Cerebrolysin 60mg Peptide

Cerebrolysin is not a single molecule. It is a standardised enzymatic breakdown product of purified porcine brain protein: a defined mixture of low molecular weight peptides and free amino acids, developed in Austria and manufactured by EVER Neuro Pharma, formerly EBEWE. That distinction is the first thing to get right, because almost every downstream question changes because of it. There is no sequence to quote, no molecular formula, no single molecular weight, and no PubChem compound record. A search of PubChem for the name returns only substance entries, because the database has nothing to draw a structure from.

What can be specified is the process and the fraction. Purified brain protein is digested under controlled proteolysis, lipid and antigenic protein are removed, and the product is filtered so that the peptide fraction lies below 10 kDa. The manufacturer's own characterisation puts the result at roughly 25% low molecular weight peptides and 75% free amino acids, a ratio quoted in the primary literature as far back as 2001 and sometimes given as 20% and 80% in later material. The registered pharmaceutical form is an aqueous ampoule solution standardised to 215.2 mg of concentrate per millilitre, not a lyophilised powder.

Cerebrolysin is registered as a prescription medicine in dozens of countries across Europe, Asia and the former Soviet states, and the closely related preparation registered in China is called cerebroprotein hydrolysate. It has never been reviewed or authorised for marketing by the US Food and Drug Administration. Its human trial record is unusually large for a compound of this kind, running to 42 studies registered on ClinicalTrials.gov as of August 2026, and it is genuinely mixed: two of the largest randomised trials reported positive primary endpoints, the largest of all reported a neutral one, and successive Cochrane reviews have declined to find a benefit in stroke while calling the vascular dementia evidence very low quality. This page reports both directions.

The working hypothesis behind Cerebrolysin is that its peptide fraction produces effects resembling those of endogenous neurotrophic factors without containing them. This is a point where vendor descriptions and physical chemistry disagree. Mature BDNF is about 27 kDa as a homodimer, NGF about 26 kDa, GDNF about 30 kDa and CNTF about 23 kDa. All of them sit well above the 10 kDa filtration cutoff that defines the preparation, so intact neurotrophins cannot survive in it. What the literature actually describes is neurotrophic-factor-like activity carried by fragments, most of it attributed to the peptide fraction rather than to any single identified sequence.

That attribution has been tested directly. In embryonic chicken telencephalon neurons held in 2% serum, both Cerebrolysin and an artificial free amino acid mixture matched to its amino acid content promoted neurite outgrowth at 4 days in vitro. By 8 days the amino acid mixture had stopped working and the cultures had degenerated, while the Cerebrolysin-treated network remained differentiated. Neither added BDNF nor serum supplementation to 5% and 10% preserved the network. The separation of the two arms at day 8 is the cleanest published argument that the peptide fraction, not the free amino acids, carries the protective activity (Hartbauer and colleagues, Journal of Neural Transmission, 2001).

Downstream, the best characterised pathway is Sonic hedgehog. Neural progenitor cells taken from the subventricular zone of adult rats proliferated and differentiated into neurons and myelinating oligodendrocytes when exposed to Cerebrolysin, with upregulation of Shh and its receptors patched and smoothened. Cyclopamine, a smoothened inhibitor, abolished the effect. In rats subjected to embolic stroke and given Cerebrolysin starting 24 hours after onset, subventricular zone progenitor proliferation, peri-infarct neurogenesis, oligodendrogenesis and axonal remodelling all increased, and neurological function improved from week 3 to week 5. In vivo cyclopamine reversed all of it (Zhang and colleagues, Stroke, 2013). A pathway that can be switched off pharmacologically and take the behavioural effect with it is a stronger mechanistic claim than a correlation with a transcript.

Other signalling nodes recur across the preclinical record with less rigorous support. In reserpine-treated rats, Cerebrolysin at 2.5 and 5 mL/kg intraperitoneally raised cortical and hippocampal BDNF protein, altered AKT, CREB and BDNF mRNA expression, reduced malondialdehyde and TNF-alpha, and raised noradrenaline, serotonin and dopamine, with effects running parallel to fluoxetine at 5 mg/kg orally in the same experiment (El-Marasy and colleagues, Chemico-Biological Interactions, 2021). In aged 23 to 24 month Wistar rats given 2.5 mL/kg intraperitoneally over 20 injections, the effect landed on the NGF system rather than BDNF: neocortical TrkA and p75NTR receptor loss was counteracted and proNGF expression fell, while total NGF and the BDNF system were essentially unchanged, and the water maze effect was minor and did not extend to memory retention (Stepanichev and colleagues, Restorative Neurology and Neuroscience, 2017). That study is worth reading precisely because it is a negative-leaning result from inside the supportive literature.

Manufacture begins with purified porcine brain protein, which is digested under standardised proteolysis, stripped of lipid and antigenic protein, and filtered so that everything remaining in the peptide fraction is below 10 kDa. Standardisation is therefore a process specification and a fraction specification, not a structural one. Two batches count as equivalent because they were made the same way and pass the same fractionation and biological assays, not because anyone showed they contain the same molecules in the same proportions.

This is why the analytical fields that a synthetic peptide record fills in are empty here, and why filling them in anyway is a warning sign. A purity figure quoted as a single HPLC percentage has no clear meaning for a material that is three-quarters free amino acids by design: a chromatogram of Cerebrolysin is supposed to be a broad envelope of species, not one peak. A stated molecular weight of roughly 3,000 g/mol describes an average that nothing in the preparation is required to sit at. A sequence field is unfillable in principle.

The one identified sub-fraction with a name in the literature is E021, the isolated peptide component. In 24 month old rats, E021 and whole Cerebrolysin both raised GluR1 immunoreactivity across most regions of the hippocampal formation and both improved Morris water maze performance, which is the clearest published evidence that the activity travels with the peptides rather than the amino acids. Below E021 the resolution stops. No individual fragment has been purified, sequenced and shown to reproduce the preparation's effects, and any ingredient list naming specific intact neurotrophins is describing something the 10 kDa cutoff excludes.

Material sold to laboratories as a lyophilised powder is a further step removed from the object the clinical literature studied. Every trial cited on this page infused an aqueous solution standardised at 215.2 mg of concentrate per millilitre. A dry powder is a different physical form with a different stability profile, and its correspondence to the ampoule preparation is a question for a certificate of analysis and an independent assay, not an assumption.

CASTA is the largest randomised trial of Cerebrolysin ever run and its primary endpoint was neutral. 1,070 patients with acute ischaemic hemispheric stroke across Asian centres were randomised within 12 hours of onset to 30 mL of Cerebrolysin daily by intravenous infusion for 10 days or to saline, both on a background of aspirin at 100 mg daily, and followed to day 90. The confirmatory endpoint, a combined global directional test of modified Rankin Scale, Barthel Index and NIH Stroke Scale, showed no significant difference between groups. A post hoc analysis in the subgroup with baseline NIHSS above 12 found odds ratios of 1.27 on both NIHSS and mRS with confidence bounds at or below 1, and 90 day cumulative mortality of 10.5% against 20.2% in that subgroup. The authors labelled the trial neutral and the subgroup finding as needing confirmation (Heiss and colleagues, Stroke, 2012).

CARS asked a narrower question in a different window and got a strong answer. Patients were randomised 24 to 72 hours after stroke onset to 30 mL intravenously once daily for 21 days or placebo, with every participant entering a standardised 21 day rehabilitation programme. The primary endpoint, Action Research Arm Test score at day 90, gave a Mann-Whitney estimator of 0.71 favouring Cerebrolysin, 95% CI 0.63 to 0.79, p below 0.0001, which is a large effect on a motor scale. Global status across 12 scales gave 0.62. Premature discontinuation was 3.8%. The paper is explicit that the study was exploratory with a relatively small sample and that confirmation in a large trial was required (Muresanu and colleagues, Stroke, 2016).

The Cochrane review reaches a different conclusion from the same field, and the reason is that it asks about hard outcomes. The 2023 update included 7 randomised controlled trials and 1,773 participants, one of them of the related preparation Cortexin. On all-cause death, Cerebrolysin or Cortexin probably make little to no difference: risk ratio 0.96, 95% CI 0.65 to 1.41, moderate certainty. None of the included studies reported poor functional outcome defined as death or dependence at end of follow-up, early death, or quality of life. On serious adverse events the total was unchanged at risk ratio 1.16, but non-fatal serious adverse events increased, risk ratio 2.39, 95% CI 1.10 to 5.23, and more sharply in the subgroup receiving 30 mL for 10 days, risk ratio 2.87. The review also records that the manufacturer supported three of the multicentre studies in full or in part (Ziganshina and colleagues, Cochrane Database of Systematic Reviews, 2023).

Those results are not contradictory so much as answers to different questions. CARS measured arm function in a rehabilitation cohort. Cochrane measured death and serious harm across all comers, and found that the functional outcomes it wanted had not been reported. The 2021 European Academy of Neurology and European Federation of Neurorehabilitation Societies guideline sits between them: after screening 17,969 unique records under GRADE, it recommends Cerebrolysin at 30 mL intravenously for a minimum of 10 days, alongside citalopram, for early neurorehabilitation after acute ischaemic stroke, while recommending against amphetamine, dextroamphetamine, fluoxetine and lithium. Anyone weighing that guideline should note that its author list includes investigators from the Cerebrolysin trial programme.

CAPTAIN II was a single-centre phase IIIb/IV prospective, randomised, double-blind, placebo-controlled trial in moderate to severe traumatic brain injury. Eligible patients had an admission Glasgow Coma Score of 7 to 12; mean age was 47.4, mean admission GCS 10.4 and mean Baseline Prognostic Risk Score 2.6. The regimen was 50 mL of Cerebrolysin or physiological saline daily for 10 days as an addition to standard care, followed by two further 10 day cycles at 10 mL daily. 142 patients enrolled and 139 entered formal analysis.

The primary endpoint was deliberately unusual: a multidimensional ensemble of 13 outcome scales tested with a multivariate directional test under a priori ordered hypotheses, rather than a single scale. It reached significance at day 90 with a combined Mann-Whitney estimator of 0.59, 95% CI 0.52 to 0.66, p equal to 0.0119, which the authors describe as a small to medium effect. Safety and tolerability observations were comparable between arms (Muresanu and colleagues, Neurological Sciences, 2020).

The ensemble design deserves attention rather than dismissal. Traumatic brain injury outcome is genuinely multidimensional, and single-scale endpoints have a long history of missing effects that are real but distributed. The design also makes the result harder to compare with trials that used the Glasgow Outcome Scale alone, and harder to reproduce. A 2023 systematic review pooling ten clinical studies and 8,749 patients, retrospective and prospective, found statistically significant changes in Glasgow Coma Score and Glasgow Outcome Scale with Cerebrolysin, and no effect on all-cause mortality or on length of stay (Jarosz and colleagues, Brain Sciences, 2023). The pattern across the traumatic brain injury literature is consistent: functional and cognitive scales move, survival does not.

The vascular dementia file is the one where a Cochrane review found a positive direction and still declined to endorse it. Pooling MMSE and ADAS-cog+ data from three studies and 420 people gave a standardised mean difference of 0.36, 95% CI 0.13 to 0.58. Global function response rates from two studies and 379 participants gave a risk ratio of 2.69, 95% CI 1.82 to 3.98. Adverse event rates did not differ. Every one of those estimates was graded very low quality. The reviewers' position, in substance: courses of intravenous Cerebrolysin improved cognition and general function, the analyses were limited by heterogeneity and by high risk of bias in the included papers, and if a benefit exists the effects may be too small to be clinically meaningful. They also record that no new vascular dementia trial had appeared since the previous version of the review (Cui and colleagues, Cochrane Database of Systematic Reviews, 2019).

A smaller randomised study in mild to moderately severe probable vascular dementia by NINDS-AIREN criteria added a biomarker dimension. Patients received intravenous infusions of 10 mL or 30 mL of Cerebrolysin or saline five days a week for four weeks, followed by a 12 week open-label extension. At the week 16 follow-up, cognitive performance on ADAS-cog+ remained improved at both 10 mL and 30 mL, and the slow-to-fast quantitative EEG power ratio, an index of EEG slowing, remained reduced at 30 mL, with a positive correlation between the two changes. 33 of the original 41 patients were assessed at week 16 (Muresanu and colleagues, Journal of the Neurological Sciences, 2010).

In Alzheimer's disease, the dose-finding trial by Alvarez and colleagues gave 100 mL intravenous infusions containing 10, 30 or 60 mL of Cerebrolysin diluted in saline, or saline alone, five days a week for four weeks and then two days a week for eight more, with endpoints at week 24. In the moderate to moderately severe subgroup, 133 patients with MMSE 14 to 20, global clinical function on CIBIC+ improved significantly at all three concentrations, and cognition, initiation of activities of daily living, and neuropsychiatric symptoms each improved at one of the three. Total activities of daily living, MMSE and the Trail Making Test did not reach significance (European Journal of Neurology, 2011). This is a positive trial with a scattered effect pattern, and its authors said the findings needed confirmation in larger trials. No such confirmatory Alzheimer's trial has been published since.

Anyone reading the Cerebrolysin preclinical literature in 2026 has to read the retraction notices alongside it. PubMed indexes ten Cerebrolysin papers under the Retracted Publication type, and a run of notices appeared through 2025 and 2026 in Acta Neuropathologica, Molecular Neurobiology, Neurochemical Research, BMC Neuroscience and the Journal of Alzheimer's Disease, plus an editorial expression of concern in Acta Neuropathologica. The affected work clusters in two places: transgenic mouse model studies of tauopathy, Rett syndrome and amyloid pathology from a laboratory whose broader published record became the subject of a widely reported research integrity investigation in 2024, and nanoparticle and spinal cord injury work from a separate group.

This does not invalidate the field. The trials cited on this page are independent of the retracted animal work, and so are the mechanistic results that carry the most weight here: the culture study separating the peptide fraction from the amino acid fraction, the cyclopamine reversal experiment, and the aged-rat E021 comparison all come from other groups. It does mean that the transgenic model literature, which is the part of the Cerebrolysin story most often summarised as evidence in Alzheimer's disease and tauopathy, should be treated as unsettled until the surviving papers are re-examined.

The second quality issue is structural rather than ethical. The Cochrane stroke review notes that the manufacturer supported three of the included multicentre studies, either fully or by supplying drug, placebo, randomisation codes, grants or statisticians, and judged two studies at high risk of other bias. The same biometrics group appears on CARS, on CAPTAIN II and on the European guideline. None of that makes the results wrong, and industry sponsorship is normal for trials of this size. It does mean that the strongest positive findings and the guideline recommending the compound come from an overlapping set of investigators, and that the independent replication the trial authors themselves called for has not arrived.

Semax and Selank are the two comparisons most often drawn, and they are structurally the opposite kind of object. Both are synthetic, sequence-defined peptides: Semax is the heptapeptide Met-Glu-His-Phe-Pro-Gly-Pro built on ACTH(4-7), Selank is a tuftsin analogue. Their molecular identity is exact, their analytical characterisation is straightforward, and their clinical files are small and concentrated in Russian-language journals with almost no registry presence. Cerebrolysin inverts every one of those properties: its identity is a process rather than a structure, its analysis is a fraction specification, and its clinical file is large, international, registry-visible and openly contested.

Cortexin is the closer relative and the more instructive comparison. It is a polypeptide preparation from bovine or porcine cerebral cortex, registered and widely used in Russia, and it is close enough in kind that the 2023 Cochrane stroke review folded a 272 participant Cortexin trial into the same analysis as the Cerebrolysin studies. Both are animal-derived brain protein hydrolysates that reach the clinic on process standardisation rather than molecular definition, and both carry the concerns that attach to any tissue-derived biological preparation.

The nearest thing to a defined synthetic answer to Cerebrolysin is P21, a modified CNTF-derived tetrapeptide linked to adamantane, developed explicitly to deliver neurotrophic activity as a single characterised molecule. P21 has rodent neurogenesis data and no human clinical trials at all. The trade is stark and worth stating plainly: Cerebrolysin has thousands of randomised patients and no molecular identity, P21 has a molecular identity and no patients.

The registered ampoule product is an aqueous solution kept below 25 degrees Celsius, protected from light, and specifically not frozen, with a shelf life of five years. That profile is the opposite of what a laboratory expects from a peptide, and it is a property of the formulation: the solution is stabilised and does not benefit from freezing, which risks concentration gradients and precipitation on thaw in a mixture holding this many species.

A lyophilised powder sold under the same name has different requirements, and the two should not be conflated when reading a storage instruction. Dry peptide material is generally held cold, desiccated and protected from light, since the dominant degradation routes for a material dominated by low molecular weight peptides and free amino acids are moisture-driven hydrolysis and oxidation of methionine, cysteine and tryptophan residues. Repeated freeze-thaw cycling of an aqueous solution is the standard thing to avoid for any peptide material, and it matters more here than for a single-species peptide, because differential precipitation changes the composition of the mixture rather than simply lowering its concentration.

There is no shortcut for verifying identity in this material. For a synthetic peptide, mass spectrometry against an expected mass answers the question in one measurement. For a hydrolysate, the achievable checks are compositional and comparative: amino acid analysis, the size distribution of the peptide fraction against the 10 kDa specification, and chromatographic fingerprint comparison against a reference lot. A certificate of analysis reporting a single purity percentage and one molecular weight has not performed any of those checks.