A synthetic seven-amino-acid peptide (sequence MEHFPGP) developed in Russia in the 1980s as a brain-protective and memory-enhancing drug. It is given as nasal drops and is officially approved in Russia for stroke and cognitive disorders — but is not approved by the FDA, EMA, or any Western regulator. Its main known effect is raising levels of BDNF (a key brain-growth protein) — the same protein that exercise and antidepressants increase. Nearly all human evidence comes from Russian-language studies that have never been independently replicated in the West.
Intranasal heptapeptide; a Pro-Gly-Pro–stabilized analog of the non-steroidogenic ACTH(4-10) fragment. Mechanistically: rapid nose-to-brain transport, robust BDNF/NGF and TrkB upregulation (Dolotov 2006), serotonergic enhancement plus dopaminergic sensitization (Eremin 2005), and multi-target anti-inflammatory neuroprotection. Russian-registered for acute ischemic stroke (1% formulation, 12–18 mg/day) and cognitive/nootropic use (0.1% formulation). The evidence base is biologically credible but methodologically limited — predominantly unblinded, single-center, Russian-language; a single Western placebo-controlled fMRI study (Lebedeva 2018, n=24) is the strongest independently accessible human signal.
Semax (Ac-none-Met-Glu-His-Phe-Pro-Gly-Pro-OH) is the ACTH(4-7) melanocortin core (MEHF) fused to a C-terminal PGP tripeptide that acts as a proteolytic shield and a pharmacologically active metabolite in its own right. CAS 80714-61-0; C₃₇H₅₁N₉O₁₀S; MW ≈ 751.9 g/mol (free acid). The duration paradox — minutes-long plasma half-life yet 20–24 h central effect — is explained by active PGP metabolite generation plus a CREB-driven neurotrophin transcriptional cascade. Developed at the Institute of Molecular Genetics, RAS (Myasoedov); manufactured by Peptogen. N-acetyl Semax amidate (NA-SA; CAS 2920938-90-3) adds N-acetylation + C-terminal amidation for ~340% greater fluid stability — popular in the Western gray market but with scarce direct human pharmacology.
The most-cited numbers across four decades of predominantly Russian research — the flagship stroke indication, the reproducible BDNF mechanism, and the honest evidence-quality caveats. Semax is the rare peptide with genuine regulatory approval in one jurisdiction and a near-total absence of independently replicated Western trials. Read every figure alongside its evidence grade.
🧠
Hippocampal BDNF · single dose · rat
↑1.4×
Dolotov et al. (2006, Brain Research, PMID 16996037): a single 50 µg/kg intranasal dose raised hippocampal BDNF protein ~1.4-fold (peak 3 h), TrkB phosphorylation ~1.6-fold, BDNF exon III mRNA ~3-fold, and TrkB mRNA ~2-fold. The most-replicated molecular signature of Semax and the mechanistic anchor for every cognitive claim.
🩸
Infarct volume reduction · MCAO models
~62%
In rat middle-cerebral-artery occlusion: pre-treatment reduced infarct volume ~62%, treatment at reperfusion ~48%, and delayed (3 h) treatment ~31%; a broader literature range of 25–50% is consistently reported. Strong, reproducible preclinical neuroprotection — the rationale for the Russian acute-stroke indication.
⏱️
Central effect duration · the paradox
20–24 h
Plasma half-life is measured in minutes, yet behavioral and neurochemical effects persist 20–24 h after one intranasal dose. Resolved by two mechanisms: the active metabolite Pro-Gly-Pro (independently neurotrophin-inducing) and a CREB-driven transcriptional cascade that outlasts the parent compound.
🧬
Default mode network · human fMRI
DMN↑
Lebedeva et al. (2018, PMID 30225715): randomized, placebo-controlled resting-state fMRI in healthy volunteers (n=24, 1% Semax 1.2 mg) showed greater volume of the DMN rostral subcomponent (medial frontal cortex) vs placebo at 5 and 20 min. The strongest independently accessible placebo-controlled human signal.
😌
Serotonin metabolite · extracellular
+180%
Eremin et al. (2005, Neurochem Int, PMID 16362768): Semax raised striatal 5-HIAA +25% in tissue and up to +180% extracellularly within 1–4 h, and amplified d-amphetamine-induced dopamine release without altering basal dopamine — a serotonin-enhancing, dopamine-sensitizing profile distinct from classical stimulants.
🏛️
Russian regulatory status
Approved
Registered pharmaceutical in the Russian Federation; placed on the List of Vital and Essential Drugs (ЖНВЛП) on December 7, 2011. Labeled for ischemic stroke, TIA, dyscirculatory encephalopathy, cognitive disorders, and optic nerve atrophy; manufactured by Peptogen under Russian GMP. Also registered in Ukraine.
⚖️
US regulatory trajectory · 2026
PCAC
Not FDA-approved. Removed from the FDA Category 2 interim 503A list on April 22, 2026 — which clears Semax (free base and acetate) for Pharmacy Compounding Advisory Committee review on July 24, 2026, a potential pathway toward lawful US compounding. Removal does not itself confer Category 1 status.
🔬
Independent Western replication
None
No non-Russian RCT replication exists; no ClinicalTrials.gov registration; no Cochrane review. The ADDF Cognitive Vitality program concluded well-conducted studies are lacking and there is no evidence for Alzheimer's disease. GRADE-equivalent rating: stroke evidence low/very-low; cognitive evidence very-low. Mechanism is credible; clinical proof is not.
02 · Mechanism of action
How a neurotrophin modulator works.
Semax is built from the middle piece of a natural hormone called ACTH — but the part that controls stress hormones is deliberately left out, so Semax does not raise cortisol or affect your adrenal glands. What it does do is switch on the brain's own growth and repair proteins, mainly BDNF and NGF — the same proteins linked to learning, memory, and recovery after brain injury. It also nudges up serotonin (mood) and "primes" the dopamine system (focus and motivation) without acting like a stimulant. Because it travels straight from the nose to the brain, effects can begin within minutes and last most of a day.
Semax engages MC4R as its initiating receptor interaction, then exerts its dominant effect indirectly: CREB-mediated transcriptional upregulation of BDNF/NGF and their receptors (TrkB/TrkA), driving MAPK/ERK and PI3K/Akt survival and plasticity signaling. Layered on top are serotonergic enhancement, dopaminergic sensitization, enkephalinase inhibition (via the PGP component), and a broad anti-inflammatory/antioxidant neuroprotective program. Critically, the molecule retains MC4R affinity but lacks MC2R/adrenal activity — confirming "non-hormonal" status with no HPA-axis or glucocorticoid effect at pharmacological doses.
Semax is best classified as a neurotrophin modulator with secondary psychostimulant-like properties — distinct from amphetamines (no direct catecholamine release, no adrenergic stress), cholinergic nootropics, and racetams (not glutamate-receptor based). Genome-wide microarray work (Medvedeva, Myasoedov et al., BMC Genomics 2014, PMID 24661604; GEO GSE163654) shows Semax modulates hundreds of genes in ischemized cortex, enhancing immune-response and vascular/angiogenic gene programs at 24 h post-occlusion. The closest mechanistic analogy is exercise-induced or antidepressant-associated BDNF elevation. Note: the molecular detail is overwhelmingly preclinical; human mechanistic confirmation is limited to fMRI surrogate endpoints and plasma-BDNF correlations in stroke cohorts.
The signature effect: a single 50 µg/kg intranasal dose raised hippocampal BDNF ~1.4-fold (peak 3 h), TrkB phosphorylation ~1.6-fold, BDNF exon III mRNA ~3-fold (Dolotov 2006). BDNF and NGF are also elevated in frontal cortex and retina at 8 h (Shadrina 2010). The BDNF–TrkB axis underlies long-term potentiation, hippocampal neurogenesis, and memory consolidation — the coherent rationale for cognitive effects.
Clinical significance: In post-stroke patients, Semax raised plasma BDNF levels that stayed elevated through the study and correlated positively with Barthel-index (functional independence) improvement (Gusev 2018). This is the closest the human literature comes to confirming the preclinical mechanism — a plasma-biomarker proxy rather than direct CNS measurement. The temporally staged neurotrophin induction (Bdnf/TrkC/TrkA at 3 h; Nt-3/Ngf at 24–72 h post-occlusion) maps onto the proposed window for post-stroke recovery support.
Molecular detail: The induction is indirect — most evidence supports CREB-mediated transcriptional upregulation of neurotrophin genes rather than direct Trk receptor binding by Semax. The major metabolite Pro-Gly-Pro independently activates Bdnf, Ngf, and TrkC transcription in ischemic cortex (PMC 11498467), so the in-vivo effect is effectively Semax + PGP pharmacology combined. Direct quantification of neurogenesis via cell-birth markers has not been performed in accessible Semax-specific publications — neurogenesis is inferred from pathway activation and behavioral normalization.
P
🎯
MC4R · initiating receptor · non-hormonal status
Semax engages melanocortin-4 receptor (MC4R), widely expressed in hypothalamus, brainstem, and limbic system — considered the primary initiating interaction upstream of the downstream neurotrophin and monoaminergic effects. MC4R agonism links to neuroprotection, energy homeostasis, and monoaminergic tone. The functional absence of adrenal/corticotropic effects despite MC engagement confirms selectivity away from MC2R.
Clinical significance: This is the key safety distinction from full-length ACTH. Because Semax spans only residues 4–10 (plus PGP) and lacks the residues 1–3 required for steroidogenesis, it is classified in Russian pharmacology as a "non-hormonal" neuropeptide: no HPA-axis disruption, no adrenal suppression, no glucocorticoid effects at pharmacological doses. Clinicians can use it without the endocrine monitoring that exogenous corticotropins demand.
Molecular detail: The ACTH(4-7) core (MEHF) contains part of the melanocortin-binding motif, though the canonical His-Phe-Arg-Trp pharmacophore's Arg-Trp residues (positions 8–9) are absent in Semax, giving reduced but measurable MC-receptor affinity. Whether MC4R agonism is a true high-affinity event or a low-affinity trigger that is amplified by the downstream transcriptional cascade remains incompletely resolved. Direct radioligand binding constants for Semax at MC4R are not well characterized in accessible literature.
Eremin et al. (2005, PMID 16362768): Semax elevated striatal 5-HIAA +25% in tissue and up to +180% extracellularly within 1–4 h (robust serotonergic stimulation). It did not change basal dopamine, but pre-treatment dramatically amplified d-amphetamine-induced dopamine release and locomotion — a priming/potentiating effect, not direct release.
Clinical significance: The serotonin-enhancement-plus-dopamine-sensitization profile partially explains the reported nootropic, mood-elevating, and attentional effects in humans, while producing a pharmacology distinct from classical stimulants. The dopamine-priming effect also carries an interaction caution: co-administered stimulants (amphetamines, high-dose caffeine) could be amplified unpredictably, and MAOIs raise a theoretical serotonin-syndrome concern.
Molecular detail: The fMRI functional-connectivity study of Semax + Selank (PMID 32342318, n=52) localized Semax's effect to dorsolateral prefrontal cortex (working memory) connectivity and Selank's to amygdala (anxiety) connectivity — consistent with the distinct clinical profiles and the proposed monoaminergic substrate. The dopamine-sensitization phenomenon (potentiation without basal change) is mechanistically reminiscent of priming effects seen with certain neuropeptides on mesolimbic reactivity, though the precise locus (presynaptic vesicular vs receptor-level) is uncharacterized.
P
🔗
Enkephalin-degradation inhibition · PGP component
Morozov et al. (2001, PMID 11443939): both Semax and Selank dose-dependently inhibit enkephalin-degrading enzymes from human serum, with Semax IC₅₀ ≈ 10 µM — far more potent than bacitracin (IC₅₀ ≈ 10 mM). By prolonging endogenous enkephalin action, this contributes anxiolytic, analgesic, and mood effects beyond direct receptor mechanisms. The PGP tripeptide (also a metabolite) drives this activity.
Clinical significance: This endogenous-opioid-modulating mechanism is the theoretical basis for the lower-dose anxiolytic effects users report, and for speculative interest in addiction/opioid-use biology. It is a secondary, supportive mechanism — not a primary therapeutic target — and there are no clinical addiction trials. It does, however, complicate the otherwise "non-opioid" framing and is worth noting in patients with opioid-use history.
Molecular detail: The shared PGP C-terminus of Semax and Selank is the structural commonality underlying their overlapping enkephalinase inhibition, even though their parent molecules (ACTH(4-10) vs tuftsin) and primary actions differ. Whether the inhibition is competitive at the peptidase active site or allosteric is not detailed in accessible literature; the micromolar IC₅₀ suggests it is pharmacologically relevant at central concentrations achieved by nose-to-brain delivery but likely minor at systemic exposures.
P
🛡️
Anti-inflammatory · antioxidant · anti-apoptotic
Multi-target neuroprotection: in stroke patients Semax shifted the cytokine balance toward anti-inflammatory (↑IL-10, TNF-α suppression, ↓IL-8, ↓CRP). In rat tMCAO it suppressed Il1a/Il1b/Il6/Ccl3/Cxcl2 mRNA; reduced post-ischemic nitric oxide; cut lipid peroxidation (MDA) ~30–40%; and preserved mitochondrial integrity in penumbral tissue.
Clinical significance: The cytokine-shift data (PMID 10358912) is a genuine small human mechanistic study (Grade C) providing biological context for the clinical stroke benefit, where limiting secondary neuroinflammatory injury in the penumbra is the therapeutic goal. The antioxidant and anti-apoptotic effects support the rationale for early administration in acute ischemia, when oxidative and nitrosative damage peaks.
Molecular detail: The inflammatory-gene suppression (Il1a/Il1b/Il6/Ccl3/Cxcl2, PMID 34097675) and the MDA reduction in spleen/thymus stress models indicate the anti-inflammatory program is not CNS-exclusive. NO regulation limits nitrosative neuronal damage; mitochondrial-membrane preservation in penumbra promotes neuronal survival signaling. These are mechanistically distinct from the neurotrophin axis but likely converge on the same net outcome — reduced infarct volume — explaining the additive effect sizes seen with combined pre-/peri-/post-ischemic dosing.
P
🧬
Genome-wide transcriptome & copper/Aβ effects
Genome-wide microarray in rat focal ischemia (BMC Genomics 2014, PMID 24661604) showed Semax affects hundreds of genes — modulating immune-cell genes at 3 h and profoundly enhancing immune-response and vascular/angiogenic gene programs at 24 h. A 2024 follow-up (PMID 39066823) confirmed ACTH-like peptides compensate the ischemia-disrupted gene-expression profile a day after stroke.
Clinical significance: The transcriptome and vascular/angiogenic findings broaden the neuroprotection story beyond single-pathway BDNF — suggesting Semax helps restore a coordinated recovery gene program rather than acting on one target. This systems-level effect is the strongest argument for the "compensation" framing of its post-stroke benefit, though it remains entirely preclinical.
Molecular detail: A rare non-Russian study (ACS Chemical Neuroscience 2022, PMID 35099917) showed Semax sequesters Cu²⁺ and inhibits Aβ:Cu²⁺-mediated fibrillogenesis in artificial membrane models — an Alzheimer's-relevant mechanism. 2025 transgenic-mouse work (PMID 41479572) showed Semax + a derivative improved cognition and reduced amyloid burden in APPswe/PS1dE9 mice. Both are mechanistically suggestive but have no human AD validation; the ADDF explicitly found no evidence for Alzheimer's disease.
Semax is administered as intranasal drops in two pharmaceutical strengths with distinct labeled indications: the 0.1% solution for outpatient cognitive/nootropic use and the 1% solution for acute, supervised neurological use. This is the engine core. Each protocol is built to the same skeleton — strength, dosing, course length, timing, administration, and explicit evidence grade — anchored by a pharmacokinetic panel (the 2–5 min / 12–24 h half-life paradox), six route/indication protocols, a course-ladder visual, a reconstitution calculator, a global dose-band table, a weight-band reference, engine-ready titration logic, and a biomarker monitoring scaffold. Dosing figures reflect Russian pharmaceutical labeling and published trials (the acute-stroke and post-stroke data are Grade B); off-label, nootropic, and injectable use is a speculative hypothesis layer (Grade C/D) unvalidated in Western trials.
Important · regulatory status
Semax is not approved by the FDA, EMA, MHRA, Health Canada, or the TGA for any indication, and is not a dietary supplement. The dosing below reflects Russian pharmaceutical labeling (Peptogen, registered medicine) and published Russian trials. In the US it is sold as a "research chemical for laboratory use only"; human use is unapproved. Injectable or compounded Semax for human use is unapproved in the US, EU, and most Western countries and carries risks of contamination, microbial infection, endotoxemia, and unverified purity. Figures are educational, not a prescription.
2026 US regulatory development
On April 22, 2026 the FDA removed Semax (free base and acetate) from the Category 2 interim 503A list. Rather than closing the door, this clears Semax for Pharmacy Compounding Advisory Committee (PCAC) review scheduled for July 24, 2026, alongside Epitalon and Emideltide/DSIP — a potential, though not guaranteed, pathway toward lawful 503A compounding if the committee recommends Category 1 inclusion. Clinicians considering future compounded Semax should monitor the PCAC outcome; removal from Category 2 does not by itself authorize compounding.
Pharmacokinetics · the half-life paradox
Why a 2–5 minute peptide acts for a day.
Semax has one of the most striking pharmacokinetic dissociations in this atlas: the intact heptapeptide is cleared from plasma in 2–5 minutes, yet its downstream effects — BDNF/NGF transcription and enkephalinase inhibition — persist for 12–24 hours. The reconciling mechanism is twofold. First, the C-terminal Pro-Gly-Pro (PGP) tail is cleaved off to form an active metabolite that itself drives neurotrophin transcription and clears far more slowly in the CNS. Second, direct nose-to-brain transport via olfactory and trigeminal pathways delivers peptide to brain tissue while bypassing the systemic compartment where degradation is fastest. Oral bioavailability is effectively zero — gastric and intestinal peptidases destroy it — which is why every route in the atlas is intranasal or parenteral. Human plasma Cmax/Tmax were never formally published in peer-reviewed Western literature and are inferred from rodent kinetics (Grade D).
Parameter
Value
Note
Primary route
Intranasal (registered)
Russian 0.1% / 1% nasal drops. SC injection is research practice; endonasal electrophoresis used for optic-nerve protocols.
Plasma t½ (parent)
2–5 minutes
Intact heptapeptide cleared by carboxy/aminopeptidases almost immediately. This is not the duration of effect.
Active metabolite
Pro-Gly-Pro (PGP)
Cleaved from C-terminus; independently activates neurotrophin transcription; accumulates in brain with a substantially longer effective window.
Effective duration
12–24 hours
Via BDNF/NGF mRNA induction (peak 3–8 h, resolving by 24 h) and enkephalinase inhibition — the basis for once/twice-daily dosing.
Nose-to-brain transport
Olfactory + trigeminal
Direct CNS delivery bypassing systemic clearance; CSF concentrations reach ~60–70% of plasma in rodents. Grade C (animal).
Oral bioavailability
≈ 0%
Destroyed by gastric/intestinal peptidases — no oral form exists. Confirms parenteral/intranasal-only routes.
CNS binding
KD = 2.4 ± 1.0 nM
Specific Ca²⁺-dependent binding sites in basal forebrain (Bmax 33.5 fmol/mg); regionally selective (hippocampus/forebrain, not cerebellum).
Cmax / Tmax (human)
Not formally published
Western peer-reviewed PK absent; inferred from rodent kinetics. Grade D.
Clearance
Carboxypeptidase (parent); slower for PGP
Rapid plasma clearance of parent; PGP metabolite cleared more slowly, accounting for prolonged effect.
Renal / hepatic adjustment
Not established
No human PK studies in organ impairment; peptide nature suggests minimal hepatic load but unconfirmed. Grade D.
Architecture note: the parent t½ and the effect duration are different fields and should never be conflated in a dosing engine. Frequency is driven by the downstream effect window (12–24 h), not the plasma half-life (minutes) — the opposite of how a small-molecule schedule would be built.
1 mg/mL (0.1%). At ~50 µL per dropper drop, each drop delivers approximately 50 µg (0.05 mg).
Standard dosing
1–2 drops in each nostril, 2–3 times per day. Total daily dose approximately 200–600 µg/day (0.2–0.6 mg/day). Mirrors the ADDF-cited regimen ("1–2 drops of 0.1% solution twice daily").
Course length
Typically 10–14 days; may be repeated after an interval. Labeled in Russia for memory/cognitive disorders, dyscirculatory encephalopathy, attention, stress, and healthy cognitive optimization.
Timing
Earlier in the day is preferable — transient insomnia is reported (~3%) when dosed late, consistent with the activating monoaminergic profile. Avoid evening dosing in sleep-sensitive patients.
Administration
Head upright or slightly back; instill drops into each nostril; brief sniff. Do not exceed the labeled per-dose drop count. Single-naris alternation is sometimes used to improve mucosal contact time.
Onset / duration
Central onset within ~15–30 min (consistent with fMRI changes at 5–20 min); single-dose central effects reported up to 20–24 h in animal models — many users dose once or twice daily rather than continuously.
Patient selection
Best characterized as adjunctive cognitive support, not a validated treatment. Screen for the contraindications in Section 05 (acute psychosis, uncontrolled hypertension, hormone-sensitive malignancy, pregnancy/lactation). No HPA-axis monitoring required given non-hormonal status.
Dose-response evidence
A Russian-language study cited secondarily (cannot be fully verified in English) reported that Semax 250–1,000 µg/kg improved attention and short-term memory with EEG changes resembling other neuroprotective drugs in healthy subjects — flagged by ADDF reviewers as "cannot access." Treat dose-response specifics as provisional.
Semax 1% · Acute Ischemic Stroke / Severe Protocol
10 mg/mL (1%). At ~50 µL per drop, each drop delivers approximately 500 µg (0.5 mg) — ten times the 0.1% strength.
Standard dosing
2–4 drops per nostril, 3–4 times per day. Total daily dose approximately 12–18 mg/day, matching the Gusev 1997 trial dosing and the ADDF-cited regimen.
Severity-tiered target
Most effective doses in the Gusev controlled trial: 12 mg/day for moderate-severity stroke and 18 mg/day for severe stroke (courses of 5 and 10 days respectively).
Timing window
Russian protocols and a 120-patient trial report administration begun within ~6–12 h of symptom onset, added on top of standard intensive stroke therapy — not as monotherapy. Earliest feasible administration aligns with the preclinical pretreatment > reperfusion > delayed efficacy gradient.
Course length
5–10 days under medical supervision. The 1% preparation is explicitly a clinical-grade formulation for acute neurological use — not for outpatient self-administration.
Expected effect
Add-on Semax increased the rate of neurological-function restoration, with particular improvement in motor disorders; meta-analysis found significant NIHSS reduction and Rivermead Mobility Index improvement (see Section 06).
Integration with standard care
Always adjunctive to guideline acute-stroke management (thrombolysis/thrombectomy eligibility, BP control, antiplatelet/anticoagulation per protocol). Semax is a putative neuroprotectant, not a reperfusion therapy, and must never delay or substitute for time-critical recanalization.
Mechanistic timing rationale
The temporally staged neurotrophin induction (Bdnf/TrkC/TrkA at 3 h; Nt-3/Ngf at 24–72 h) and the 3 h-vs-24 h transcriptome shifts support a multi-day course rather than a single dose, and provide a biological rationale for both acute-phase and early-rehabilitation administration.
⚠ Caveat · evidence quality
The defining stroke trials are non-randomized or open-label, single-center, Russian-language, with small Semax arms. Even the Russian meta-analysis authors called for a "multicenter, powerful double-blind study." A frequently-cited Stroke 2007 RCT (n=184, claimed 34% NIHSS benefit, PMID 17395875) could not be independently confirmed against the original article — verify before relying on those effect sizes.
Gusev 2018 (PMID 29798983): 2 courses of 6,000 µg/day (6 mg/day) for 10 days each, with a 20-day interval between courses — a lower-dose maintenance approach versus the acute high-dose protocol.
Rehabilitation timing
Early rehabilitation combined with Semax showed synergistic benefit; the study compared early vs late rehabilitation timing and favored early integration. Pair courses with structured physiotherapy/occupational therapy.
Biomarker correlate
Plasma BDNF rose and stayed elevated throughout the study; high BDNF accelerated Barthel-index improvement, and BDNF correlated positively with Barthel score — a rationale for using BDNF as an exploratory response biomarker where assayable.
Population
Study cohort n=110 (43 men, 67 women; mean age 58). Open-label, no placebo arm, Russian-language — Grade C. Treat the regimen as a documented practice pattern, not a proven maintenance standard.
Practical structuring
A clinician building a profile might frame: acute course (1%, 12–18 mg/day × 5–10 d) → 20-day interval → maintenance courses (6 mg/day × 10 d ×2). Reassess function (NIHSS, Barthel, Rivermead) at each course boundary; discontinue if no measurable functional trajectory.
Open question
No data define the optimal number of maintenance courses, the durability of BDNF elevation between courses, or whether the BDNF–Barthel correlation is causal or merely associative. The neonatal-rat finding of lasting behavioral change after early Semax raises an unresolved question about long-term neuroplastic remodeling with repeated courses.
Polunin et al. (2000, PMID 10741256): Semax tested in vascular, toxic, allergic, and inflammatory optic nerve disease including partial optic nerve atrophy, added to standard neurotrophic + anti-inflammatory therapy.
Delivery routes studied
Two active arms: (1) intranasal nasal drops; (2) Semax by endonasal electrophoresis. Both were compared against a control receiving basic therapy only.
Expected effect
Improved visual-function recovery rate and quality: visual acuity, total visual field, electric sensitivity and conductivity of the optic nerve, and color vision — described as particularly effective in the acute stage.
Concentration
Russian labeling lists optic nerve conditions under the lower-severity 0.1% formulation indications; severe acute presentations may use the 1% strength under specialist supervision.
Positioning
Strictly adjunctive to ophthalmologic standard of care; not a substitute for treating the underlying vascular/inflammatory etiology. Sample size was unspecified in the accessible English abstract and there was no described blinding — Grade C at best.
Mechanistic plausibility
Retinal BDNF/NGF elevation at 8 h (Shadrina 2010) provides a neurotrophic rationale for optic-nerve and retinal-ganglion-cell support, consistent with the broader neuroprotection program — but this remains a small, single-center signal.
Off-Label / Research-Grade Community Practice
Anecdotal · unvalidated · plain Semax vs NA-SA · cycling
Grade D
Typical intranasal dosing
Research-grade intranasal Semax is commonly used in the 300–1,000 µg/day range, often split across 2 administrations — broadly comparable to the 0.1% labeled cognitive dose per day. Some reports cite up to 3,000 µg/day.
NA-SA (N-acetyl Semax amidate)
Community preference for NA-SA rests on purported greater per-µg potency and longer action from enhanced stability — so users often dose it lower than plain Semax. The potency claim (2–5× per µg) is inferred, not demonstrated in controlled human trials.
Injectable (unapproved)
Community subcutaneous reports range 100–500 µg/day. No approved injectable exists anywhere; this is unapproved compounded use with unknown sterility — explicitly discouraged.
Cycling
Empirical cycles of 10–30 days on with off-periods (e.g., 10 on / 10–20 off) to assess effect and pre-empt theoretical adaptation. There is no clinical tolerance data — cycling is a community heuristic, not evidence-based.
Quality risk
Gray-market product carries real risks: mislabeling (plain vs NA-SA), unverifiable HPLC certificates, no pharmaceutical sterility, and possible endotoxin/contamination. Identity and purity are not guaranteed.
Clinician framing
If a patient discloses gray-market Semax use, the harm-reduction priorities are: confirm intranasal (not injectable) route, screen contraindications (Section 05), check for stimulant/MAOI co-use, and counsel that efficacy and long-term safety are unestablished. Do not endorse injectable use.
NA-SA pharmacology gap
NA-SA mechanism-of-action is largely inferred from plain-Semax pharmacodynamics; direct human studies are scarce (Grade P/D). A January 2026 industry-cited stability study (340% longer fluid stability; 6–8 h BDNF elevation) needs independent verification.
⚠ Not validated
All figures in this tab are anecdotal community practice from forums and vendor reports, not validated in controlled trials. There is no long-term human safety dataset for healthy-user dosing of either plain Semax or NA-SA. This information is harm-reduction context, not an endorsement.
For laboratory reference · sterility not guaranteed in gray market
Grade D
Reconstitution medium
Bacteriostatic water or sterile saline to the desired concentration. Standard research concentration is 1–5 mg/mL. For intranasal-equivalent strengths, 1 mg/mL approximates the 0.1% formulation and 10 mg/mL the 1%.
Appearance
Lyophilized peptide is a white to off-white powder; the reconstituted solution is aqueous and clear. Discard if cloudy, discolored, or particulate.
Storage
Lyophilized: stable at −20 °C. Reconstituted: 2–8 °C, used within ~14–30 days. Avoid repeated freeze-thaw cycles — they degrade activity.
pH / stability
Most stable at slightly acidic-to-neutral pH (5–7). Primary degradation is by serine and metallopeptidases, yielding PGP and smaller fragments — relevant to potency loss in poorly stored solution.
Quality verification
Where available, verify vendor HPLC purity certificates (≥99% claimed) and request endotoxin testing for any preparation intended for biological assay. Certificate authenticity cannot always be confirmed.
NA-SA stability note
NA-SA's N-acetylation and C-terminal amidation are reported to maintain structural integrity >7 h in reconstituted form vs ~90 min for unmodified Semax — the basis for its handling advantage, though this is vendor/industry-cited rather than independently confirmed.
Use the calculator
The reconstitution calculator below converts a chosen vial mass and diluent volume into solution concentration, per-drop dose (at ~50 µL/drop intranasal), and the number of labeled-equivalent daily doses per vial. It is a reference tool, not a clinical authorization.
Acute stroke → maintenance course structure
Course model: from acute neuroprotection to rehabilitation.
Hr 0–121%InitiateAdd-on to standard care · ASAP post-onset
Day 1–1012–18 mgAcute course12 mg moderate · 18 mg severe · 5–10 d
Day 11–3020-dayIntervalOff-drug interval · begin early rehab
ReassessNIHSSBarthel · RivermeadContinue only on measurable trajectory
L2 · Reconstitution & intranasal dose math
Reconstitution & Intranasal Dose Calculator
For laboratory reference only. The Russian pharmaceutical product is supplied as pre-made 0.1% and 1% nasal drops — no reconstitution needed. This tool covers research-grade lyophilized peptide and assumes a standard ~50 µL intranasal dropper drop. It is not a clinical authorization; human use of research-grade Semax is unapproved in the US.
Concentration
—
Per drop
—
Drops per dose
—
Doses per vial
—
Strength class
—
Worked reference: a 5 mg vial + 2 mL diluent = 2.5 mg/mL ≈ 0.1% strength, delivering ~125 µg per 50 µL drop, so a 250 µg nootropic dose ≈ 2 drops and the vial yields ~20 doses. A 10 mg vial + 1 mL = 10 mg/mL ≈ 1% strength (~500 µg/drop) reproduces the acute-stroke preparation. For SC research dosing the dossier reference is 10 mg + 3.0 mL = 3.33 mg/mL, where 1 unit on a U-100 syringe ≈ 33.3 µg, so a 500 µg dose ≈ 15 units (0.15 mL). The Russian pharmaceutical product ships pre-made and needs no reconstitution; this tool is for research-grade lyophilate only.
Global dose bands · labeled use + research practice
Three dose tiers & weight-band reference.
Semax dosing spans a wide range — from sub-milligram nootropic micro-doses to 12–18 mg/day acute-stroke regimens, a roughly 30–70× spread. The engine anchors three bands: a Low cognitive band drawn from the Russian 0.1% label, a Standard general-neuroprotection band, and a High band reflecting the supervised acute-stroke and post-stroke maintenance protocols. The clinical-outcome data behind the High band (acute stroke, n=110–187) are Grade B; the Low/Standard nootropic figures rest on practice-pattern and Russian-label use (Grade C/D). Semax is dosed by flat indication-based amount, not body weight — the weight table below is per-kg framing only and is not a validated weight-adjusted schedule.
Band
Dose / day (intranasal)
Primary use
Basis
Grade
Low
250–600 µg/day
Cognitive / nootropic; attention & working memory
Russian 0.1% label (1–2 drops 2–3×/day).
C
Standard
600–1,000 µg/day
General neuroprotection; off-label research community
The High band is a clinical-grade, medically-supervised regimen using the 1% formulation — not an outpatient self-administration target. There is no validated nootropic dose above ~1,000 µg/day; the gap between the Standard and High bands reflects an indication change (cognition → acute neuroprotection), not a continuous escalation path.
Weight-band reference (per-kg context only — flat dosing applies)
Interpolated on a ~7 µg/kg/day basis for the Standard band; Low and High columns reproduce the fixed labeled ranges for orientation. The rodent mechanistic dose (50 µg/kg) does not translate to humans by simple allometric scaling, so these values are calculator scaffolding, not dosing instructions (Grade D).
Body weight
Low band (µg/day)
Standard band (µg/day)
High band (µg/day)
55 kg (121 lb)
250
~385
≤ 12,000–18,000*
65 kg (143 lb)
300
~455
≤ 12,000–18,000*
75 kg (165 lb)
350
~525
≤ 12,000–18,000*
85 kg (187 lb)
400
~595
≤ 12,000–18,000*
95 kg (209 lb)
450
~665
≤ 12,000–18,000*
105 kg (231 lb)
500
~735
≤ 12,000–18,000*
*The High band is not weight-scaled — it is a fixed severity-tiered stroke regimen (12 mg moderate, 18 mg severe) given under supervision regardless of body mass. Per-kg values for the Standard band are research-comparison framing only and are not validated for weight-adjusted dosing of Semax.
Titration logic · engine-ready decision rules
Escalation, hold & stop logic.
Unlike the incretin peptides, Semax is not titrated by slow upward steps to a maintenance ceiling — it is dosed in fixed courses (typically 10–14 days) at an indication-appropriate strength, then cycled with a washout. The decision rules below are extrapolated from Russian clinical protocols and research-community practice (Grade D unless a clinical endpoint anchors them). The most important nodes are the hard stops tied to its activating monoaminergic/melanocortin profile: psychiatric destabilization and cardiovascular stress.
Permanent stop (absolute) → discontinue; do not initiate.
No safety data; neuropeptide effects on fetal/neonatal development unknown.
D
Special populations: avoid in anyone under 18 (no pediatric safety data) and in active CNS malignancy (BDNF/NGF upregulation could theoretically support TrkB/TrkA-driven tumor growth). Acute-stroke dosing is supervised-only and follows the fixed Gusev course rather than this outpatient logic.
Biomarker scaffold · validation status flagged
Response & safety monitoring bundles.
Semax has no validated clinical monitoring panel for nootropic use — almost every marker below is flagged validated_for_semax = false, the hallmark of an unapproved compound. The single partial exception is plasma BDNF, used as an exploratory response marker in the post-stroke cohort (where it correlated with Barthel-index recovery), and visual acuity / field in the optic-nerve indication, which carries Grade B support. The rest are class-borrowed safety surveillance checked on clinical indication only.
Biomarker
Frequency
Threshold / action
Validated?
Plasma BDNF (response)
Baseline; day 10; end of course (research only)
Rising/sustained BDNF correlated with Barthel-index recovery in post-stroke cohort (n=110)
Research only
Visual acuity / field (optic-nerve)
Baseline; day 7; day 14
Improvement in acuity, field, and optic-nerve conductivity
Functional endpoint; no Semax-specific validated battery exists
Not validated
Nasal mucosa inspection
Weekly
Erosion / bleeding / atrophy → hold intranasal route
Not validated
Cortisol (baseline)
Baseline (if HPA concern)
Melanocortin-axis context; no defined action threshold
Not validated
CBC / CMP
Baseline
General health screen; no Semax-specific signal
Not validated
Architecture note: store each marker with a source_endpoint tag and a validated_for_semax boolean. Only plasma BDNF (research) and the optic-nerve visual metrics (Grade B) resolve to anything other than false — which is exactly what marks Semax as a speculative-use entry rather than an approved-drug page.
04 · Combination protocols
Stacking Semax.
Semax's best-known combination is with Selank — its anxiolytic sibling from the same Russian institute — a pairing supported by a single human functional-connectivity study showing the two act on neurologically distinct regions. Cholinergic and other nootropic combinations are mechanistically defensible but evidence-free. Several combinations warrant genuine caution given the serotonergic and dopamine-sensitizing profile. Below: the mechanistically supported, the theoretical, and the avoid list.
Semax + Selank · the classic pairing
High Synergy
SemaxSelank
The most-discussed Semax stack. Mechanistically complementary: Semax provides dopamine sensitization, BDNF upregulation, and a nootropic/activating profile; Selank provides GABA-A modulation, anxiety reduction, and enkephalinase inhibition — nootropic enhancement plus anxiolytic stability without sedation. The PMID 32342318 functional-connectivity study (n=52) assessed both peptides by resting-state fMRI, finding Semax acts on DLPFC (working-memory) connectivity and Selank on amygdala (anxiety) connectivity — indirect human evidence that the two have neurologically distinguishable, complementary effects.
Component
Primary action
Evidence
Semax
MC4R → BDNF/TrkB · nootropic
fMRI DLPFC (C)
Selank
GABA-A · enkephalinase · anxiolytic
fMRI amygdala (C)
Semax + Cerebrolysin · neurotrophin loading
Moderate Synergy
SemaxCerebrolysin
A mechanistically complementary, not redundant, neuroprotection pairing. Cerebrolysin supplies direct neurotrophic-factor activity (BDNF/NGF mimicry) — it reduced caspase-3/7 activation and apoptosis in CoCl₂-stressed PC12 cells via GSK3β inhibition — whereas Semax adds transcriptional BDNF/NGF induction but produced no direct trophic effect on PC12 cells, indicating the two work through different arms of the same pathway. The theoretical appeal is additive neurotrophin elevation in stroke/vascular-dementia contexts. No combined human RCT exists (Grade C/D), and the combined TrkB/TrkA loading is an absolute contraindication in active CNS malignancy.
Component
Mechanism
Evidence
Semax
Transcriptional BDNF/NGF induction
Preclinical (P/C)
Cerebrolysin
Direct trophic mimicry (PC12 ↓apoptosis)
Approved (multi-country)
Semax + BPC-157 · central + systemic
Theoretical
SemaxBPC-157
A community neurorecovery concept pairing non-overlapping targets: BPC-157 acts peripherally on angiogenesis, the gut-brain axis, and vagal signalling, while Semax acts centrally on neurotrophin and monoamine systems — plausible complementarity rather than mechanistic overlap. The rationale is reasonable but entirely unproven: no human combination studies and no published safety data for the pairing exist (Grade D). BPC-157 itself is FDA Category-2 listed and unapproved; treat the combination as fully uncharacterized.
Component
Primary target
Evidence
Semax
Central · neurotrophin / monoamine
Russian clinical (B/C)
BPC-157
Peripheral · angiogenesis / gut-brain
Animal; limited human (C/P)
Semax + Cholinergics
Theoretical
SemaxAlpha-GPCCiticoline (CDP-choline)
A theoretically plausible learning/memory stack: Semax elevates BDNF (which supports cholinergic neuron survival and function) while cholinergics supply acetylcholine precursor. A BDNF–ACh synergism is biologically reasonable for memory consolidation. No published evidence exists for this specific combination — it is mechanistic speculation (Grade D), commonly assembled in the nootropic community but never tested in a trial.
Gusev 2018 reported that early rehabilitation combined with Semax produced synergistic functional benefit, with high plasma BDNF accelerating Barthel-index recovery. The biological rationale — BDNF-primed synaptic plasticity coinciding with task-driven rehabilitation during the early recovery window — is the most defensible "stack" in the entire Semax literature, pairing a pharmacological plasticity signal with the activity-dependent learning that consolidates it.
Component
Role
Evidence
Semax
BDNF-mediated plasticity priming
Open-label (C)
Early rehab
Activity-dependent consolidation
Standard of care (A)
⚠ Combine With Caution / Avoid
Caution
MAO inhibitorsAmphetamines / stimulantsHigh-dose caffeineOther BDNF-elevators (bipolar)Anticoagulants
Semax's serotonergic and dopamine-sensitizing pharmacology drives most interaction concerns. Because Semax amplifies amphetamine-induced dopamine release and locomotion in rodents, co-use with stimulants could increase their effects unpredictably. MAOIs raise a theoretical serotonin-syndrome concern given the +180% extracellular 5-HIAA effect. In bipolar disorder, stacking with other BDNF-elevating agents (high-dose lithium, certain antidepressants) is a theoretical manic-switch concern. Some sources note mild platelet modulation — a theoretical additive bleed risk with anticoagulants.
Combination
Concern
Action
MAO inhibitor
Serotonin syndrome (theoretical)
Avoid / specialist input
Stimulants
Unpredictable DA amplification
Caution · monitor
BDNF-elevators (bipolar)
Manic switching (theoretical)
Caution · psychiatry input
Anticoagulants
Additive platelet effect
Monitor
Injectable / compounded
Sterility · contamination
Avoid
Semax vs. the neuropeptide & nootropic field
Agent
Origin / class
Primary mechanism
Primary use
Human evidence
Status
Semax
ACTH(4-10) heptapeptide
MC4R → BDNF/NGF · serotonin · DA sensitization
Stroke · nootropic · neuroprotection
Russian-centric · limited
Russia-approved · FDA-unapproved
Selank
Tuftsin heptapeptide (PGP)
GABA-A · enkephalinase inhibition
Generalized anxiety · cognitive clarity
Russian-centric · limited
Russia-approved · FDA-unapproved
Noopept (GVS-111)
Russian dipeptide
AMPA/NMDA modulation · indirect BDNF/NGF
Nootropic · neuroprotection
Very limited
Russia-marketed · FDA-unapproved
Cerebrolysin
Porcine brain peptide mixture
Direct neurotrophin mimicry (mixture)
Stroke · vascular dementia
More extensive; some Western
Approved in many countries · IV
Piracetam (racetams)
Pyrrolidinone
AMPA modulation · membrane fluidity
Historical nootropic
Older, mixed human data
Varies by country · not FDA
Modafinil
Wakefulness agent
DAT inhibition · orexin
Narcolepsy · wakefulness
Strong RCT (on-label)
FDA-approved (narcolepsy)
Donepezil (AChEI)
Cholinesterase inhibitor
Acetylcholinesterase inhibition
Alzheimer's symptomatic
Strong RCT
FDA-approved
N-Acetyl Semax Amidate
Modified Semax (Ac- / -NH₂)
As Semax · longer fluid stability
Nootropic (community)
Inferred from Semax · scarce
Research chemical only
ACTH(4-10) (parent)
Native melanocortin fragment
MC receptor · rapidly degraded
Research substrate
Historical · minutes t½
Not a marketed drug
05 · Safety profile & contraindications
Favorable short-term profile; long-term data essentially absent.
In decades of Russian clinical experience, Semax is described as well tolerated with very low discontinuation rates and no prominent serious adverse events at approved intranasal doses — helped by its non-hormonal status (no HPA-axis or glucocorticoid effects). The most common issues are local and mild. The dominant safety concern is not a known toxicity but a knowledge gap: there are no long-term human safety studies, no Western pharmacovigilance, and theoretical concerns about sustained BDNF elevation remain uncharacterized. Adverse-event frequencies below derive from Russian clinical data and secondary compilations.
Common / Mild Adverse Events (Russian clinical data)
Nasal cavity discoloration~10% — the most commonly reported effect (Kolomin 2013 review). Local, route-specific, and generally cosmetic/reversible. Rotate nostrils and avoid exceeding labeled drop counts.
Mild nasal irritation~12.4% — expected with any chronic intranasal peptide. Usually transient; persistent irritation warrants a drug holiday and inspection for mucosal injury.
Nasal congestion~8.1% — local mucosal response. Distinguish from intercurrent rhinitis/infection, which is a route-specific relative contraindication (impaired delivery + irritation).
Headache~5.7% — typically early and self-limited. Encourage hydration; reassess if persistent or severe.
Transient insomnia~3.2% — when dosed late in the day, consistent with the activating monoaminergic profile. Shift dosing earlier; avoid evening administration.
Restlessness at higher dosesReported with higher (community) doses — again reflecting serotonergic/dopaminergic activation. Dose-reduce if it occurs.
Elevated blood glucose in diabetics~7.4% — clinically relevant in diabetic patients. Monitor glucose during courses in diabetes; not described as significant in non-diabetics.
Rare allergic reactions~1.1% — standard peptide hypersensitivity caveat. Discontinue on any hypersensitivity sign; avoid in known ACTH-analog allergy.
Discontinuation rateDescribed as very low in Russian clinical experience, suggesting generally good short-term tolerability at approved doses.
Knowledge Gaps & Theoretical Risks
No long-term human safety dataThe single most important safety statement. There are no long-term studies in healthy individuals; rare or delayed effects would not be captured by any Western pharmacovigilance system, which does not track Semax.
Sustained BDNF elevation · theoreticalProlonged neurotrophin elevation has been associated (in other contexts) with hypomanic episodes, growth promotion in BDNF-sensitive tumors, and excessive synaptic sprouting. These are theoretical, not demonstrated with Semax dosing — but they are genuine, unstudied gaps.
Active malignancy · cautionListed in expert sources as a relative contraindication specifically because of BDNF upregulation and growth-promoting pathways — particularly hormone-sensitive tumors.
Epilepsy · listed contraindicationListed as a contraindication, though paradoxically one animal study suggests anti-epileptic effects at 50 µg/kg. Treat as caution pending clarity; involve neurology.
Neonatal exposure signalOne animal study found neonatal Semax produced lasting behavioral changes in adult rats — raising an unresolved question about long-term neuroplastic remodeling, relevant to any pediatric or repeated-course use.
No overdose datasetNo documented overdose cases. Given the short plasma half-life and indirect gene-level mechanism, acute toxicity risk is considered low — but there is no formal toxicological overdose dataset in humans.
Gray-market quality riskFor non-pharmaceutical product: mislabeling, unverifiable purity, non-sterile preparation, and endotoxin risk are real and independent of the molecule's intrinsic safety.
Contraindication reference
Compiled from translated Russian prescribing information and expert secondary sources. The current Russian SmPC should be consulted directly for definitive contraindications and pediatric age thresholds.
No approved injectable anywhere · sterility, identity, endotoxin risk.
Suggested monitoring for Semax protocols
Baseline
Indication confirmation, BP, nasal mucosa inspection, psychiatric history (screen for psychosis, bipolar disorder), malignancy history (esp. hormone-sensitive), pregnancy status, concurrent serotonergic/stimulant/anticoagulant medications. Fasting glucose/HbA1c in diabetics. No HPA-axis or cortisol testing required — Semax is non-hormonal.
During acute stroke course
Within a supervised stroke pathway: NIHSS trajectory, motor function, BP, and tolerability. Confirm Semax remains strictly adjunctive to reperfusion/standard therapy. Watch for restlessness/insomnia and local nasal effects.
Cognitive/nootropic course
Symptom review at each course (sleep, mood, restlessness, nasal irritation/discoloration). Glucose in diabetics. Subjective cognitive/functional benefit vs side-effect burden — discontinue if no benefit. Earlier-in-day dosing to protect sleep.
Post-stroke maintenance
Functional scales (Barthel, Rivermead, NIHSS) at each course boundary; plasma BDNF where assayable as an exploratory response marker. Continue maintenance courses only on a measurable functional trajectory.
Repeated / long-term use
No validated long-term protocol exists. Given the absent long-term safety data, periodic reassessment of ongoing benefit, mood stability, and (in at-risk patients) malignancy surveillance is prudent. Empirical cycling is community practice, not evidence-based.
Stop / hold criteria
Hypersensitivity reaction → discontinue. New psychosis or manic symptoms → discontinue, psychiatry. Significant nasal mucosal injury → hold. New malignancy diagnosis → reassess risk/benefit. Pregnancy → discontinue. Uncontrolled hypertension → hold until controlled.
06 · Key studies & research program
A robust preclinical base; a thin, Russian-centric clinical one.
Semax's evidence base is inverted relative to a typical Western drug: the preclinical mechanistic literature is genuinely strong and reproducible, while the human clinical literature is dominated by small, unblinded, single-center, Russian-language studies that have never been independently replicated. Approximately 12 distinct human studies exist across all indications; very few are English-accessible full text, only 1–2 are placebo-controlled, and none are registered on ClinicalTrials.gov or covered by a Cochrane review. Below: the studies that define — and limit — the clinical positioning of the molecule, with every grade and caveat explicit.
Stroke · flagship
12–18mg
Gusev 1997 controlled trial · n=30 Semax vs 80 controls · faster neurological restoration, esp. motor · the highest-evidence indication (still B/C)
Meta-analysis · stroke
n=654
8 studies screened, 3 (n=181) pooled · significant NIHSS & Rivermead improvement · Russian-literature only · authors urged a real double-blind trial
fMRI · healthy
n=24
Lebedeva 2018 · RCT, placebo-controlled · DMN rostral subcomponent expansion at 5 & 20 min · strongest independently accessible human signal
Mechanism · anchor
↑1.4×
Dolotov 2006 · BDNF protein, TrkB phosphorylation & mRNA in rat hippocampus · the molecular foundation of every cognitive claim
BControlled trial · acute stroke · flagship
Gusev, Skvortsova, Myasoedov et al. (1997)
Controlled (non-randomized) trial: 30 Semax patients vs 80 matched controls. Semax 1% intranasal 12–18 mg/day added to intensive therapy increased neurological-function restoration, with particular motor improvement. Most effective: 12 mg/day (moderate) and 18 mg/day (severe). Limitations: non-randomized, unblinded allocation, small Semax arm, Russian-language (English abstract only).
Semax in acute stroke — meta-analysis (Bull Rehab Med 2018)
Screened 167 PubMed + 197 elibrary.ru articles; 8 studies (n=654) met quality thresholds, 3 (n=181) shared endpoints. Semax 1% (12–18 mg/day, 10–14 d) produced significant NIHSS reduction in moderate/severe stroke at days 10–14 and 21, mRS improvement, and Rivermead Mobility Index improvement across all severities. Limitations: entirely Russian literature, heterogeneous designs, small studies, not Cochrane. Authors explicitly called for a multicenter double-blind trial.
Referenced Phase 2/3 RCT (cited as Stroke 2007, n=184)
Secondary sources cite a 184-patient RCT in which Semax 0.1% vs placebo reduced NIHSS decline 34%, improved 90-day mRS 28%, and limited infarct-volume increase 23%. ⚠ Citation caution: the PMID and effect sizes appear in secondary peptide reviews but could not be independently confirmed against the original Stroke article. Verify PMID 17395875 directly before citing. If accurate, this would be the strongest human RCT evidence in stroke.
Gusev et al. (2018) — Semax, BDNF & rehabilitation
n=110 stroke patients (mean age 58). Regimen: 2 courses of 6,000 µg/day × 10 days, 20-day interval; early vs late rehabilitation. Semax raised plasma BDNF (sustained); high BDNF accelerated Barthel-index recovery and correlated with Barthel score; early rehab + Semax was synergistic. Limitations: open-label, no placebo, Russian-language — Grade C.
Randomized, placebo-controlled resting-state fMRI in healthy volunteers (n=24; 14 Semax, 10 placebo; mean age 43.9). 1% Semax, total 1.2 mg; imaging at baseline, 5, and 20 min. Semax group showed greater DMN rostral-subcomponent (medial frontal cortex) volume at 5 and 20 min. One of very few placebo-controlled human studies of a central Semax effect. Grade C — small n, single study, surrogate endpoint.
The landmark mechanistic study. A single 50 µg/kg intranasal dose in rats raised hippocampal BDNF protein ~1.4-fold (peak 3 h), TrkB phosphorylation ~1.6-fold, BDNF exon III mRNA ~3-fold, TrkB mRNA ~2-fold, and increased conditioned-avoidance learning. The reproducible molecular signature underlying the entire cognitive/neuroprotective rationale.
Eremin et al. (2005, Neurochemistry International)
Semax raised striatal 5-HIAA +25% (tissue) and up to +180% (extracellular) within 1–4 h; did not change basal dopamine but amplified d-amphetamine-induced dopamine release and locomotion — a serotonin-enhancing, dopamine-sensitizing/priming profile distinct from classical stimulants. Underpins the mood/attention effects and the stimulant-interaction caution.
Medvedeva, Myasoedov et al. (2014, BMC Genomics) + 2024 follow-up
Genome-wide microarray in rat focal ischemia: Semax modulates hundreds of genes — immune-cell genes at 3 h, immune-response and vascular/angiogenic programs at 24 h (GEO GSE163654). A 2024 follow-up (Biomedicines, PMID 39066823) confirmed ACTH-like peptides including Semax compensate the ischemia-disrupted gene-expression profile a day after stroke. Systems-level support for the "compensation" model of neuroprotection.
ACS Chemical Neuroscience (2022) + APP/PS1 mouse (2025)
A rare non-Russian study (PMID 35099917) showed Semax sequesters Cu²⁺ and inhibits Aβ:Cu²⁺-mediated fibrillogenesis in membrane models. 2025 transgenic work (PMID 41479572) found Semax + a derivative improved cognition (open field, novel object, Barnes maze) and reduced amyloid burden in APPswe/PS1dE9 mice. Suggestive only — no human AD trials; ADDF found no evidence for Alzheimer's disease.
Semax is best characterized as biologically credible but clinically unvalidated by Western standards. The preclinical mechanistic literature is genuinely robust — reproducible BDNF/NGF induction, neurotrophin-receptor signaling, cytokine modulation, and gene-level neuroprotection across multiple well-designed rodent models. The human clinical base, however, is dominated by Russian-language publications from a small cluster of institutions (chiefly the Institute of Molecular Genetics and collaborating neurology departments) that are mostly unblinded, small (n < 200), lacking Western-standard endpoints, inaccessible in full text to non-Russian readers, and entirely unreplicated independently.
A single fMRI study (PMID 30225715, n=24) is perhaps the strongest independently accessible placebo-controlled human evidence — it confirms a measurable brain effect but falls far short of establishing clinical efficacy. The ADDF Cognitive Vitality program, reviewing the English-accessible evidence, found that well-conducted studies are lacking and that there is no evidence for Alzheimer's disease.
Applying GRADE methodology, a skeptical Western clinician would rate the stroke evidence low to very low (downgraded for risk of bias, indirectness, imprecision) and the cognitive evidence very low — while acknowledging the preclinical rationale is strong enough to justify a properly designed trial. The practical conclusion: a pharmacologically plausible neuropeptide used in Russia for decades with an acceptable short-term safety profile, separated from Western regulatory acceptance by an evidence gap that has not narrowed in decades and is unlikely to close without strategic pharmaceutical investment that the expired-patent, commodity-market landscape makes improbable.
Critical framing: nearly all human Semax data originates from Russian institutions, is published in Russian-language journals, and has not been independently replicated in Western peer-reviewed RCTs. Evidence grades below reflect that reality honestly.