# Mexidol, bemethyl and phenylpiracetam

Rating: Early — Mostly animal or cell studies; few or uncontrolled human data
Type: Guide
Updated: 2026-11-02
Canonical: https://heretic.bio/articles/mexidol-bemethyl-and-phenylpiracetam.html

> Mexidol (emoxypine succinate) has two randomized placebo-controlled stroke trials, EPICA (n=150) and MIR (n=304), both reporting less disability. Both appear in one Russian journal, and MIR was sponsored by Pharmasoft; its main effect was a 0.44-point difference on the modified Rankin Scale. Neither tested cognition in healthy people. Bemethyl has a pharmacokinetic study and one heat-tolerance study, with no cognition trial. Phenylpiracetam (fonturacetam) has one placebo-controlled epilepsy trial without reported effect sizes and an asthenia meta-analysis with no stated placebo arm, and it is on the 2026 WADA Prohibited List.

Three Russian drugs turn up again and again in nootropic forums and Eastern European pharmacies: Mexidol (emoxypine succinate, chemically ethylmethylhydroxypyridine succinate), bemethyl (bemitil, sold as Metaprot) and phenylpiracetam (phenotropil, now called fonturacetam and sold as Actitropil). All three are registered medicines in Russia under the brand names above. None has FDA approval. This page lays out what the human studies actually show, who ran them, and where the gaps are.

The short version: Mexidol has two placebo-controlled stroke trials, both run in Russia, both reported in one journal, and the larger one sponsored by Pharmasoft, the company named on the Mexidol trial registration. Bemethyl has almost no human data. Phenylpiracetam has one placebo-controlled trial with no effect sizes in its abstract, a meta-analysis with no placebo comparison, and a place on the 2026 WADA Prohibited List.

Not medical advice. Doses below are what the cited trials gave their participants, not a regimen.

## What these compounds are

Name in papers| Chemistry| Often confused with  
---|---|---  
Mexidol, emoxypine succinate, ethylmethylhydroxypyridine succinate| The succinate salt of a 3-hydroxypyridine, structurally compared with vitamin B6 ([PMID 35992374](https://pubmed.ncbi.nlm.nih.gov/35992374/)). The stroke trials used the branded drug.| Emoxypine base sold online. It is a related compound, not the preparation that was tested.  
Bemethyl, bemitil, Metaprot| 2-ethylthiobenzimidazole, sold in capsules ([PMID 21870773](https://pubmed.ncbi.nlm.nih.gov/21870773/))| [Bromantane](bromantane.html), another Soviet "actoprotector" with a different structure  
Phenylpiracetam, phenotropil, fonturacetam, carphedon, Actitropil| 2-(2-oxo-4-phenylpyrrolidin-1-yl)acetamide. Clinical use is the racemate, a mix of two mirror-image forms ([PMID 21689376](https://pubmed.ncbi.nlm.nih.gov/21689376/)).| Piracetam. The class is covered on [racetams](racetams.html).  
  
## How to read this literature

Almost every positive clinical paper on Mexidol and phenylpiracetam appears in one Russian journal, the _Zhurnal Nevrologii i Psikhiatrii imeni S. S. Korsakova_. It is a real, indexed journal. But when nearly all the good news for a drug comes from one journal and one country, and the sponsor is rarely named in the abstract, independent replication is missing. That is not a reason to dismiss the work. Western regulators ignore much of it for a different reason: nobody with a patent to protect will pay for a Western trial of an old Russian molecule (see [off-patent compounds and missing trials](off-patent-compounds-and-missing-trials.html)). The fair reading is the one used on [Russian pharmacology](russian-pharmacology.html): check the chemical identity, the mechanism at real-world doses, the human evidence beyond custom, and the governance (sponsor, doping status, dependence).

## Mexidol: two stroke trials

### EPICA (2017)

A randomized, double-blind, placebo-controlled, multicenter trial in 150 adults aged 40 to 79 with hemispheric ischemic stroke ([PMID 28665371](https://pubmed.ncbi.nlm.nih.gov/28665371/)). Patients received Mexidol by intravenous infusion for 10 days and then tablets for 8 weeks, or a matching placebo, on top of standard care. At the end of treatment the Mexidol group had a lower mean disability score on the modified Rankin Scale (mRS, p=0.04), more patients with little or no disability (mRS 0 to 2, p=0.039) and a lower stroke-severity score on the NIHSS (p=0.035). Side effects were no more common than on placebo. The abstract gives p-values but not the scores themselves or the size of the difference, and it does not name a sponsor.

A 2020 re-analysis of the same patients split them by age ([PMID 33016677](https://pubmed.ncbi.nlm.nih.gov/33016677/)). It reported the clearest difference from placebo in the oldest group, aged 76 to 90 (p<0.001). The English and Russian versions of that abstract give different limits for the middle age band (60 to 65 versus 60 to 75), so that subgroup result cannot be pinned down from the abstract.

### MIR (2025)

The larger trial: 304 patients at 17 centers in Russia, Kazakhstan and Uzbekistan, randomized and double-blind, adults aged 40 to 75 with a first hemispheric ischemic stroke treated within 48 hours and without clot-removal or clot-dissolving therapy ([PMID 40898634](https://pubmed.ncbi.nlm.nih.gov/40898634/); [PMID 41456191](https://pubmed.ncbi.nlm.nih.gov/41456191/); [NCT06437626](https://clinicaltrials.gov/study/NCT06437626)). Mexidol was given intravenously for 10 days and then as tablets for 60 days, again on top of standard care.

The ClinicalTrials.gov record, which has posted results, names Pharmasoft as the lead industry sponsor. The journal abstracts do not. Here is what the registry shows at about day 71:

  * **Disability (primary outcome):** mean mRS improved by 2.45 points on Mexidol (n=141) and 2.01 on placebo (n=138). The difference is 0.44 points on a 0 to 6 scale. The companion paper reports p=0.003.
  * **Little or no disability (mRS 0 to 1):** 85 of 141 on Mexidol versus 57 of 138 on placebo (60% versus 41%).
  * **Severe disability (mRS above 3):** 22 of 141 versus 39 of 138.
  * **Stroke severity (NIHSS):** the median improvement was 7 points in both groups. The primary paper still reports a significant NIHSS difference (p<0.001); the registry's identical medians do not show one.
  * **Mobility and cognition:** median gains of 10 versus 9 points on the Rivermead Mobility Index and 4 versus 3 points on the MoCA cognitive test. The authors call the MoCA result a trend, not a significant finding.
  * **Safety:** adverse events in 23% of each group. Two deaths on Mexidol and one on placebo; six versus four patients with a serious adverse event. These numbers are too small to compare.



Two smaller discrepancies remain between the paper and the registry: the primary paper describes its main contrast with medians while the registry reports means, and the upper confidence limit for the Mexidol group is given as −2.32 in the paper and −2.23 in the registry.

Taken together, Mexidol has two positive placebo-controlled trials in acute stroke, which is more than most compounds on this site. Both come from the same research network and journal, the larger one is industry-funded, and the main effect is a modest shift on the disability scale. A trial run by an independent group, ideally outside Russia, would settle it. Neither trial tested Mexidol as a cognitive enhancer in healthy people, which is how it is usually sold online.

### What is not a Mexidol trial

Two 2020 papers in the same journal ([PMID 32678558](https://pubmed.ncbi.nlm.nih.gov/32678558/), [PMID 33449535](https://pubmed.ncbi.nlm.nih.gov/33449535/)) are reviews, not new patient data. The second one says early treatment within six hours works best; neither trial tested that, and MIR enrolled patients up to 48 hours after onset. A 2022 review ([PMID 35992374](https://pubmed.ncbi.nlm.nih.gov/35992374/)) covers patents, iron binding in the lab and antioxidant theory. The Italian "EPICA" atrial-fibrillation studies are unrelated and only share the acronym. ClinicalTrials.gov lists other completed Mexidol studies (chronic cerebral ischemia, childhood ADHD, vascular cognitive impairment, glaucoma), but only MIR has posted results.

## Bemethyl: almost no human data

Bemethyl was developed in the Soviet Union as an "actoprotector", a drug meant to keep soldiers and workers performing under heavy physical load. Its reputation rests on that history, not on published trials. ClinicalTrials.gov has no registered trial under any of its names, and PubMed has no human cognition trial.

The human records that do exist are thin:

  * A pharmacokinetic study in healthy volunteers after a single 250 mg capsule found peak blood levels about one hour after the dose and very uneven levels between people ([PMID 21870773](https://pubmed.ncbi.nlm.nih.gov/21870773/)). The abstract gives no number of participants or half-life.
  * A 1997 study in men doing physical work in insulated protective suits reported that one dose of bemethyl or phenibut raised heat tolerance, with phenibut the stronger of the two ([PMID 9162292](https://pubmed.ncbi.nlm.nih.gov/9162292/)). No sample size is given.
  * A study after brain injury tested bemethyl only in combination with pyrazidol, so it says nothing about bemethyl alone ([PMID 16404942](https://pubmed.ncbi.nlm.nih.gov/16404942/)).



A 2021 rat study mapped nine breakdown products in urine, mainly for drug testing ([PMID 34445727](https://pubmed.ncbi.nlm.nih.gov/34445727/)). It notes that WADA has had bemethyl on its Monitoring Program since 2018. That program tracks substances for possible future bans; it is not a ban.

## Phenylpiracetam (fonturacetam)

### Human studies

  * **Epilepsy, add-on (placebo-controlled):** 90 adults with focal epilepsy took phenylpiracetam or placebo alongside their usual medicines ([PMID 25591651](https://pubmed.ncbi.nlm.nih.gov/25591651/)). The authors report fewer seizures and better cognition in patients without epileptiform EEG changes, but the abstract gives no arm sizes, seizure counts or p-values. In 40% of patients it did not reduce the side effects of the epilepsy drugs.
  * **After brain lesions:** 99 patients treated for a month, with no control group ([PMID 16447562](https://pubmed.ncbi.nlm.nih.gov/16447562/)). It describes improvement over time, which cannot be separated from natural recovery.
  * **Heart patients with anxiety or low mood:** an open comparison of 35 patients on phenylpiracetam and 35 on another Russian drug, pantogam active, for 12 weeks, with no placebo ([PMID 25403298](https://pubmed.ncbi.nlm.nih.gov/25403298/)). Pantogam active did better on several measures, and the calming effect of phenylpiracetam faded after four to eight weeks.
  * **Fatigue (asthenia), 2025 meta-analysis:** 11 studies, 549 patients, all at the same daily dose ([PMID 40047835](https://pubmed.ncbi.nlm.nih.gov/40047835/)). Fatigue scores (MFI-20) fell by an average of 16.3 points after a month (95% CI 8.85 to 23.85), and 5.5% had side effects described as transient. This is a before-and-after change in treated patients. The abstract does not say whether any included study had a placebo group, so the result cannot be read as an effect of the drug. The paper's objective names the Actitropil brand.



Phenylpiracetam has no trial in healthy people, the group that buys it as a nootropic.

### Animal work

In mice, the R-form increased movement and improved memory in a passive-avoidance test, and the S-form did not improve memory ([PMID 21689376](https://pubmed.ncbi.nlm.nih.gov/21689376/)). Rat and mouse studies compare it with piracetam after reduced blood flow to the brain ([PMID 17523446](https://pubmed.ncbi.nlm.nih.gov/17523446/)) and test the R-form against inflammation ([PMID 32279140](https://pubmed.ncbi.nlm.nih.gov/32279140/)). Animal doses do not translate to human doses, and the claim that the R-form is "the active one" in people rests on this mouse work alone.

### Banned in sport

The [2026 WADA Prohibited List](https://www.wada-ama.org/sites/default/files/2025-09/2026list_en_final_clean_september_2025.pdf) names fonturacetam (4-phenylpiracetam, carphedon) as a non-specified stimulant, alongside modafinil. It is prohibited in competition. Whether the asthenia studies are positive is a separate question from its doping status.

## Cortexin and Pinealon

These two are often sold alongside the drugs above. Cortexin is a peptide extract from animal brain cortex, a mixture rather than a single molecule, so a result from one batch does not automatically carry over to another; its stroke studies are covered on [Russian pharmacology](russian-pharmacology.html) and [Cerebrolysin](cerebrolysin-porcine-neurotrophic-peptide-complex.html). Pinealon (Glu-Asp-Arg) is a Khavinson bioregulator peptide with cell, rat and computer-model studies but no human trial under that name; see [Khavinson peptide bioregulation](peptide-bioregulation-khavinson.html).

## Claims versus evidence

Claim| What the evidence shows| Verdict  
---|---|---  
Mexidol improves recovery after ischemic stroke| Two placebo-controlled trials (150 and 304 patients) report less disability. Same journal and research network; the larger trial is industry-sponsored; the main effect is 0.44 points on the mRS.| Supported, not independently replicated  
Mexidol works best within six hours| Stated in a review. Neither trial tested it.| Unsupported  
Mexidol is a cognitive enhancer| No trial in healthy people. In MIR, cognition was a non-significant trend.| Unsupported  
Bemethyl boosts physical and mental performance| One heat-tolerance study and one PK study, neither with a reported sample size. No cognition trial.| Unsupported  
Phenylpiracetam improves cognition and fatigue| One placebo-controlled epilepsy trial without reported effect sizes; a meta-analysis of before-and-after changes with no stated placebo arm.| Weak  
R-phenylpiracetam is the active form| Mouse memory test only. Clinical studies use the racemic mix.| Animal only  
  
## Practical points

  * The stroke evidence for Mexidol used the pharmaceutical product, given intravenously and then by mouth under medical care. Emoxypine powders sold online are not the same preparation.
  * Phenylpiracetam is prohibited in competition. Athletes who are tested should treat it as a doping substance.
  * Bemethyl is not bromantane, and neither has trial data for cognition.
  * None of these drugs has FDA approval. Products sold online as research chemicals are not the pharmaceutical preparations used in the trials.

## Who paid, and how it got here

### How it got here

- The 2026 WADA list names fonturacetam (4-phenylpiracetam, carphedon) among non-specified stimulants. It is prohibited in competition. ([WADA 2026 Prohibited List](https://www.wada-ama.org/sites/default/files/2025-09/2026list_en_final_clean_september_2025.pdf))
- The FDA drug-approval database has no application for emoxypine, bemethyl or phenylpiracetam under any of their names. ([openFDA Drugs@FDA](https://open.fda.gov/apis/drug/drugsfda/))

### Who paid for the key studies

- MIR Mexidol stroke trial; n=304: Less disability on Mexidol; 0.44-point mean mRS difference. Funding and interests: Pharmasoft, the lead industry sponsor on the trial registration. The journal abstracts do not name it. ([Trial registry](https://clinicaltrials.gov/study/NCT06437626))
- EPICA Mexidol stroke trial; n=150: Lower disability and stroke-severity scores than placebo. Funding and interests: No funding statement in the PubMed record. ([PubMed funding record](https://pubmed.ncbi.nlm.nih.gov/28665371/))
- Fonturacetam (Actitropil) asthenia meta-analysis; 549 patients: Fatigue scores fell 16.3 points after a month; no placebo comparison reported. Funding and interests: No funding statement in the PubMed record. ([PubMed funding record](https://pubmed.ncbi.nlm.nih.gov/40047835/))

Checked 2026-11-02.

## Sources

1. PubMed 35992374 (PubMed 35992374)
2. PubMed 21870773 (PubMed 21870773)
3. PubMed 21689376 (PubMed 21689376)
4. PubMed 28665371 (PubMed 28665371)
5. PubMed 33016677 (PubMed 33016677)
6. PubMed 40898634 (PubMed 40898634)
7. PubMed 41456191 (PubMed 41456191)
8. NCT06437626 (NCT06437626)
9. PubMed 32678558 (PubMed 32678558)
10. PubMed 33449535 (PubMed 33449535)
11. PubMed 9162292 (PubMed 9162292)
12. PubMed 16404942 (PubMed 16404942)
13. PubMed 34445727 (PubMed 34445727)
14. PubMed 25591651 (PubMed 25591651)
15. PubMed 16447562 (PubMed 16447562)
16. PubMed 25403298 (PubMed 25403298)
17. PubMed 40047835 (PubMed 40047835)
18. PubMed 17523446 (PubMed 17523446)
19. PubMed 32279140 (PubMed 32279140)