Cerebrolysin

01 · What it is

Cerebrolysin is the most heavily trialled compound in this project, and the only one whose most recent Cochrane review reports a possible harm as its most confident finding.

That review — the seventh version of a systematic review first published in 2010 — found moderate- certainty evidence of an increase in non-fatal serious adverse events, risk ratio 2.39 with a 95% confidence interval of 1.10 to 5.23 across three trials and 1,335 participants ¹⁵. In the subgroup receiving 30 mL a day for ten days, the estimate rose to 2.87, 95% CI 1.24 to 6.69 ¹⁵. Over the same seven trials it found no benefit on all-cause death.

What it is made of. Cerebrolysin is not a peptide. It is a mixture of low-molecular-weight peptides and amino acids derived from porcine brain ¹⁵, manufactured by EVER Neuro Pharma GmbH, formerly Ebewe Pharma, whom the Cochrane reviewers contacted directly during their searches ⁴. Like Thymalin elsewhere in this project, it is a tissue extract rather than a defined molecule, and no paper cited here reports a composition specification, a batch-release assay or a list of constituent species. It is widely used for acute ischaemic stroke in Russia, Eastern Europe, China and other Asian and post-Soviet countries ¹⁵.

The evidence base is large, weak, and has stopped growing in one of its two indications. The stroke review has been published seven times between 2010 and 2023 ^{1,4,5,6,12,15} by a group led by Liliya Ziganshina, working from Kazan and Moscow. The vascular dementia review has appeared twice, in 2013 and 2019 ^2,11, from West China Hospital in Chengdu. The 2019 update states plainly that no new studies had become eligible since the 2013 version ¹¹ — six randomised trials and 597 participants, unchanged in six years.

The two Cochrane reviews do not point the same way, and the difference is instructive. In vascular dementia the reviewers found a benefit: cognition improved, standardised mean difference 0.36 (95% CI 0.13 to 0.58) across three studies and 420 people, and global function response improved, risk ratio 2.69 (95% CI 1.82 to 3.98) across two studies and 379 participants ¹¹. They graded both as very low-quality evidence, noted that every included study whose funding was identifiable had been supported by the pharmaceutical industry, and wrote that if there are benefits the effects may be too small to be clinically meaningful ¹¹. In acute ischaemic stroke, where the trials are larger and better documented, no benefit on death was found and a possible harm was ¹⁵.

And a separate group of meta-analyses reports benefit in stroke. These carry authors affiliated with the manufacturer ^3,9,10. They are meta-analyses of real randomised trials in peer-reviewed journals, and they pool substantially the same studies the Cochrane reviewers assessed.

The unaffiliated analyses side with Cochrane, in plain language. Two meta-analyses published seven weeks apart in 2017, by unrelated groups in Hangzhou and Shanghai, both concluded that routine use in acute ischaemic stroke cannot be supported by the available randomised evidence ^7,8.

02 · Evidence at a glance

Evidence grade

Mixed Evidence — abundant randomised human data whose conclusions conflict by analyst and by indication. See the note below

What it is

A mixture of low-molecular-weight peptides and amino acids from porcine brain ¹⁵ — not a defined molecule

Manufacturer

EVER Neuro Pharma GmbH, formerly Ebewe Pharma ⁴

Composition specification published

None identified

Cochrane reviews

Seven on acute ischaemic stroke, 2010–2023 ^{1,4,5,6,12,15}; two on vascular dementia, 2013 and 2019 ^2,11

Most confident finding of the 2023 stroke review

Increase in non-fatal serious adverse events: RR 2.39, 95% CI 1.10–5.23; 3 trials, 1,335 participants, moderate certainty ¹⁵

That signal at 300 mL cumulative dose

RR 2.87, 95% CI 1.24–6.69; 2 trials, 1,189 participants ¹⁵

All-cause death, same review

RR 0.96, 95% CI 0.65–1.41; 6 trials, 1,689 participants, moderate certainty ¹⁵

Stroke trials reporting death or dependence

None. No included study reported poor functional outcome, early death, quality of life, or time to restoration of capacity for work ¹⁵

Vascular dementia, cognition

SMD 0.36, 95% CI 0.13–0.58; 3 studies, 420 people, very low-quality evidence ¹¹

Vascular dementia, global function

RR 2.69, 95% CI 1.82–3.98; 2 studies, 379 participants, very low quality ¹¹

New vascular dementia trials since 2013

None ¹¹

Industry involvement

Manufacturer supported three multicentre stroke studies ¹⁵; every vascular dementia trial with identifiable funding was industry-supported ¹¹

Unaffiliated 2017 meta-analyses

Two, seven weeks apart, both concluding routine use is not supported ^7,8

Largest independent efficacy pooling

14 trials, 2,884 patients: NIHSS mean difference 1.39 (95% CI 0.53–2.25); functional independence RR 1.31 (0.90–1.91), not significant ¹⁸

Position against other agents

35 studies, 18,423 cases, 9 interventions: reduced mortality vs conventional care, but placed among agents with a higher adverse-effect rate and absent from the safest-and-most-effective group ¹⁷

Approval status

Marketed in numerous countries; not approved in the United States. Specific authorisation records not verified here

On the grade. The rubric’s Approved Pharma tier requires a marketing authorisation in at least one jurisdiction. Cerebrolysin is manufactured commercially and used clinically across many countries, so that tier may well apply — but this guide has not consulted a regulatory record for any jurisdiction and does not assert one. It is graded Mixed Evidence because that tier describes what is actually distinctive here: human and preclinical data exist in quantity and their conclusions conflict.

03 · Mechanism of action

An extract, so the mechanism is a claim about a mixture

Cerebrolysin’s proposed mechanism is neurotrophic. Cochrane’s own description is that it is a porcine brain-derived preparation said to have neurotrophic and neuroprotective activity ¹¹ — the qualifier is the reviewers’, and it is doing real work.

The difficulty is definitional rather than empirical. A mixture derived from enzymatically digested brain tissue contains many species, and no paper in this file identifies which of them is responsible for any observed effect, or reports what fraction of the preparation any named species represents. A 2025 review of post-stroke neuroinflammation places Cerebrolysin among neuroprotective agents alongside edaravone and citicoline and calls it a neurotrophic factor ²⁰ — a description of what it is proposed to resemble rather than what it has been shown to contain.

Every mechanistic statement about this compound is a statement about an unspecified mixture, and that limitation propagates into every dose, every batch and every trial.

The Cortexin problem

The 2023 Cochrane review’s inclusion criterion is randomised trials of Cerebrolysin or Cerebrolysin-like agents started within 48 hours of stroke onset ¹⁵. In that update the reviewers added a trial of Cortexin, a different preparation, contributing 272 of the 1,773 participants ¹⁵.

A review that has to define its intervention as “or similar peptide mixtures” is telling the reader something about the category. Where a compound is defined by tissue of origin and manufacturing process rather than by structure, the boundary between it and its neighbours is a judgement call, and Cochrane made that call explicitly rather than pretending it did not exist.

What the trials were designed to detect, and what they were not

The clinical programme has used functional and global outcome scales rather than mechanistic endpoints: neurological improvement in the early post-stroke period ¹⁰, motor recovery during rehabilitation ⁹, cognitive scales in dementia ^3,11, and functional recovery measures in acute stroke ^7,8,18.

But the 2023 Cochrane review records that none of its seven included studies reported poor functional outcome — death or dependence at the end of follow-up — nor early death, nor quality of life, nor time to restoration of capacity for work ¹⁵.

Forty years of trials in a condition whose defining consequence is dependence, and not one of the included studies reported dependence. That is a fact about study design, not about the compound, and it limits what any amount of pooling can establish.

04 · Key research findings

The 2023 Cochrane review of acute ischaemic stroke. Seven randomised trials, 1,773 participants, searching Cochrane’s stroke register, CENTRAL, MEDLINE, Embase, Web of Science, LILACS and a set of Russian databases, plus reference lists, trial registers and conference proceedings ¹⁵. All-cause death showed no benefit: RR 0.96, 95% CI 0.65 to 1.41, six trials, 1,689 participants, moderate certainty. Total serious adverse events showed no difference, RR 1.16 (0.81 to 1.66), and fatal serious adverse events none either, RR 0.90 (0.59 to 1.38). Non-fatal serious adverse events increased: RR 2.39, 95% CI 1.10 to 5.23, moderate certainty, rising to 2.87 (1.24 to 6.69) in the subgroup given a cumulative 300 mL ¹⁵.

The reviewers’ own summary is that Cerebrolysin probably has no beneficial effect on preventing all-cause death, probably has no effect on the total number of people with serious adverse events, and that moderate-certainty evidence indicates a potential increase in non-fatal serious adverse events. A dose-related harm signal at moderate certainty is a stronger finding than anything on the benefit side of this file.

Its risk-of-bias assessment, which is the more damaging half. Across the six Cerebrolysin studies, risk of bias for selective outcome reporting was judged unclear in every one. Generation of the allocation sequence was low risk in one study and unclear in six; allocation concealment low in one and unclear in six; incomplete outcome data low in three and high in four. The manufacturer supported three multicentre studies, either wholly or by supplying drug and placebo, randomisation codes, research grants or statisticians ¹⁵.

One of seven trials has a documented adequate randomisation procedure. That is the foundation the entire meta-analytic literature rests on, including the analyses that report benefit.

The 2019 Cochrane review of vascular dementia, which is positive. Six randomised trials, 597 participants, follow-up from 15 days to three years, conducted in China (three), Russia (one) and Romania (one) ¹¹. Combining Mini-Mental State Examination and ADAS-cog+ data across three studies and 420 people, cognition improved: SMD 0.36, 95% CI 0.13 to 0.58. Global function response improved across two studies and 379 participants: RR 2.69, 95% CI 1.82 to 3.98. Adverse events showed no difference, RR 0.91 (0.29 to 2.85). One trial reported mortality and recorded no deaths ¹¹.

Every one of those estimates is graded very low-quality evidence. Where funding was described, all studies were supported by the pharmaceutical industry. The reviewers wrote that the analyses were limited by heterogeneity, that the included papers had high risk of bias, that if there are benefits the effects may be too small to be clinically meaningful, and that the compound continues to be used and promoted while the supporting evidence base remains weak ¹¹.

This is the strongest efficacy result in the file and it comes with the weakest certainty grade Cochrane issues. Both halves are the finding.

No new vascular dementia trials in six years. The 2019 update identified no studies eligible for inclusion that were not already in the 2013 review ¹¹.

A commercially successful product used across several continents generated no new randomised evidence in its dementia indication between 2013 and 2019. Where a compound sells without new trials, new trials are a cost with no commercial return, and the literature stops moving.

Two independent 2017 meta-analyses, both negative, both explicit. The first pooled 1,649 patients from six randomised trials and found no significant effect on functional recovery at day 90 by any of three measures: modified Rankin Scale response, RR 1.33 (95% CI 0.79 to 2.24, p = 0.28); National Institutes of Health Stroke Scale response, 1.03 (0.83 to 1.28, p = 0.77); Barthel Index response, 0.95 (0.84 to 1.08, p = 0.44). Adverse events, serious adverse events and mortality were likewise null. Its conclusion is that routine administration in acute ischaemic stroke cannot be supported by the available randomised evidence, and its authors declare no conflicts of interest ⁷.

The second, appearing seven weeks later, restricted itself to trials intervening within 72 hours of onset and pooled seven studies and 1,779 patients. It failed to demonstrate significant superiority on the modified Rankin Scale or the Barthel Index, found neutral effects on mortality and serious adverse events, flagged the small number of includable studies as a source of possible publication bias, and concluded that although the preparation appeared safe, routine use to improve long-term rehabilitation after stroke could not be supported by the available evidence ⁸.

Two groups, two independent literature searches, overlapping but non-identical trial sets, and the same sentence at the end of both.

The manufacturer-adjacent meta-analyses. Nine ischaemic stroke trials pooled for global neurological improvement in early post-stroke recovery ¹⁰. The two CARS trials pooled for motor recovery during early rehabilitation ⁹. Randomised double-blind placebo-controlled trials in mild-to-moderate Alzheimer’s disease ³. Several share authors — Vester, Rahlfs, Doppler, Moessler, Guekht, Bornstein, Muresanu — and Doppler and Moessler are affiliated with the manufacturer.

These are peer-reviewed meta-analyses of real randomised trials using declared statistical methods, and they reach conclusions the independent analyses of overlapping trials do not. Both halves of that sentence belong together.

The largest independent pooling, 2025. Fourteen randomised trials and 2,884 patients. The NIHSS mean difference was significant at 1.39 (95% CI 0.53 to 2.25, p = 0.020). Functional independence, defined as a modified Rankin Scale score of 0 to 2, was not: RR 1.31 (95% CI 0.90 to 1.91). Serious adverse events, RR 1.08 (0.84 to 1.40), and mortality, RR 0.86 (0.68 to 1.09), showed no significant difference ¹⁸.

A deficit scale moves and an independence measure does not — the pattern that recurs across this project. Note also that the 2017 analysis found NIHSS null when expressed as a responder risk ratio, 1.03 (0.83 to 1.28) ⁷. Whether this compound improves NIHSS depends on how NIHSS is expressed, which is not a reassuring property for a headline result.

Where the compound sits against its competitors. A 2024 network meta-analysis pooled 35 studies and 18,423 cases across nine interventions: citicoline, edaravone, edaravone dexborneol, cinepazide maleate, cerebrolysin, minocycline, ginkgolide, ginkgo diterpene lactone meglumine and conventional treatment ¹⁷. Cerebrolysin was among the agents associated with reduced mortality against conventional treatment. It was not among those showing a high total treatment effective rate, and it was among those with a higher adverse-effect rate than conventional treatment. Edaravone, ginkgolide and edaravone dexborneol were identified as safest and most effective; cerebrolysin was not in that group.

Direct comparison against eight alternatives is a test most compounds in this project never face. Its adverse-effect finding runs the same direction as the 2023 Cochrane harm signal ¹⁵, which is worth noting given that the dedicated placebo-controlled safety analyses found nothing ^7,8,13,18.

Other indications. A systematic review and meta-analysis in traumatic brain injury reported findings supporting beneficial effects on clinical outcome ¹⁴. A systematic review and meta-analysis in subarachnoid haemorrhage, by the same Polish group, noted that no prior meta-analysis had addressed that population ¹⁶.

Two 2023 meta-analyses in new indications, published while the stroke question was still open enough to warrant a seventh Cochrane update in the same year ¹⁵.

The independent verdict in vascular cognitive impairment. A 2025 review extracted 173 trials covering 22,347 participants, four categories of vascular cognitive impairment, 91 interventions and 145 outcomes ¹⁹. Cerebrolysin showed small to moderate cognitive improvement — as did acetylcholinesterase inhibitors, memantine, propentofylline, physical exercise and cognitive rehabilitation. Ginkgo biloba extracts showed the largest effects. The authors report substantial methodological heterogeneity, state that few interventions were reproducibly tested in adequately sized and designed studies, and describe certainty as low overall. None declares a conflict relating to this compound.

Published in a major dementia journal by a group with no stake in any of the 91 interventions it assessed. It places Cerebrolysin in a crowded field of modest, uncertain effects, behind a botanical extract and level with unpaid interventions like exercise.

05 · Evidence overview

06 · Safety profile

The harm signal. The 2023 Cochrane review found moderate-certainty evidence of an increase in non-fatal serious adverse events: RR 2.39, 95% CI 1.10 to 5.23, across three trials and 1,335 participants. In the subgroup receiving 30 mL daily for ten days — a cumulative 300 mL — the estimate was 2.87, 95% CI 1.24 to 6.69, across two trials and 1,189 participants ¹⁵. The reviewers state the finding as a potential increase, and the confidence intervals are wide, but the point estimates are above two and the certainty grade is the highest assigned anywhere in this file.

The null findings, which are numerous and independent. All-cause death, RR 0.96 (0.65 to 1.41), six trials, 1,689 participants, moderate certainty ¹⁵. Total serious adverse events, RR 1.16 (0.81 to 1.66) ¹⁵. Total adverse events, RR 1.03 (0.92 to 1.14), four trials, 1,607 participants, low certainty ¹⁵. A dedicated safety meta-analysis of twelve randomised trials and 2,202 patients found no statistically significant differences from placebo across main and subgroup analyses ¹³. The fourteen-trial pooling reports serious adverse events at 1.08 (0.84 to 1.40) and mortality at 0.86 (0.68 to 1.09) ¹⁸. Both 2017 analyses agree ^7,8. In vascular dementia, adverse events showed no difference, RR 0.91 (0.29 to 2.85), on very low-quality evidence ¹¹.

Reading the two together. Total serious adverse events do not differ; fatal ones do not differ; the non-fatal subset does. The pooled analyses that report reassuring safety generally report totals, which is where a shift within a total can hide. The 2024 network meta-analysis, comparing against active agents rather than placebo, also places Cerebrolysin among those with a higher adverse-effect rate than conventional treatment ¹⁷. No publication reconciles these, and this guide is not in a position to.

What is genuinely unknown. What the preparation contains — no composition specification, batch assay or constituent list appears in any paper cited here, which means the reproducibility of every trial depends on manufacturing consistency that has not been published. Whether the non-fatal serious adverse events driving the 2.39 estimate share a mechanism or a body system: the Cochrane abstract does not enumerate them, and this guide has not read the full review. Whether the dose-response appearance at 300 mL is real or an artefact of two trials. Immunogenicity and transmissible agent risk from porcine brain sourcing — standard considerations for any animal-tissue-derived biological and absent from this literature entirely, which is conspicuous for a central nervous system tissue product. Long-term exposure beyond the three-year maximum follow-up in the dementia trials ¹¹. Whether death or dependence is affected at all, since no stroke trial has reported it ¹⁵. And the composition and provenance of any material obtained outside the manufacturer’s supply chain, on which nothing is published.

07 · US regulatory status

Current as of 6 September 2026. Cerebrolysin is not approved as a drug in the United States and is not a controlled substance there.

It is manufactured commercially by EVER Neuro Pharma GmbH, formerly Ebewe Pharma ⁴, and Cochrane records that it is widely used for acute ischaemic stroke in Russia, Eastern Europe, China and other Asian and post-Soviet countries ¹⁵. This guide does not enumerate those authorisations or state their terms, because doing so accurately would require regulatory records that have not been consulted. That limitation is why the compound is graded Mixed Evidence rather than Approved Pharma, and it is logged.

Under the World Anti-Doping Code, Cerebrolysin does not appear as a named prohibited substance in the classes reviewed for this guide. Competitors should consult the current Prohibited List directly rather than rely on secondary summaries, including this one.

08 · Limitations of the evidence

1. The highest-certainty finding in the entire literature is a harm, not a benefit. Moderate- certainty evidence of increased non-fatal serious adverse events, RR 2.39 (95% CI 1.10 to 5.23), with a larger estimate at higher cumulative dose ¹⁵. Nothing on the efficacy side of this file carries a comparable certainty grade in the opposite direction.
2. The positive results carry the weakest certainty grade Cochrane issues. Cognition, SMD 0.36, and global function, RR 2.69, in vascular dementia are both graded very low-quality evidence, drawn from 420 and 379 people respectively ¹¹. Very low certainty means the true effect is likely to be substantially different from the estimate.
3. No stroke trial has ever reported the outcome that matters. Not one included study reported poor functional outcome — death or dependence at the end of follow-up — nor early death, quality of life, or time to restoration of capacity for work ¹⁵. Every efficacy claim in acute stroke rests on rating scales rather than on whether people ended up dependent.
4. Independent replication free of manufacturer involvement is not established. The manufacturer supported three of the seven stroke trials, in some cases supplying drug, placebo, randomisation codes, grants and statisticians ¹⁵; every vascular dementia trial with identifiable funding was industry-supported ¹¹. Meta-analysis cannot manufacture independence that the underlying trials lack.
5. Randomisation quality is documented in one trial out of seven. Allocation sequence generation was low risk in one study and unclear in six; allocation concealment the same; incomplete outcome data was high risk in four of seven; selective outcome reporting was unclear across all ¹⁵.
6. Independent and manufacturer-adjacent analyses of overlapping trials reach different conclusions. Two unaffiliated 2017 meta-analyses concluded outright that routine use is not supported ^7,8; manufacturer-affiliated meta-analyses report benefit ^3,9,10. No publication reconciles them.
7. Publication bias is named as a concern by the reviewers themselves. The 2017 analysis restricted to trials starting within 72 hours flags the small number of includable studies as a source of possible publication bias and of inaccurate between-study variance ⁸. Where a trial base is small, commercially sponsored and reported across decades, the studies that reached print are not a random sample of those that were run.
8. It is a mixture with no published specification. A porcine brain digest ¹⁵ cannot be characterised by molecular weight, sequence or purity, and no paper cited here provides a composition, an assay or a batch standard. Every trial result depends on manufacturing consistency that has not been demonstrated in the literature — and the current Cochrane review pools it with other “Cerebrolysin-like agents” including Cortexin ¹⁵, because at this level of definition the category has no sharp edge.
9. The deficit scale and the independence measure disagree, and the deficit scale disagrees with itself. The largest independent pooling found NIHSS significant as a mean difference and functional independence not ¹⁸; the 2017 analysis found NIHSS null as a responder risk ratio ⁷.
10. Where an independent effect is reported, it is small and not distinctive. Small to moderate cognitive improvement, level with acetylcholinesterase inhibitors, memantine, physical exercise and cognitive rehabilitation, and behind Ginkgo biloba extracts in the same analysis ¹⁹.
11. The evidence base stopped growing in one indication. No vascular dementia trial became eligible for Cochrane inclusion between 2013 and 2019 ¹¹.
12. The mechanism is a resemblance claim. Cochrane’s own phrasing is that the preparation is said to have neurotrophic and neuroprotective activity ¹¹. No paper here identifies an active constituent, a receptor, or a species-specific effect.
13. Porcine brain sourcing raises questions this literature does not address. Immunogenicity and transmissible agent risk are standard considerations for animal-tissue-derived biologicals and are absent from the record entirely.
14. This guide has read abstracts, not full texts. For a compound whose central problem is disagreement between analyses, the methodological detail that would adjudicate sits in the full texts of nine Cochrane reviews and several meta-analyses, none of which has been read here. In particular, the composition of the non-fatal serious adverse events behind the 2.39 estimate is not available from the abstract.
15. Nothing is known about non-manufacturer material. No published analysis has examined the identity or composition of any Cerebrolysin-labelled product obtained outside the commercial supply chain.

09 · References

1. Ziganshina LE, Abakumova T, Kuchaeva A. Cerebrolysin for acute ischaemic stroke. Cochrane Database Syst Rev. 2010 Apr 14;(4):CD007026.

   PMID 20393954 ↗

2. Chen N, Yang M, Guo J, Zhou M, Zhu C, He L. Cerebrolysin for vascular dementia. Cochrane Database Syst Rev. 2013 Jan 31;(1):CD008900.

   PMID 23440834 ↗

3. Gauthier S, Proaño JV, Jia J, Froelich L, Vester JC, Doppler E. Cerebrolysin in mild-to-moderate Alzheimer’s disease: a meta-analysis of randomized controlled clinical trials. Dement Geriatr Cogn Disord. 2015;39(5–6):332–347.

   PMID 25832905 ↗

4. Ziganshina LE, Abakumova T. Cerebrolysin for acute ischaemic stroke. Cochrane Database Syst Rev. 2015 Jun 17;(6):CD007026.

   PMID 26083192 ↗

5. Ziganshina LE, Abakumova T, Vernay L. Cerebrolysin for acute ischaemic stroke. Cochrane Database Syst Rev. 2016 Dec 5;12(12):CD007026.

   PMID 27918088 ↗

6. Ziganshina LE, Abakumova T, Vernay L. Cerebrolysin for acute ischaemic stroke. Cochrane Database Syst Rev. 2017 Apr 21;4(4):CD007026.

   PMID 28430363 ↗

7. Wang Z, Shi L, Xu S, Zhang J. Cerebrolysin for functional recovery in patients with acute ischemic stroke: a meta-analysis of randomized controlled trials. Drug Des Devel Ther. 2017 Apr 19;11:1273–1282.

   PMID 28458521 ↗

8. Zhang D, Dong Y, Li Y, Chen J, Wang J, Hou L. Efficacy and safety of Cerebrolysin for acute ischemic stroke: a meta-analysis of randomized controlled trials. Biomed Res Int. 2017 Jun;2017:4191670. Epub 2017 Jun 5.

   PMID 28656143 ↗

9. Guekht A, Vester J, Heiss WD, Gusev E, Hoemberg V, Rahlfs VW, Bajenaru O, Popescu BO, Doppler E, Winter S, Moessler H, Muresanu D. Safety and efficacy of Cerebrolysin in motor function recovery after stroke: a meta-analysis of the CARS trials. Neurol Sci. 2017 Oct;38(10):1761–1769.

   PMID 28707130 ↗

10. Bornstein NM, Guekht A, Vester J, Heiss WD, Gusev E, Hömberg V, Rahlfs VW, Bajenaru O, Popescu BO, Muresanu D. Safety and efficacy of Cerebrolysin in early post-stroke recovery: a meta-analysis of nine randomized clinical trials. Neurol Sci. 2018 Apr;39(4):629–640.

    PMID 29248999 ↗

11. Cui S, Chen N, Yang M, Guo J, Zhou M, Zhu C, He L. Cerebrolysin for vascular dementia. Cochrane Database Syst Rev. 2019 Nov 11;2019(11):CD008900.

    PMID 31710397 ↗

12. Ziganshina LE, Abakumova T, Hoyle CH. Cerebrolysin for acute ischaemic stroke. Cochrane Database Syst Rev. 2020 Jul 14;7(7):CD007026.

    PMID 32662068 ↗

13. Strilciuc S, Vécsei L, Boering D, Pražnikar A, Kaut O, Riederer P, Battistin L. Safety of Cerebrolysin for neurorecovery after acute ischemic stroke: a systematic review and meta-analysis of twelve randomized-controlled trials. Pharmaceuticals (Basel). 2021 Dec 13;14(12):1297.

    PMID 34959697 ↗

14. Jarosz K, Kojder K, Andrzejewska A, Solek-Pastuszka J, Jurczak A. Cerebrolysin in patients with TBI: systematic review and meta-analysis. Brain Sci. 2023 Mar 17;13(3):507.

    PMID 36979317 ↗

15. Ziganshina LE, Abakumova T, Nurkhametova D, Ivanchenko K. Cerebrolysin for acute ischaemic stroke. Cochrane Database Syst Rev. 2023 Oct 11;10(10):CD007026.

    PMID 37818733 ↗

16. Kojder K, Jarosz K, Bosiacki M, Andrzejewska A, Zacha S, Solek-Pastuszka J, Jurczak A. Cerebrolysin in patients with subarachnoid hemorrhage: a systematic review and meta-analysis. J Clin Med. 2023 Oct 20;12(20):6638.

    PMID 37892776 ↗

17. Li M, Huo X, Chang Q, Liu X, Zhang J, Mao Z. Efficacy analysis of neuroprotective drugs in patients with acute ischemic stroke based on network meta-analysis. Front Pharmacol. 2024 Nov 7;15:1475021.

    PMID 39575393 ↗

18. Patel PN, Mangal D, Patel K. Safety and efficacy of Cerebrolysin for neurorecovery after acute ischemic stroke: a systematic review and meta-analysis of 14 randomized controlled trials. Cureus. 2025 Aug 26;17(8):e91054.

    PMID 41018475 ↗

19. Masserini F, Gendarini C, Baso G, Salvadori E, Pantoni L. Therapeutic strategies in vascular cognitive impairment: a systematic review and meta-analysis. Alzheimers Dement. 2025 Nov;21(11):e70840.

    PMID 41198594 ↗

20. Chan TYH, Ma BY, Hung TK, Wong JSY, Lo BWY. Microglial polarization and therapeutic strategies in post-stroke neuroinflammation. Neurol Ther. 2025 Dec;14(6):2277–2293.

    PMID 40971139 ↗