke — 110 patients, Pirogov Russian National Research Medical University, Moscow, Russia, 2018. PMID 29798983 · DOI 10.17116/jnevro20181183261-68. Title as indexed: [The efficacy of semax in the tretament of patients at different stages of ischemic stroke]. (typo is in the PubMed record, quoted as-is)
2. [Effectiveness of semax in acute period of hemispheric ischemic stroke (a clinical and electrophysiological study)]. Russia, 1997. PMID 11517472 (no DOI in the PubMed record)
3. [Semax in prevention of disease progress and development of exacerbations in patients with cerebrovascular insufficiency]. Russia, 2005. PMID 15792140 (no DOI in the PubMed record)
4. [The study of chronic partial denervation and quality of life in patients with motor neuron disease treated with semax]. Russia, 2007 — the only record typed Randomized Controlled Trial; negative on its primary endpoint. PMID 18379501 (no DOI in the PubMed record)
5. [Investigation of mechanisms of neuro-protective effect of semax in acute period of ischemic stroke]. Russia, 1999 — retrospective analysis. PMID 10358912 (no DOI in the PubMed record)
6. [Use of semax at a follow-up of patients with posthypoxic encephalopathy]. Russia, 1999 — 73 patients; EEG observation. PMID 10199046 (no DOI in the PubMed record)
7. Preclinical comparison in models of acute hemic and histotoxic hypoxia; bell-shaped dose-effect relationship reported. PMID 20486550 (no DOI in the PubMed record)
8. 2026 review, United States (Pacific Coast Sports Medicine, Los Angeles; Lenox Hill Hospital, New York; Redox Medical Group, Beverly Hills) — there is a current lack of clinical trials. PMID 41490200 · DOI 10.5435/JAAOSGlobal-D-25-00236
9. 2026 review searching regulatory databases through January 2026; classifies Semax among Non-approved peptides. PMID 42021992
10. [Evaluation of therapeutic effect of new Russian drug semax in optic nerve disease]. Russia, 2000 — the fourth clinical-trial-typed record. PMID 10741256 (no DOI in the PubMed record)

Methodology: this page draws only on primary PubMed records retrieved through NCBI E-utilities and PubChem, transcribed verbatim; study design, country and institution were read from each record rather than from secondary summaries. Last verified August 26, 2026.

All compounds sold by Artemis Labs are for laboratory research use only. Nothing on this page is medical advice, and no statement has been evaluated by the FDA.